PROSTATE CANCER DETECTION METHODS

Abstract

The present invention provides methods of detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer comprising determining the average methylation ratio at 10 or more genomic regions as set out in the application, and associated methods of selecting a treatment or ascertaining whether a treatment is effective. The present invention also provides a method for determining a solid cancer circulating free DNA (cfDNA) methylome signature for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of the solid cancer in a sample obtained from a subject comprising determining the average methylation ratio at 10 or more genomic regions as set out in the application.

Claims

1. A method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; calculating a methylation score using the average methylation ratio for each of the genomic regions; analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

2. The method of claim 1, wherein each of the genomic regions is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 characterized methylome sequences.

3. The method of claim 1 or 2, wherein each of the genomic regions is covered by at least 10 sequence reads, for example at least 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 sequence reads, and preferably wherein each sequence read or the majority of the sequence reads (for example at least 50%, 60%, 70%, 80% or 90% of the sequence reads) are from different characterized methylome sequences.

4. The method of any one of claims 1 to 3, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises: determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

5. The method of any one of claims 1 to 4, wherein analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores, wherein a reference methylation score is a methylation score calculated for the same genomic regions (for example, calculated using the average methylation ratio for the same genomic regions) in one or more of the following a cfDNA sample from a healthy subject, for example a healthy age-matched subject; a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; a sample of white blood cells from a subject, for example the subject or a healthy subject; a cfDNA sample from a different subject having prostate cancer, preferably wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and more preferably wherein each cfDNA sample has a different level of prostate cancer fraction); a characterized methylome sequence of a white blood cell; a characterized methylome sequence of a prostate cancer cell line; a characterized methylome sequence of a cancerous prostate cell; and/or a characterized methylome sequence of a non-cancerous prostate cell.

6. The method of any one of claims 1 to 5, comprising determining the average methylation ratio at 25 or more, 50 or more, 100 or more, 150 or more, 200 or more, 300 or more, 400 or more, 500 or more, 600 or more, 700 or more, 800 or more, or 900 or more genomic regions (for example comprising determining the average methylation ratio at 25, 50, 100, 150, 200, 300, 400, 500, 600, 700, 800, 900 or 1000 genomic regions).

7. The method of any one of claims 1 to 6, wherein the genomic regions have a 100 bp genomic location defined in any one of Tables 1 to 4, Table 5, Table 6 or Table 7.

8. The method of any one of claims 1 to 7, wherein at least 25% of the genomic regions are prostate tissue specific genomic regions.

9. The method of any one of claims 1 to 8, wherein the prostate cancer is acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer (for example wherein the prostate cancer is acinar adenocarcinoma prostate cancer or ductal adenocarcinoma prostate cancer).

10. The method of any one of claims 1 to 9 wherein the prostate cancer is castration resistant prostate cancer and/or is metastatic prostate cancer.

11. The method of any one of claims 1 to 10, wherein the sample comprising cfDNA is a blood or plasma sample.

12. The method of any one of claims 1 to 11, further comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of prostate cancer fraction in the two samples.

13. The method of any one of claims 1 to 12, further comprising treating the subject for prostate cancer using a therapeutic agent for the treatment of prostate cancer; or ceasing or altering treatment with a therapeutic agent for the treatment of prostate cancer; or initiating a non-therapeutic agent treatment for prostate cancer (for example initiation of treatment by surgery or radiation).

14. An in-vitro diagnostic kit for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer, comprising one or more reagents for detecting the presence or absence of at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4, or comprising at least one CpG locus defined in Table 5, or comprising at least one CpG locus defined in Table 6, or comprising at least one CpG locus defined in Table 7.

15. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform the method of any one of claims 1 to 12; or a computer-executable software for performing the method of any one of claims 1 to 12 or a computer-implemented method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in the sample; and wherein the computer readable medium comprises instructions that, when executed by the processor, causes the computer to perform a method of any one of claims 1 to 12

16. A therapeutic agent for the treatment of prostate cancer for use in the treatment of prostate cancer, whereby i) the method of any one of claims 1 to 12 is performed to determine the level of prostate cancer prostate cancer DNA in a subject; ii) the therapeutic agent is administered if the subject has a level of prostate cancer.

17. A method of determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer comprising performing the method of any one of claims 1 to 12; determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to the level of prostate cancer, whereby one therapeutic agent is suitable for a subject with no level of prostate cancer fraction (for example an undetectable level of prostate cancer fraction) or a level of prostate cancer fraction of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of prostate cancer DNA (for example a percentage level of prostate cancer fraction of at least 0.01%); or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with no level of prostate cancer DNA (for example an undetectable level of prostate cancer DNA) or a level of prostate cancer DNA of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of prostate cancer DNA (for example a percentage level of prostate cancer fraction of at least 0.01%).

18. A method or therapeutic agent as claimed in any one of claim 16 or 17, wherein the therapeutic agent for the treatment of prostate cancer is selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, a chemotherapy agent.

19. A method for determining a solid cancer circulating free DNA (cfDNA) methylome signature for use in detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of the solid cancer, the method comprising: (i) characterizing the methylome sequence of a plurality of cfDNA molecules in a first sample comprising cfDNA from a subject known to have the solid cancer, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; (ii) determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample by aligning the methylome sequences; (iii) determining the methylation ratio of each CpG locus and/or average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample; repeating steps (i) to (iii) for one or more further samples comprising cfDNA each from subjects known to have the solid cancer; performing a variance analysis of all or a selection of the methylation ratios of the CpG loci and/or all or a selection of average methylation ratios of the genomic regions of the samples; selecting a group of CpG loci and/or genomic regions associated with a feature of the samples; selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature.

20. The method of claim 19, wherein the solid cancer is prostate cancer.

21. The method of claim 19 or 20, wherein the variance analysis performed is a dimensionality reduction.

22. The method as claimed in claim 21 wherein the variance analysis performed is a principal component analysis.

23. The method as claimed in claim 22, wherein selecting a group of CpG loci and/or genomic regions associated with a feature of the samples comprises selecting one of principal component 1, principal component 2, principal component 3, principal component 4, principal component 5, principal component 6, principal component 7, principal component 8 or a higher principal component.

24. The method of any one of claims 18 to 23, wherein selecting the CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting the CpG loci and/or genomic regions in the group that have strong association with the feature, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions most correlated with the feature in the group (for example selecting CpG loci and/or genomic regions that are within the top 8000, 5000, 3000, 2000, 1000, 800, 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions most correlated with the feature in the group).

25. The method of any one of claims 18 to 24, wherein selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting at least 5 CpG loci (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) and/or at least 5 genomic regions (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) in the group to provide a cfDNA methylome signature.

26. The method of claim 22 or 23, or claim 24 or 25 when dependent on claim 22 or 23, wherein selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting a plurality of CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 5000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 4000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 3000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 2000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8.

27. A method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus, and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; calculating a methylation score using the average methylation ratio for each of the genomic regions; analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

28. The method of claim 27, wherein the prostate cancer is an androgen-insensitive subtype of the prostate cancer.

Description

DESCRIPTION OF THE DRAWINGS

[0027] FIGS. 1 and 2 shows the patient characteristics of the plasma samples used in Example 1.

[0028] FIG. 3 shows the targeted methylome sequencing matrix (total reads, mapped reads, % mapped reads, % bisulfite conversion).

[0029] FIG. 4 shows the LP-WGBS matrix (total reads, mapped reads, % mapped reads, % bisulfite conversion).

[0030] FIG. 5 shows a plot showing coverage distribution in target regions by bisulfite high-coverage next-generation sequencing (NGS) in plasma samples.

[0031] FIG. 6 shows the fraction of data to drop on different window sizes (10 bp, 100 bp, 1000 bp, 10000 bp).

[0032] FIG. 7 shows the distribution of methylation ratio by different segment size (10 bp, 100 bp, 1,000 bp, 10,000 bp).

[0033] FIG. 8 shows the correlation of median methylation ratio of selected segments with genomically-determined tumour fraction. Y-axis shows the correlation value and the X-axis denotes the number of top correlated segments.

[0034] FIG. 9 shows the correlation of methylation ratios of selected segments. Y-axis shows the standard deviation and the X-axis denotes the number of top correlated segments.

[0035] FIG. 10 shows a schematic overview of the work-flow for integrating NGS of the plasma methylome and genome of Example 1.

[0036] FIG. 11 shows the genomically-determined tumour fraction in baseline and progression samples from pre- and post-chemotherapy patients receiving abiraterone or enzalutamide.

[0037] FIG. 12 shows a box plot showing methylation ratio distribution for baseline (A) and progression (B) plasma samples and white blood cells (C) presented separately.

[0038] FIG. 13 shows the genomic annotation based on location of methylation segments in the custom targeted panel and in segments covered >10× in all 19 baseline samples.

[0039] FIG. 14 shows the methylation ratio density (upper panel) and Quantile-Quantile plot (bottom panel) analysis based on the genomic annotation of methylation segments in promoter or other regions. Data from white blood cells (WBC) or plasma collected at baseline (BL) or progression (PD) from mCRPC patients or from healthy volunteers (HV) are presented separately. In the bottom panel, the upper line in both plots that diverges from the course of the other two lines corresponds to the HV data, the other two tracking lines in both plots correspond to the BL and PD data.

[0040] FIG. 15 shows a schematic workflow of methylation data analysis of Example 1.

[0041] FIG. 16 shows a bar-chart showing the variance associated to each Principal Component (PC) (black columns show significant principal components); and a scree plot (the dotted line) indicating cumulative explained variance.

[0042] FIG. 17 shows the correlation between PCs and tumour fraction (bottom panel). Size and the colour of each circle show Pearson correlation and background shading denotes P value).

[0043] FIG. 18 shows the correlation of genomically determined tumour fraction (y-axis) and principal component 1 (PC1) values (x-axis) from high-coverage targeted methylation sequencing on 19 baseline, 16 progression plasma samples, and control samples (n=4 healthy volunteer plasma samples, LNCaP prostate cancer cell line).

[0044] FIG. 19 shows the functional enrichment analysis of genes (n=231) in ct-MethSig segments. The p-value was corrected for multiple statistical testing (Benjamini-Hochberg).

[0045] FIG. 20 shows the Bland-Altman plot showing agreement for tumour fraction estimation by genomically-determined tumour fraction and on LP-WGBS.

[0046] FIG. 21 shows the top 1000 segments (ct-MethSig) with the highest correlation coefficient between PC1 and methylation ratio.

[0047] FIG. 22 shows the ct-MethSig methylation ratio distribution by patient plasma sample split by negatively correlated (i.e. hypermethylated) and positively correlated (i.e. hypomethylated) segments.

[0048] FIGS. 23A to 23C shows the methylation ratios of GSTP1 (FIG. 23A), APC (FIG. 23B), and RASSF1A (FIG. 23C) across different tissue types—healthy volunteer plasma, white blood cells, CRPC plasma samples, LNCaP cell line.

[0049] FIGS. 24A and 24B show the ct-MethSig segment methylation ratio derived from mCRPC tissues lined by tumour fraction (A: negatively correlated (i.e. hypermethylated) segments; B: positively correlated (i.e. hypomethylated) segments).

[0050] FIG. 25 shows the correlation between HSPC tissue tumour fraction estimation by ct-MethSig and molecularly-defined tumour fraction.

[0051] FIG. 26 shows a Venn diagram showing the overlap of negatively (dark blue/darker shading) correlated genes in ct-MethSig segments with targets of EED, SUZ12, and ES (Embryonic Stem cells) with H3K27ME3 marks. The numbers highlighted in white bold denote the number of genes in the ct-MethSig negatively correlated group.

[0052] FIG. 27 shows the permutation test on genes overlapping with ct-MethSig; the dot represents the gene enrichment test (Fisher Exact test) P value in genes overlapping with ct-MethSig and the box represents P values of the permutation test with 1000-time iteration.

[0053] FIG. 28 shows the circulating tumour fraction methylation signature comprises segments specific to either normal or malignant prostate epithelium. Left panel: Methylation ratios of ct-MethSig negatively (i.e. hypermethylated, N=520) and positively (i.e. hypomethylated, N=480) correlated group from LNCaP (N=4), healthy volunteer (H.V., N=4), and normal prostate epithelium (PrEC). The right panel shows ct-MethSig negative (i.e. hypermethylated) and positive (i.e. hypomethylated) groups can be split into prostate cancer specific segments and prostate epithelium specific.

[0054] FIG. 29 shows the CASCADE patient and sample characteristics.

[0055] FIGS. 30A and B shows the methylation ratio distribution of circulating normal prostate specific or prostate cancer specific component in localized prostate cancer from TCGA.

[0056] FIG. 31 shows the top 1000 segments with highest correlation coefficient between the third principal component (PC3) and methylation ratio.

[0057] FIG. 32 shows the methylation ratio of top 1000 segments highly correlated with PC3 values derived from plasma, white blood cell, HSPC tissue, and CRPC metastases (CASCADE trial).

[0058] FIG. 33 shows the comparison of intra-individual changes in the top correlated segments defined by targeted methylation NGS on plasma DNA and changes in tumour fraction. Y-axis denotes the difference (A) of mean methylation ratio of the top correlated segments between baseline and progression samples and the X-axis denotes the difference (A) in tumour fraction.

[0059] FIG. 34 shows the median methylation ratio of the top correlated segments of different metastatic sites by patient from the CASCADE rapid warm autopsy program.

[0060] FIG. 35 shows the median methylation ratio of 993 MethSig3 segments positively (i.e. hypomethylated) correlated with PC3 values across different sample types—plasma, white blood cells, cell lines (LNCaP, LNCaP95, VCaP), CASCADE tumours (mCRPC biopsies) and CSPC tumours are plotted against the median methylation ratio of top correlated segments with ct-MethSig.

[0061] FIG. 36A shows the genes overlapping with AR-MethSig; and FIG. 36B shows the functional enrichment of top correlated segments with principal component 3 (PC3).

[0062] FIG. 37 shows the AR binding motif that is over-represented in regions adjacent to the top correlated segments (top panel). The consensus AR binding motif is shown as a reference (bottom panel).

[0063] FIG. 38 shows the performance of Gaussian Mixture Model (k-fold cross-validation, k=100).

[0064] FIG. 39 shows copy number alteration plots from LP-WGS on plasma DNA with and without bisulfite treatment.

[0065] FIG. 40 shows the prevalence of gain and loss events lined by chromosome position extracted from LP-WGBS on mCRPC plasma samples.

[0066] FIG. 41 shows the analysis of copy number profiles on low-pass whole genome bisulfite sequencing. Matrix shows gains (red) and losses (blue) ordered by chromosomal position (columns) for individual patient samples (one per row) ordered by tumour fraction. Bar chart on the left shows tumour fraction per sample. Bar chart on the right shows the number of gain (red) or loss (blue) events per sample.

[0067] FIG. 42 shows the contingency tables showing ct-MethSig and AR-MethSig segments in copy number aberrant regions.

[0068] FIG. 43 shows a Manhattan plot showing the level of significance of the association between PC1 value distribution and copy number alterations ordered by chromosome position. The segment containing AR is circled with a dotted line (not significant, P=0.18). Dark dots represent that the segment belongs to the odd numbered chromosome indicated, and light dots represent that the segment belongs to the even numbered chromosome indicated.

[0069] FIG. 44 shows a Manhattan plot showing the level of significance of the association between PC3 value distribution and copy number alterations ordered by chromosome position. The segment containing AR is highlighted circled with a dotted line (P=0.018, Kruskal-Wallis test). Dark dots represent that the segment belongs to the odd numbered chromosome indicated, and light dots represent that the segment belongs to the even numbered chromosome indicated.

[0070] FIG. 45 shows the methylation ratio of AR-MethSig segments of AR gain and non-gain groups

[0071] FIG. 46 shows a Bland-Altman plot showing agreement between targeted methylation NGS and LP-WGBS on AR-MethSig median methylation ratio.

[0072] FIG. 47 shows the overall survival analysis (start of ADT to death) for AR-MethSig low group versus AR-MethSig high group (Mantel-Cox log-rank test). The line in the graph that extends beyond 150 months on ADT corresponds to AR-MethSig high, the other line corresponds to AR-MethSig low.

[0073] FIG. 48 shows the correlation of genomically-determined tumour fraction and PC1 values derived from PCA on different window sizes (10 bp, 100 bp, 1000 bp, 10000 bp).

[0074] FIG. 49 shows a schematic of the workflow of building a classification model of Example 1.

[0075] FIG. 50A shows the accuracy of Random Forest Classification model (=10) on 1000-time cross validation; FIG. 50B shows the accuracy of RFC (number of trees in the forest=100) on 1000-time cross validation.

[0076] FIGS. 51A to 51D show the accuracy of RFC model (number of trees in the forest=100) on 1000-time cross validation trained on 1 (FIG. 51A), 10 (FIG. 51B), or 100 (FIG. 51C), randomly (rdm) selected ct-MethSig segments or the ct-MethSig segments (FIG. 51D).

DEFINITIONS

[0077] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly recognized by one of ordinary skill in the art to which this invention belongs.

[0078] As used herein “DNA methylation” refers to the addition of a methyl group to a DNA nucleotide. DNA methylation most commonly occurs on the 5′ carbon of cytosine residues (i.e. 5-methylcytosines) of a CpG dinucleotide (referred to herein as a “CpG locus”). DNA methylation may also occur in cytosines in other contexts, for example CHG and CHH, where H is adenine, cytosine or thymine. Cytosine methylation may also be in the form of 5-hydroxymethylcytosine. Non-cytosine methylation, such as N6-methyladenine, may also occur.

[0079] As used herein, the term “CpG locus” refers to a region of DNA where a cytosine nucleotide is followed by a guanine nucleotide in the linear sequence of bases along its 5′ to 3′ direction. A CpG site can become methylated in human and other animal DNA.

[0080] As used herein, a “methylome” is the set of nucleic acid methylation modifications in a subject's genome in a particular cell, tissue or cancer. The methylome may correspond to all of the genome, a substantial part of the genome, or relatively small portion(s) of the genome.

[0081] As used herein the term “plasma methylome” is the methylome determined from the plasma or serum of a subject (e.g., a human). The plasma methylome is an example of a “cell-free DNA methylome” since plasma includes cfDNA. The plasma methylome is an example of a mixed methylome because the plasma may comprise cfDNA from a variety of sources, for example, cfDNA from different tissues, non-cancerous and cancerous tissues.

[0082] As used herein the term “methylation profile” is the information related to DNA methylation for a DNA molecule. Information related to DNA methylation can include, but not limited to, a methylation index of a CpG locus, a methylation density of CpG sites in a DNA molecule, a distribution of CpG sites over a contiguous region, a pattern or level of methylation for each individual CpG site within a region that contains more than one CpG site, and non-CpG methylation.

[0083] As used herein the term “methylome sequence” is the DNA sequence and the methylation profile of the whole or a portion of a DNA molecule, for example a cfDNA molecule. For example, the methylome sequence may be the methylome sequence of the whole or a portion of a cfDNA molecule. The methylome sequence may correspond to all of the genome, a substantial part of the genome, or portion(s) of the genome.

[0084] As used herein the term “circulating free DNA” (cfDNA) means the DNA fragments that have been released into the blood plasma and are found freely circulating the blood stream, as well as in the urine. cfDNA is generally double-stranded DNA consisting of small fragments (70 to 200 bp).

[0085] As used herein the term “sequence read” refers to a sequence of the base pairs inferred from the whole or a portion of single molecule of DNA, for example the whole or a portion of a single molecule of cfDNA. A single read may be of 20 to 500 base pairs, or even up to 1500 base pairs. The sequence of a specific single molecule of DNA may be read once or read multiple times and each sequence is taken to be representative of a single molecule of DNA.

[0086] As used herein the term “tumour fraction cfDNA” is cfDNA derived from DNA of a cancer cell. As used herein the term “prostate cancer fraction cfDNA” is cfDNA derived from DNA of a prostate cancer cell.

[0087] As used herein, the term “genomic region” refers to a region of a genome, e.g. the genome of a subject, for example a human. A genomic region may also be referred to as a “segment”. It may be referred to using the genomic location of the region, for example using the coordinates of the start position and end position of the location in a specific chromosome. For a human subject a genomic region is suitably described by a genomic location, and in particular a genomic location with reference to a reference genome (for example, a digital nucleic acid sequence database, assembled a representative example of a species' set of genes).

[0088] As used herein, the term “genomic location” refers to the location of a region of a genome, e.g. the genome of a subject, for example a human. It may be referred to using the coordinates of the start position and end position of the location in a specific chromosome. For a human subject a genomic location is suitably described by reference to a reference genome (for example, a digital nucleic acid sequence database, assembled from a representative example of a species' set of genes). For example, for a human subject, with reference to the human reference genome GRCh37 (also referred to as Human Genome 19 (hg19)) or human reference genome GRCh38 (also referred to as Human Genome 38 (hg38)). For the present inventions, preferably the reference genome is human reference genome GRCh37 (also known as hg19). As such, a genomic location for a human may be described using the coordinates of the start position and end position of the location in a specific chromosome with reference to the Genome Reference Consortium Human Build 37 (GRCh37) (also referred to as Human Genome 19 (hg19)). Suitably, a genomic location according to the present invention is a location that covers 2 to 200 bp of DNA. A genomic location according to the present invention preferably includes at least one CpG locus, and suitably includes at least two CpG loci, for example 2, 3, 4, 5, 6, 7 or 8 CpG loci, and preferably 2, 3, 4, 5 or 6 CpG loci.

[0089] As used herein the term “plurality” is at least 2, for example at least 10, at least 100, at least 1000, at least 10,000, at least 100,000, at least 10.sup.6, at least 10.sup.7, at least 10.sup.8 or at least 10.sup.9 or more.

[0090] As used herein the term “a level of prostate cancer fraction” is the level of cfDNA derived from prostate cancer cells in a cfDNA sample compared to the cfDNA that is not derived from prostate cancer cells. cfDNA that is not derived from prostate cancer cells in a cfDNA sample may be derived from blood cells, for example white blood cells (leukocyte), and other non-prostate tissues.

[0091] As used herein the term “blood cell fraction cfDNA” is cfDNA derived from DNA of a blood cell, for example a white blood cell (leukocyte).

[0092] As used herein, a “subject” refers to an animal, including mammals such as humans. Preferably, the subject is a human subject. As used herein, an “individual” can be a subject. As used herein, a “patient” refers to a human subject. In one embodiment, the subject is known or suspected to have a cancer (for example prostate cancer), and/or is known or suspected to have a risk of developing cancer (for example prostate cancer), or is known to have cancer and is known or suspected to have metastatic cancer (for example prostate cancer) or to have a risk of developing metastatic cancer (for example prostate cancer). In some embodiments, the subject is a subject who has been identified as being at risk of developing a cancer, in particular at risk of developing a prostate cancer.

[0093] As used herein, a “healthy subject” refers to a subject that has not been diagnosed with a type of cancer (for example prostate cancer), and preferably has not been diagnosed with any type of cancer. Thus, for example, a method of relating to prostate cancer, a “healthy subject” has no prostate cancer, and preferably no other type of cancer. Preferably, a healthy subject has not been diagnosed with a type of cancer (for example prostate cancer), and is not suspected of having a type of cancer, and suitably has not been diagnosed with any type of cancer (for example prostate cancer), and is not suspected of having any type of cancer.

[0094] The term “sample” as used herein means a biological sample derived from a patient to be screened in a method of the invention. The biological sample may be any suitable sample known in the art in which cfDNA can be detected and/or isolated. Included are individual cells and cell populations obtained from bodily tissues or fluids. Examples of suitable body fluids that may be used as samples according to the present invention are plasma, blood, and urine.

[0095] As used herein the term “methylation ratio” refers to the proportion of cytosine residues (C) that are methylated at all sequence reads covering a CpG locus (“G” is a guanine residue) within a population or pool of DNA, such as a sample of cfDNA obtained from the plasma of a subject. When the methylation profile is measured using bisulfite conversion the un-methylated CpG loci are converted to UpG (“U” is a uracil residue), while methylated CpG sites remain the same. The uracil residues are read as thymine residues during the DNA sequencing step following bisulfite conversion. The methylation ratio may be calculated using formula (X), which take the cytosine (C) and thymidine (T) counts from multiple sequence reads of a specific CpG locus:

[00001]$\begin{array}{cc}\mathrm{Methylation}\mathrm{Ratio}=\frac{\begin{array}{c}\mathrm{cytosine}\left(C\right)\mathrm{counts}\mathrm{from}\\ \begin{array}{c}\mathrm{all}\mathrm{sequence}\mathrm{reads}\\ \mathrm{of}a\mathrm{CpG}\mathrm{locus}\end{array}\end{array}}{\begin{array}{c}\mathrm{cytosine}\left(C\right)\mathrm{and}\mathrm{thymidine}\\ \begin{array}{c}\left(T\right)\mathrm{counts}\mathrm{from}\mathrm{all}\mathrm{seqeunce}\\ \mathrm{reads}\mathrm{of}\mathrm{the}\mathrm{CpG}\mathrm{loci}\end{array}\end{array}}& \left(X\right)\end{array}$

[0096] For example, a CpG locus having a methylation ratio of 0.5 is methylated in 50% of the sequencing reads covering the specific CpG locus and unmethylated in 50% of the reads covering the specific CpG. A CpG locus having a methylation ratio of 0.75 is methylated in 75% of the sequencing reads covering the specific CpG locus and unmethylated in 25% of the reads covering the specific CpG. A CpG locus having a methylation ratio of 1.0 is methylated in 100% of the sequencing reads covering the specific CpG locus and unmethylated in 0% of the reads covering the specific CpG.

[0097] The methylation ratio of a specific CpG locus describes the degree of methylation of that specific CpG locus in the population or pool of DNA (for example the degree of methylation of that specific CpG locus in a sample of cfDNA obtained from the plasma of a subject).

[0098] Tools such as BSMAP (PMID: 19635165), Bismark (PMID: 21493656), gemBS (PMID: 30137223), Arioc (PMID: 29554207), BS-Seeker2 (PMID: 24206606), MethylCoder (PMID: 21724594) or BatMeth2 (PMID: 30669962) can be used to determine methylation ratios. These programs can also align the sequencing reads from bisulfite sequencing before determining methylation ratios.

[0099] As used herein the term “reference methylation ratio” is the methylation ratio of a CpG locus in a reference sample or reference methylome, for example the methylation ratio of a CpG locus in one of the following: [0100] a cfDNA sample from a healthy subject, for example a healthy age-matched subject (for example, a sample from a subject before they have developed cancer); [0101] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0102] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0103] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0104] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0105] a cfDNA sample from a different subject having a cancer (for example prostate cancer), preferably wherein the level of cancer fraction in the cfDNA sample from the different subject is known (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200 or 500 samples) each from a different subject having cancer, preferably wherein the level of cancer fraction in each cfDNA sample from the different subjects is known, and more preferably wherein each cfDNA sample has a different level of cancer fraction); [0106] a cfDNA sample from a different subject having cancer (for example prostate cancer), wherein preferably the sample is known to comprise cfDNA derived from a cancer subtype (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having cancer, wherein preferably the each sample is known to comprise cfDNA derived from the cancer subtype, and for example wherein each cfDNA sample has a different level of cfDNA derived from the cancer subtype); [0107] a characterized methylome sequence of a white blood cell; [0108] a characterized methylome sequence of a cancer cell line (for example prostate cancer cell line); [0109] a characterized methylome sequence of a cancerous cell; and/or [0110] a characterized methylome sequence of a non-cancerous cell.

[0111] As regards using a cfDNA sample from a different subject having prostate cancer, wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known, the level of prostate cancer fraction in a cfDNA sample from a different subject can be determined by, for example, using methods that estimate tumour fraction using genomic markers. Due to the low sensitivity of such methods, generally the lowest percentage level of tumour fraction in a cfDNA sample that can be detected are around 5 to 10% tumour fraction.

[0112] As used herein the term “average methylation ratio” is the average of the methylation ratios of all the CpGs within a given genomic region. The average methylation ratio can be calculated by determining the sum of the methylation ratios of all CpGs within a given genomic region and dividing the sum by the number of CpGs within the given genomic region. The average methylation ratio may also be referred to as the mean methylation ratio. If a genomic region has only 1 CpG locus, the average methylation is the same as the methylation ratio for the single CpG locus in the genomic region. Programs such as methylKit R package v1.6.2 (Akalin, A. et al. Genome Biol 13, R87 (2012)) can be used to calculate average methylation ratio.

[0113] The average methylation ratio of a specific genomic region describes the degree of methylation of that specific genomic region in the population or pool of DNA (for example the degree of methylation of that specific genomic region in a sample of cfDNA obtained from the plasma of a subject).

[0114] The term “hypermethylated region” as used herein refers to a genomic region of cfDNA that is indicative of cancer when there is an increase in the average methylation ratio in the region (i.e. hypermethylation) compared to the average methylation ratio of the same genomic region in one or more of the following: [0115] a cfDNA sample from a healthy subject, for example a healthy age-matched subject (for example, a sample from a subject before they have developed cancer); [0116] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0117] a characterized methylome sequence of a white blood cell; [0118] a cfDNA sample from a different subject having prostate cancer, wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known, and wherein the level of prostate cancer fraction in the cfDNA sample from a different subject is lower (for example at least 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40% or 50% lower) compared to the sample from the subject (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known and wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is lower (for example at least 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40% or 50% lower) compared to the sample from the subject).

[0119] The term “hypomethylated region” as used herein refers to a genomic region of cfDNA that is indicative of cancer when there is a decrease in the average methylation ratio in the region (i.e. hypomethylation) compared to the average methylation ratio of the same genomic region in one or more of the following: [0120] a cfDNA sample from a healthy subject, for example a healthy age-matched subject (for example, a sample from a subject before they have developed cancer); [0121] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0122] a characterized methylome sequence of a white blood cell; [0123] a cfDNA sample from a different subject having prostate cancer, wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known, and wherein the level of prostate cancer fraction in the cfDNA sample from a different subject is higher (for example at least 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40% or 50% higher) compared to the sample from the subject (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known and wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is higher (for example at least 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40% or 50% higher) compared to the sample from the subject).

[0124] The term “methylation score” as used herein is a value that is indicative of the methylation state of a sub-population or fraction of DNA in a sample. For example a “methylation score” may be indicative of the methylation state of the genomic regions in a sample that have the average methylation ratio determined. The methylation score may be, for example: [0125] the median or the mean of the average methylation ratios for the genomic regions that have had average methylation ratios determined; [0126] the median or the mean of the average methylation ratios for a first group of genomic regions that have had average methylation ratios determined (resulting in a first methylation score) and/or the median or the mean of the average methylation ratios for a second group of genomic regions that have had average methylation ratios determined (resulting in a second methylation score) (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions, and the second group of genomic regions are all of the hypomethylated genomic regions); or [0127] the methylation ratio score for each genomic region that have the average methylation ratio determined, wherein a methylation ratio score is determined by comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region.

[0128] In certain embodiments, preferably the methylation score is, for example: [0129] the median of the average methylation ratios for the genomic regions that have the average methylation ratio determined; [0130] the median of the average methylation ratios for a first group of genomic regions that have had average methylation ratios determined (resulting in a first methylation score), and/or the median of the average methylation ratios for a second group of genomic regions that have had average methylation ratios determined (resulting in a second methylation score) (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions, and the second group of genomic regions are all of the hypomethylated genomic regions)

[0131] The term “reference methylation score” as used herein is a methylation score for a reference sample or a reference methylome. The reference sample or reference methylome may selected from the group consisting of: [0132] a cfDNA sample from a healthy subject, for example a healthy age-matched subject (for example, a sample from a subject before they have developed cancer); [0133] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0134] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0135] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0136] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0137] a cfDNA sample from a different subject having cancer (for example prostate cancer), preferably wherein the level of cancer fraction in the cfDNA sample from the different subject is known (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having cancer, wherein preferably the level of cancer fraction in each cfDNA sample from the different subjects is known, and more preferably wherein each cfDNA sample has a different level of cancer fraction); [0138] a cfDNA sample from a different subject having cancer (for example prostate cancer), wherein preferably the sample is known to comprise cfDNA derived from a cancer subtype (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having cancer, wherein preferably the each sample is known to comprise cfDNA derived from the cancer subtype, and for example wherein each cfDNA sample has a different level of cfDNA derived from the cancer subtype); [0139] a characterized methylome sequence of a white blood cell; [0140] a characterized methylome sequence of a cancer cell line; [0141] a characterized methylome sequence of a cancerous cell; and [0142] a characterized methylome sequence of a non-cancerous cell.

[0143] The reference methylation score is preferably calculated (for example calculated using the average methylation ratio) for the same genomic regions as the genomic regions for a methylation score to which the reference methylation score is to be compared with.

[0144] For example, if a methylation score is the median of the average methylation ratios for all genomic regions that have had the average methylation ratios determined, then preferably the reference methylation score is the median of the average methylation ratios for the same genomic regions in a reference sample or reference methylome. If a methylation score is the mean of the average methylation ratios for all genomic regions that have had the average methylation ratios determined, then preferably the reference methylation score is the mean of the average methylation ratios for the same genomic regions in a reference sample or reference methylome

[0145] If a methylation score is the median of the average methylation ratios for a first group of genomic regions (resulting in a first methylation score) and/or the median of the average methylation for a second group of genomic regions (resulting in a second methylation score), then preferably the reference methylation score is the median of the average methylation ratios for the same first group of genomic regions (resulting in a first reference methylation score) and/or the median of the average methylation ratios for the same second group of genomic regions (resulting in a second reference methylation score) in a reference sample or reference methylome.

[0146] If a methylation score is the mean of the average methylation ratios for a first group of genomic regions (resulting in a first methylation score) and/or the mean of the average methylation for a second group of genomic regions (resulting in a second methylation score), then preferably the reference methylation score is the mean of the average methylation ratios for the same first group of genomic regions (resulting in a first reference methylation score) and/or the mean of the average methylation ratios for the same second group of genomic regions (resulting in a second reference methylation score) in a reference sample or reference methylome.

[0147] If a methylation score is the methylation ratio score for each genomic region that have the average methylation ratio determined, wherein a methylation ratio score is determined by comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region, the reference methylation score is preferably the reference methylation ratio score for each of the same genomic regions in a reference sample.

[0148] As used herein an “abnormal level of PSA” is a level of PSA in the blood indicative of a risk of a patient having prostate cancer. For example an abnormal level of PSA in the blood may be a level of at least 4.0 ng/mL. An “abnormal level of PSA” may additionally be an increase in the level of PSA in the blood compared to the level at initial diagnosis or the level at the previous time PSA was tested in the subject (for example an increase of 0.1 ng/mL or more, 0.2 ng/mL or more, 0.5 ng/mL or more, 1.0 ng/mL or more compared to the level at initial diagnosis or the level at the previous time PSA was tested in the subject).

[0149] The term “oligonucleotide(s)” are nucleic acids that are usually between 5 and 100 contiguous bases, for example between 5-10, 5-20, 10-20, 10-50, 15-50, 15-100, 20-50, or 20-100 contiguous bases. An oligonucleotide may be capable of hybridising to a target of interest, e.g., a sequence that is at least 10 nucleotides in length. An oligonucleotide for hybridising to a target may comprise at least 10, at least 15 nucleotides, at least 20 nucleotides, at least 30 nucleotides, at least 40 nucleotides, at least 50 nucleotides or at least 60 nucleotides. An oligonucleotide can be used as a primer, a probe, included in a microarray, or used in polynucleotide-based identification methods. An oligonucleotide may be capable of hybridising to a DNA genomic region of the invention, for example a DNA genomic region as defined in Tables 1 to 4, or DNA genomic region comprising a DNA genomic region as defined in Tables 1 to 4, or a 2 to 99 bp DNA genomic region within a DNA genomic region defined in Tables 1 to 4 and comprising at least one CpG locus.

[0150] The term “comprising” as used in this specification and claims means “consisting at least in part of” or “consisting of”, that is to say when interpreting statements in this specification and claims which include the term, the features, prefaced by that term in each statement, all need to be present but other features can also be present. Related terms such as “comprise” and “comprised” are to be interpreted in a similar manner.

[0151] As used herein a “subtype of a cancer” (for example a “subtype of prostate cancer”) is a subset of a type of cancer based on characteristics of the cancer cells, and in particular molecular and genetic characteristics of the cells. Different cancer subtypes can have different disease progression and can respond or not respond to different treatments. The subtype of a cancer is, for example, used to assist in planning treatment and determine prognosis of the patient having that cancer subtype.

[0152] As used herein a “solid cancer cfDNA methylome signature” is a set of CpG loci and/or genomic regions that have a certain state of methylation in cfDNA derived from solid cancer cells. The pattern or fingerprint of methylation at the set of CpG loci and/or genomic regions is indicative of the solid cancer, and can provide information relating to the solid cancer, for example the level of solid cancer fraction in the cfDNA sample, a subtype of cancer (for example a genomic subtype), the aggression of the cancer, the prognosis of the cancer, and/or the tumour response to a treatment. A CpG locus or genomic region of a solid cancer cfDNA methylome signature may be tissue specific (for example, a certain state of methylation present in a particular tissue type, i.e. the tissue from which the cancer is derived) and/or cancer specific (for example, a certain state of methylation present in a particular cancer type). A CpG locus or genomic region of a solid cancer cfDNA methylome signature may have increased methylation compared to, for example, the methylation of the same locus or genomic region in a white blood cell and/or non-tumour cell and/or a different tissue to the cancer tissue, and especially compared to the methylation of the same locus or genomic region in a white blood cell. A CpG locus or genomic region of a solid cancer cfDNA methylome signature may have decreased methylation compared to, for example, the methylation of the same locus or genomic region in a white blood cell and/or non-tumour cell and/or a different tissue to the cancer tissue, and especially compared to the methylation of the same locus or genomic region in a white blood cell.

DETAILED DESCRIPTION OF THE INVENTION

[0153] Tumour DNA circulates in the plasma of cancer patients admixed with DNA from non-cancerous cells. The genomic landscape of plasma DNA has been characterized in prostate cancer, for example, metastatic castration-resistant prostate cancer (mCRPC) but the plasma methylome has not been extensively explored. The identification of circulating methylation biomarkers can be challenging due to the heterogeneities of methylation. The traditional way to identify methylation markers started with the comparison between cancer tissue and normal tissue methylation patterns, and cancer-specific methylation loci are chosen and later validated in plasma samples. The present inventors have used an innovative approach and workflow to characterize the plasma methylome in mCRPC and identify a unique set of methylation markers due to the innovative experimental design which uses an unbiased approach to investigate the methylation profile of tumour derived cfDNA. The inventors' approach starts from profiling plasma pan-methylome. They then applied unbiased dimensional reduction algorithms, such as principal component analysis (PCA), and selected the regions most highly correlated with genomically-determined tumour fraction or the subtype of interest. The methylation markers identified by this approach markers can be used as cancer-specific methylation signatures in methods of the invention for high sensitivity and accurate tracking of tumour DNA in subjects with, for example, suspected or confirmed untreated or treated prostate cancer and/or for subtyping prostate cancer patients.

[0154] Furthermore, due to the large number of regions that the inventors have found to highly correlate with prostate and prostate cancer specific DNA methylation patterns and that show the greatest variance when compared to non-cancer plasma DNA in age-matched men, the inventors have been able to develop methods that are applicable to, for example, low-pass whole genome bisulfite sequencing data, and thus will be cost-effective and clinically scalable methods for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer.

[0155] Additionally, due to the methylation markers of the signatures of the present invention being based on variance compared to non-cancer plasma DNA in age-matched men, the signatures can be used in the methods described here to provide increased sensitivity and specificity for determining the level of prostate cancer fraction in a cfDNA sample, and in particular to detect significantly lower levels than is possible using genomic screening of cfDNA. Also, as methylation markers are not affected by clonal hematopoiesis in older populations (i.e. the formation of a genetically distinct subpopulation of blood cells), which can introduce false positives in genomic alternation-based tests, the methods of the present invention are applicable to subjects of all ages. Furthermore, as the methods of the invention determine methylation levels at multiple different methylation markers of the signatures of the present invention, the methods are not biased by inter-patient differences and genomic changes that could occur in normal cells and that could introduce a false positive result in the case of genomic testing.

[0156] Surprisingly, and due to the innovative workflow of the present invention, the methylation signatures of the present invention include methylation markers that are specific to either normal prostate tissue or prostate cancer tissue. The approach can be adapted and applied to other tumour types to identify circulating tumour-specific methylation signatures that can be used to accurately detect a tumour at earlier stages and quantitate tumour fraction. Also surprisingly, the signature found by the inventors did not include genes whose methylation status has been previously reported as diagnostic of prostate cancer such as, GSTP1, APC, RASSF1 and HOXD3 (Massie, C. E, et al, J Steroid Biochem Mol Biol 166, 1-15 (2017)). Although not wishing to be bound by theory, the present inventors postulate that this finding could be explained by highly variable methylation levels at the genomic regions of the signature in non-cancer plasma DNA compared to cancer plasma DNA. The inventors therefore understand that, in view of the signatures being found by the innovative workflow of the present invention, only the most stably methylated regions in non-cancer plasma DNA are identified as discriminators between non-cancer plasma DNA and cancer plasma DNA and are included in the signatures.

[0157] The present invention finds particular utility in risk stratification of men diagnosed with localised prostate cancer. Men with prostate cancer DNA detected in plasma using methods of the present invention can be staged, classified, and/or offered additional treatment with the aim of maximising cure whilst minimising over-treatment of men who do not require it. Furthermore, the methods of the invention can be used to identify patients with poorer prognosis so that a more intensive primary treatment can be selected, i.e. patients with a high tumour fraction level in the plasma, or who have an aggressive subtype of cancer. The methods can also be used for monitoring whether a treatment for prostate cancer is working or not, and for selecting further treatment, if necessary. Also, the half-life of Plasma DNA is approximately 1 hour so changes can be seen within days when a cancer is responding/not responding. Thus testing, after start of treatment (for example days or weeks after start of treatment) could be used to identify men for whom treatment is ineffective and to guide a change to a more effective alternative, potentially improving outcomes and minimising unnecessary side-effects.

[0158] Currently PSA testing is used to determine bio-chemical progression, and whole-body MRI scanning/PSMA testing for detecting metastases. PSA testing has come under much scrutiny for its reliability and overdiagnosis. Imaging modalities cannot detect metastatic disease as early as a ctDNA test. Imaging can only detect lesions >0.5-1 cm, i.e. 1 million cells or more. On the other hand, it is possible detect DNA from a few 100 tumour cells in circulation. The methods of the present invention can therefore complement or replace imaging for more accurate detecting, screening, monitoring, staging, classification and prognostication of prostate cancer, and in particular metastatic prostate cancer.

[0159] Furthermore, the methods and approaches employed by the present inventors to find the signatures described herein can be used in methods to find further signatures useful for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of other solid cancers in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA).

[0160] DNA cytosine methylation, also called DNA methylation or CpG methylation, plays an important part in multiple biological processes by interacting with specific methyl-CpG binding proteins or specific methyl-CpG binding domains (MBDs), a key messenger to other transcriptional regulators which result in histone modification, chromatin re-arrangement, and differential gene expressions (Ballestar, E. & Esteller, M. Biochem Cell Biol 83, 374-384 (2005); Nakayama, T. et al. Lab Invest 80, 1789-1796 (2000)). Some DNA methylation is believed to remain constant in tumour clones, and have the unique inheritance, while some methylation consequences may be later events and result in more malignant form of cancer (Beltran, H. et al. Nat Med 22, 298-305 (2016)). Therefore, it has been hypothesized that DNA methylation signatures could be an important indicator for both early carcinogenesis and advanced tumour progression.

Methods of the Invention to Determine the Level of Prostate Cancer Fraction in a cfDNA Sample

[0161] The present invention provides a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: [0162] characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0163] determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0164] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0165] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, [0166] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0167] calculating a methylation score using the average methylation ratio for each of the genomic regions; [0168] analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0169] Tables 1 to 4 are provided below. The genomic locations have been separated into 4 tables based on whether a region including, having, or within the genomic location is a hypermethylated region (i.e. indicative of cancer when there in an increase in methylation level for the region) or a hypomethylated region (i.e. indicative of cancer when there is a decrease in methylation level for the region) when used in the method, and a region including, having, or within the genomic location is indicative of a methylation pattern specific to prostate tissue or is indicative of a methylation pattern specific to prostate cancer. The genomic locations of Tables 1 (and Table 1b) to 4 are locations with reference to hg19.

TABLE-US-00001 TABLE 1 Hypermethylated region, prostate tissue specific genomic locations Chromosome start end gene 1 24649401 24649501 GRHL3 1 36043651 36043751 TFAP2E 1 45274001 45274101 BTBD19 1 45274051 45274151 BTBD19 1 47909951 47910051 n/a 1 64937351 64937451 CACHD1 1 67600301 67600401 C1orf141 1 67600451 67600551 C1orf141 1 68154601 68154701 n/a 1 119526151 119526251 TBX15 1 119526201 119526301 TBX15 1 119526251 119526351 TBX15 1 119526301 119526401 TBX15 1 119526351 119526451 TBX15 1 119526401 119526501 TBX15 1 119531601 119531701 TBX15 1 119531651 119531751 TBX15 1 119531701 119531801 TBX15 1 119532751 119532851 TBX15 1 119532801 119532901 TBX15 1 119532851 119532951 TBX15 1 235147651 235147751 n/a 1 235147701 235147801 n/a 10 22765851 22765951 n/a 10 22765901 22766001 n/a 10 29096801 29096901 LINC01517 10 29096851 29096951 LINC01517 10 31423451 31423551 n/a 10 31423501 31423601 n/a 10 31423551 31423651 n/a 10 101280651 101280751 n/a 10 101280701 101280801 n/a 10 126336751 126336851 FAM53B 10 126336801 126336901 FAM53B 11 2920151 2920251 SLC22A18AS 11 46367051 46367151 DGKZ 11 46367101 46367201 DGKZ 11 47939651 47939751 n/a 11 47939701 47939801 n/a 11 47939751 47939851 n/a 11 63973901 63974001 FERMT3 11 111809551 111809651 DIXDC1 12 19389701 19389801 PLEKHA5 12 19389751 19389851 PLEKHA5 12 45443801 45443901 DBX2 12 54441001 54441101 HOXC4 12 54441051 54441151 HOXC4 12 81471601 81471701 ACSS3 12 81471651 81471751 ACSS3 12 90150551 90150651 n/a 12 90150601 90150701 n/a 12 95941801 95941901 USP44 12 116997001 116997101 MAP1LC3B2 12 116997051 116997151 MAP1LC3B2 12 116997101 116997201 MAP1LC3B2 13 95357651 95357751 LOC101927248 13 95357701 95357801 LOC101927248 13 99959551 99959651 GPR183 13 99959601 99959701 GPR183 14 24808701 24808801 RIPK3 14 37124151 37124251 n/a 14 37124201 37124301 n/a 14 60973301 60973401 n/a 14 60973351 60973451 n/a 14 60976851 60976951 SIX6 14 60976901 60977001 SIX6 14 60977001 60977101 SIX6 14 60977051 60977151 SIX6 14 61109951 61110051 n/a 15 42227251 42227351 EHD4 15 96909701 96909801 n/a 17 46667051 46667151 HOXB-AS3 17 46673851 46673951 HOXB-AS3 17 46673901 46674001 HOXB-AS3 17 59534551 59534651 TBX4 17 59534601 59534701 TBX4 17 59534651 59534751 TBX4 17 59534701 59534801 TBX4 17 75471401 75471501 40057 17 80944051 80944151 B3GNTL1 18 12307251 12307351 TUBB6 19 16394401 16394501 n/a 19 16394451 16394551 n/a 19 18508551 18508651 LRRC25 19 35396351 35396451 n/a 19 41316751 41316851 n/a 19 41316801 41316901 n/a 19 46526251 46526351 PGLYRP1 19 46917001 46917101 CCDC8 19 46917051 46917151 CCDC8 19 53039001 53039101 ZNF808 19 55592451 55592551 EPS8L1 19 58728401 58728501 n/a 19 58728451 58728551 n/a 2 8379951 8380051 LINC00299 2 8380001 8380101 LINC00299 2 26521751 26521851 n/a 2 26521801 26521901 n/a 2 45465301 45465401 LINC01121 2 46613551 46613651 EPAS1 2 54900801 54900901 n/a 2 54900901 54901001 n/a 2 54900951 54901051 n/a 2 54901001 54901101 n/a 2 63282251 63282351 OTX1 2 63282651 63282751 OTX1 2 63283851 63283951 OTX1 2 63283901 63284001 OTX1 2 71116551 71116651 LINC01143 2 71116601 71116701 LINC01143 2 71116651 71116751 LINC01143 2 71116701 71116801 LINC01143 2 71126251 71126351 n/a 2 71131551 71131651 VAX2 2 71131601 71131701 VAX2 2 71134851 71134951 VAX2 2 172945251 172945351 METAP1D 2 176964051 176964151 HOXD12 2 177012551 177012651 n/a 2 177012601 177012701 n/a 2 177012651 177012751 n/a 2 177012701 177012801 n/a 2 201450501 201450601 AOX1 2 201450551 201450651 AOX1 2 201450601 201450701 AOX1 2 201450651 201450751 AOX1 2 206551401 206551501 NRP2 2 206551451 206551551 NRP2 2 228324851 228324951 n/a 2 228324901 228325001 n/a 2 228324951 228325051 n/a 2 228325001 228325101 n/a 2 238777551 238777651 RAMP1 2 242908101 242908201 LINC01237 21 37802451 37802551 n/a 21 37802501 37802601 n/a 21 37802551 37802651 n/a 3 33701201 33701301 CLASP2 3 46448851 46448951 CCRL2 3 46448901 46449001 CCRL2 3 127453801 127453901 MGLL 3 127453851 127453951 MGLL 3 167742601 167742701 GOLIM4 3 170746251 170746351 n/a 4 20256801 20256901 SLIT2 4 20256851 20256951 SLIT2 4 54959951 54960051 n/a 4 54960001 54960101 n/a 4 54975201 54975301 n/a 4 54975251 54975351 n/a 4 74809851 74809951 n/a 4 75230551 75230651 EREG 4 75230601 75230701 EREG 4 81107201 81107301 PRDM8 4 87281351 87281451 MAPK10 4 87281401 87281501 MAPK10 4 108814501 108814601 LOC101929595 4 108814551 108814651 SGMS2 4 188917101 188917201 ZFP42 5 297251 297351 PDCD6 5 1608551 1608651 LOC728613 5 1608601 1608701 LOC728613 5 72676801 72676901 n/a 5 87439351 87439451 n/a 5 87439401 87439501 n/a 5 134735451 134735551 H2AFY 5 134880301 134880401 n/a 5 140800801 140800901 PCDHGA11 5 170735101 170735201 n/a 5 170735251 170735351 TLX3 5 172673051 172673151 n/a 6 6901051 6901151 n/a 6 10887701 10887801 SYCP2L 6 26088151 26088251 HFE 6 27107251 27107351 HIST1H2BK 6 27107301 27107401 HIST1H4I 6 27107651 27107751 HIST1H4I 6 27107701 27107801 HIST1H2BK 6 27107751 27107851 HIST1H4I 6 27858251 27858351 HIST1H3J 6 27858301 27858401 HIST1H3J 6 27858551 27858651 HIST1H3J 6 139795501 139795601 LINC01625 6 147235051 147235151 STXBP5-AS1 7 27289101 27289201 n/a 7 38361201 38361301 n/a 7 45066651 45066751 CCM2 7 45066701 45066801 CCM2 7 73132051 73132151 STX1A 7 73132101 73132201 STX1A 7 116140351 116140451 CAV2 7 129423101 129423201 n/a 7 129425301 129425401 n/a 7 129425351 129425451 n/a 7 149112151 149112251 n/a 8 55066251 55066351 n/a 8 55066301 55066401 n/a 9 971451 971551 n/a 9 22005201 22005301 CDKN2B 9 22005251 22005351 CDKN2B-AS1 9 22005301 22005401 CDKN2B-AS1 9 22005501 22005601 CDKN2B-AS1 9 22005551 22005651 CDKN2B-AS1 9 22005601 22005701 CDKN2B 9 112810301 112810401 PALM2-AKAP2 9 126775151 126775251 LHX2 9 135462201 135462301 BARHL1 9 139129901 139130001 QSOX2

TABLE-US-00002 TABLE 2 Hypermethylated region, prostate cancer specific genomic locations Chromosome start end gene 1 12404851 12404951 VPS13D 1 19278651 19278751 IFFO2 1 27944951 27945051 FGR 1 27945001 27945101 FGR 1 28218201 28218301 RPA2 1 54562051 54562151 TCEANC2 1 54562101 54562201 TCEANC2 1 54562151 54562251 TCEANC2 1 65399401 65399501 JAK1 1 65399451 65399551 JAK1 1 66839051 66839151 PDE4B 1 66839101 66839201 PDE4B 1 68154551 68154651 n/a 1 117046951 117047051 n/a 1 117047001 117047101 n/a 1 117058401 117058501 CD58 1 117058451 117058551 CD58 1 150971801 150971901 FAM63A 1 154376251 154376351 n/a 1 154376301 154376401 n/a 1 155506001 155506101 ASH1L 1 156462151 156462251 MEF2D 1 156509751 156509851 IQGAP3 1 156509801 156509901 IQGAP3 1 181031851 181031951 n/a 1 202130701 202130801 PTPN7 1 207103601 207103701 PIGR 1 207103651 207103751 PIGR 1 217313801 217313901 n/a 10 497351 497451 DIP2C 10 22766101 22766201 n/a 10 22936901 22937001 PIP4K2A 10 22936951 22937051 PIP4K2A 10 88632601 88632701 BMPR1A 10 88632651 88632751 BMPR1A 10 94450801 94450901 HHEX 10 94450851 94450951 HHEX 10 94450901 94451001 HHEX 10 94450951 94451051 HHEX 10 94451001 94451101 HHEX 11 31817851 31817951 PAX6 11 62455001 62455101 LRRN4CL 11 70211351 70211451 PPFIA1 11 70211401 70211501 PPFIA1 11 70211451 70211551 PPFIA1 11 70211501 70211601 PPFIA1 11 70248651 70248751 CTTN 11 70248701 70248801 CTTN 12 1642601 1642701 n/a 12 6665301 6665401 IFFO1 12 6665351 6665451 IFFO1 12 6665401 6665501 IFFO1 12 7060151 7060251 PTPN6 12 7060201 7060301 PTPN6 12 7062051 7062151 PTPN6 12 7062101 7062201 PTPN6 12 7062151 7062251 PTPN6 12 47610151 47610251 PCED1B-AS1 12 47610201 47610301 PCED1B 12 109899001 109899101 KCTD10 12 109899051 109899151 KCTD10 12 111536901 111537001 CUX2 12 111536951 111537051 CUX2 12 115135751 115135851 n/a 12 123707851 123707951 MPHOSPH9 13 113437951 113438051 ATP11A 13 113438001 113438101 ATP11A 13 113438051 113438151 ATP11A 14 37124901 37125001 n/a 14 37124951 37125051 n/a 14 37125001 37125101 n/a 14 37125801 37125901 PAX9 14 37125851 37125951 PAX9 14 37125901 37126001 PAX9 14 37126051 37126151 PAX9 14 38725601 38725701 CLEC14A 14 38725651 38725751 CLEC14A 14 95237351 95237451 GSC 14 95237401 95237501 GSC 15 86098551 86098651 AKAP13 15 86098601 86098701 AKAP13 15 96887101 96887201 n/a 15 101777751 101777851 CHSY1 15 101777801 101777901 CHSY1 15 101991451 101991551 PCSK6 16 2737251 2737351 KCTD5 16 2737301 2737401 KCTD5 16 29675101 29675201 SPN 16 88038901 88039001 BANP 16 88866601 88866701 n/a 17 41799001 41799101 n/a 17 41799051 41799151 n/a 17 43242651 43242751 HEXIM2 17 55533301 55533401 MSI2 17 55533351 55533451 MSI2 17 55562901 55563001 MSI2 17 55562951 55563051 MSI2 17 56407101 56407201 BZRAP1-AS1 17 59532801 59532901 TBX4 17 59532851 59532951 TBX4 17 59536601 59536701 TBX4 17 70715551 70715651 SLC39A11 17 72776151 72776251 TMEM104 17 72776201 72776301 TMEM104 17 78724151 78724251 RPTOR 17 79422501 79422601 BAHCC1 17 79422551 79422651 BAHCC1 17 79422601 79422701 BAHCC1 17 80740751 80740851 TBCD 17 81039651 81039751 METRNL 17 81039701 81039801 METRNL 19 2776001 2776101 SGTA 19 31843301 31843401 n/a 19 33162801 33162901 ANKRD27 19 33162851 33162951 ANKRD27 19 33162901 33163001 ANKRD27 2 3246151 3246251 TSSC1 2 3246201 3246301 TSSC1 2 10687901 10688001 n/a 2 10687951 10688051 n/a 2 10688251 10688351 n/a 2 10688601 10688701 n/a 2 10688651 10688751 n/a 2 11674501 11674601 GREB1 2 11674551 11674651 GREB1 2 27298301 27298401 n/a 2 30489401 30489501 n/a 2 36776251 36776351 CRIM1 2 36776301 36776401 CRIM1 2 55339151 55339251 n/a 2 63279651 63279751 OTX1 2 71132301 71132401 VAX2 2 106415051 106415151 NCK2 2 106415101 106415201 NCK2 2 111875851 111875951 n/a 2 111875901 111876001 n/a 2 171569251 171569351 LINC01124 2 172945201 172945301 METAP1D 2 172974151 172974251 DLX2-AS1 2 172974201 172974301 DLX2-AS1 2 198063601 198063701 ANKRD44 2 198063651 198063751 ANKRD44 2 202126301 202126401 CASP8 2 202126351 202126451 CASP8 2 204571201 204571301 CD28 2 204571301 204571401 CD28 2 206004801 206004901 PARD3B 2 206004851 206004951 PARD3B 2 232186801 232186901 ARMC9 2 232186851 232186951 ARMC9 2 232186901 232187001 ARMC9 2 232186951 232187051 ARMC9 2 236774051 236774151 AGAP1 2 237623801 237623901 n/a 2 241504751 241504851 n/a 2 242048301 242048401 PASK 2 242048351 242048451 PASK 2 242785201 242785301 n/a 2 242908201 242908301 LINC01237 20 31123201 31123301 NOL4L 20 31123251 31123351 NOL4L 20 39127001 39127101 n/a 22 23923201 23923301 IGLL1 22 45575251 45575351 NUP50 22 50618551 50618651 PANX2 22 50618601 50618701 PANX2 3 32993101 32993201 CCR4 3 32993151 32993251 CCR4 3 46448551 46448651 CCRL2 3 46448751 46448851 CCRL2 3 46448801 46448901 LOC102724297 3 53700101 53700201 CACNA1D 3 72227101 72227201 n/a 3 72227251 72227351 n/a 3 73620851 73620951 PDZRN3 3 121796551 121796651 CD86 3 123063401 123063501 ADCY5 3 160475201 160475301 PPM1L 3 160475251 160475351 PPM1L 3 167659101 167659201 n/a 3 167659151 167659251 n/a 3 177397701 177397801 LINC00578 3 177397751 177397851 LINC00578 3 184504551 184504651 n/a 3 190363701 190363801 IL1RAP 3 194868651 194868751 XXYLT1-AS2 3 194868701 194868801 XXYLT1-AS2 3 194868751 194868851 XXYLT1-AS2 4 1221751 1221851 CTBP1 4 1221801 1221901 CTBP1 4 1742401 1742501 TACC3 4 13544651 13544751 NKX3-2 4 13544701 13544801 NKX3-2 4 53862451 53862551 SCFD2 4 53862501 53862601 SCFD2 4 57522951 57523051 HOPX 4 74713351 74713451 n/a 4 77226301 77226401 STBD1 4 77226351 77226451 STBD1 4 101438751 101438851 EMCN 4 101438801 101438901 EMCN 4 183795701 183795801 n/a 4 183795751 183795851 n/a 4 184320501 184320601 n/a 4 186559651 186559751 SORBS2 5 1107151 1107251 SLC12A7 5 10445501 10445601 ROPN1L 5 10445551 10445651 ROPN1L 5 14331751 14331851 TRIO 5 14331801 14331901 TRIO 5 14676401 14676501 OTULIN 5 31470851 31470951 DROSHA 5 32734901 32735001 NPR3 5 32734951 32735051 NPR3 5 37834151 37834251 GDNF 5 80050751 80050851 MSH3 5 80050801 80050901 MSH3 5 92931551 92931651 MIR548AO 5 92931901 92932001 MIR548AO 5 134826301 134826401 n/a 5 134826351 134826451 n/a 5 135266751 135266851 FBXL21 5 135266801 135266901 FBXL21 5 138714601 138714701 SLC23A1 5 138714651 138714751 SLC23A1 5 150593601 150593701 CCDC69 5 150593651 150593751 CCDC69 5 176758401 176758501 n/a 5 176758451 176758551 n/a 5 179344551 179344651 n/a 5 179344601 179344701 n/a 6 2733751 2733851 MYLK4 6 10393751 10393851 n/a 6 10425551 10425651 n/a 6 34252751 34252851 n/a 6 34252801 34252901 n/a 6 36209701 36209801 n/a 6 41394751 41394851 n/a 6 41394801 41394901 n/a 6 41394851 41394951 n/a 6 45296051 45296151 RUNX2 6 45296101 45296201 SUPT3H 6 135516851 135516951 MYB 6 135516901 135517001 MYB 6 157184151 157184251 ARID1B 6 168107101 168107201 n/a 6 170531301 170531401 n/a 6 170531351 170531451 n/a 7 5518851 5518951 FBXL18 7 5518901 5519001 FBXL18 7 5518951 5519051 FBXL18 7 6475951 6476051 DAGLB 7 6476001 6476101 DAGLB 7 6476051 6476151 DAGLB 7 6476101 6476201 DAGLB 7 27281401 27281501 EVX1 7 27289151 27289251 n/a 7 75957101 75957201 YWHAG 7 96631701 96631801 DLX6-AS1 7 96631751 96631851 DLX6-AS1 7 96650001 96650101 DLX5 7 96650051 96650151 DLX5 7 105662751 105662851 CDHR3 7 128579851 128579951 IRF5 7 129411001 129411101 MIR182 7 129411051 129411151 MIR182 7 129411101 129411201 MIR182 7 129411151 129411251 MIR182 8 76318451 76318551 n/a 8 99950851 99950951 STK3 8 99950901 99951001 STK3 8 102149801 102149901 n/a 8 117487001 117487101 n/a 8 134072501 134072601 SLA 8 140945801 140945901 TRAPPC9 8 141584651 141584751 AGO2 8 141584701 141584801 AGO2 8 144408451 144408551 TOP1MT 8 144408501 144408601 TOP1MT 8 144408551 144408651 TOP1MT 9 91006601 91006701 SPIN1 9 91006651 91006751 SPIN1 9 91006701 91006801 SPIN1 9 91006751 91006851 SPIN1 9 96080251 96080351 WNK2 9 96080301 96080401 WNK2 9 96080351 96080451 WNK2 9 98790151 98790251 n/a 9 101876601 101876701 TGFBR1 9 101876651 101876751 TGFBR1 9 110399201 110399301 n/a 9 110399251 110399351 n/a 9 124045101 124045201 GSN 9 125796751 125796851 RABGAP1 9 125796801 125796901 GPR21 9 125797101 125797201 GPR21 9 125797151 125797251 RABGAP1 9 132650501 132650601 FNBP1 9 132650551 132650651 FNBP1 9 132650601 132650701 FNBP1 9 132650651 132650751 FNBP1 9 134151501 134151601 FAM78A 9 140586151 140586251 EHMT1 9 140586201 140586301 EHMT1

TABLE-US-00003 TABLE 3 Hypomethylated region, prostate tissue specific genomic locations Chromosome start end gene 1 2839151 2839251 n/a 1 2876451 2876551 n/a 1 2876501 2876601 n/a 1 2876551 2876651 n/a 1 43637151 43637251 EBNA1BP2 1 95172951 95173051 LINC01057 1 95173001 95173101 LINC01057 1 95173051 95173151 LINC01057 1 110676801 110676901 n/a 1 155904851 155904951 KIAA0907 1 158465951 158466051 n/a 1 160079651 160079751 n/a 1 162527401 162527501 n/a 1 162527451 162527551 n/a 1 169696601 169696701 SELE 1 175490501 175490601 TNR 1 203829801 203829901 SNRPE 1 204165251 204165351 KISS1 1 204349851 204349951 n/a 1 248153651 248153751 OR2L1P 10 6779901 6780001 n/a 10 6779951 6780051 n/a 10 126713301 126713401 CTBP2 10 126713351 126713451 CTBP2 11 19681701 19681801 n/aV2 11 19681751 19681851 n/aV2 11 27536001 27536101 BDNF-AS 11 27536051 27536151 MIR8087 11 57519401 57519501 BTBD18 11 60680451 60680551 TMEM109 11 67615951 67616051 n/a 11 76371951 76372051 LRRC32 11 78900901 78901001 TENM4 11 78900951 78901051 TENM4 11 88019001 88019101 n/a 11 88019051 88019151 n/a 11 128737201 128737301 KCNJ1 11 132912251 132912351 OPCML 11 133445651 133445751 n/a 12 1702101 1702201 FBXL14 12 4029951 4030051 n/a 12 4030001 4030101 n/a 12 4030051 4030151 n/a 12 5156351 5156451 n/a 12 8438301 8438401 n/a 12 8438351 8438451 n/a 12 8438401 8438501 n/a 12 130711351 130711451 n/a 12 131941401 131941501 n/a 12 132848201 132848301 GALNT9 12 132848251 132848351 GALNT9 12 132848301 132848401 GALNT9 13 112906801 112906901 n/a 14 52219001 52219101 n/a 14 52219051 52219151 n/a 14 52219101 52219201 n/a 14 59296551 59296651 LINC01500 14 59296601 59296701 LINC01500 14 59296651 59296751 LINC01500 14 93412701 93412801 ITPK1 14 97497051 97497151 n/a 14 100046451 100046551 CCDC85C 14 104742051 104742151 n/a 14 104742101 104742201 n/a 14 104742151 104742251 n/a 14 104889001 104889101 n/a 14 104889051 104889151 n/a 15 22799001 22799101 n/a 15 22799051 22799151 n/a 15 28051101 28051201 OCA2 16 23988951 23989051 PRKCB 16 86457501 86457601 n/a 16 88218201 88218301 n/a 17 39472101 39472201 KRTAP17-1 17 66951501 66951601 ABCA8 17 79694451 79694551 n/a 17 79694501 79694601 n/a 19 15901151 15901251 n/a 19 16178551 16178651 TPM4 19 54177201 54177301 MIR498 19 54177251 54177351 MIR498 19 54778401 54778501 LILRB2 19 55104551 55104651 LILRA1 19 55104601 55104701 LILRA1 2 879401 879501 n/a 2 879451 879551 n/a 2 879501 879601 n/a 2 2581201 2581301 n/a 2 2581251 2581351 n/a 2 2581301 2581401 n/a 2 59477151 59477251 LOC101927285 2 74153201 74153301 DGUOK 2 100426901 100427001 AFF3 2 100426951 100427051 AFF3 2 107456801 107456901 ST6GAL2 2 147788651 147788751 n/a 2 208795601 208795701 PLEKHM3 2 208795651 208795751 PLEKHM3 2 208795701 208795801 PLEKHM3 2 232455551 232455651 n/a 2 232455601 232455701 n/a 20 1975251 1975351 PDYN 20 1975301 1975401 PDYN 20 1975351 1975451 PDYN 20 1975401 1975501 PDYN 20 19866651 19866751 RIN2 20 19866701 19866801 RIN2 20 62111301 62111401 n/a 21 43735451 43735551 TFF3 21 43735501 43735601 TFF3 21 43735551 43735651 TFF3 21 44375551 44375651 n/a 22 24979501 24979601 GGT1 22 24979551 24979651 GGT1 22 24979601 24979701 GGT1 22 49020401 49020501 FAM19A5 22 49800101 49800201 n/a 22 49800151 49800251 n/a 22 50481601 50481701 n/a 3 29494901 29495001 RBMS3 3 29494951 29495051 RBMS3 3 33757901 33758001 CLASP2 3 36360601 36360701 n/a 3 36360651 36360751 n/a 4 1047001 1047101 n/a 4 1047051 1047151 n/a 4 3895101 3895201 n/a 4 3895151 3895251 n/a 4 5368251 5368351 STK32B 4 5368301 5368401 STK32B 4 5526701 5526801 LINC01587 4 9104401 9104501 n/a 4 9104451 9104551 n/a 4 9104551 9104651 n/a 4 16708251 16708351 LDB2 4 16708301 16708401 LDB2 4 79971101 79971201 LINC01088 4 79971151 79971251 LINC01088 4 79971201 79971301 LINC01088 4 100576551 100576651 n/a 4 100576601 100576701 n/a 4 120502101 120502201 PDE5A 4 190283101 190283201 n/a 5 759101 759201 n/a 5 3188351 3188451 n/a 5 19531501 19531601 CDH18 5 19531551 19531651 CDH18 5 171808201 171808301 SH3PXD2B 5 171808251 171808351 SH3PXD2B 5 178594601 178594701 ADAMTS2 6 87830101 87830201 n/a 6 87830151 87830251 n/a 6 133689901 133690001 EYA4 6 152804701 152804801 SYNE1 6 152804751 152804851 SYNE1 6 159872201 159872301 n/a 6 159872251 159872351 n/a 7 39056301 39056401 POU6F2 7 39056351 39056451 POU6F2 7 39056401 39056501 POU6F2 7 158059651 158059751 PTPRN2 7 158059701 158059801 PTPRN2 7 158059901 158060001 PTPRN2 8 49984651 49984751 C8orf22 8 49984701 49984801 C8orf22 8 49984751 49984851 C8orf22 8 52754401 52754501 PCMTD1 8 105988201 105988301 n/a 8 119073651 119073751 EXT1 8 119073701 119073801 EXT1 8 120779851 120779951 TAF2 8 130365251 130365351 CCDC26 8 130365301 130365401 CCDC26 8 133573401 133573501 HPYR1 8 139124351 139124451 n/a 8 139124401 139124501 n/a 8 139124451 139124551 n/a 8 139784451 139784551 COL22A1 8 142289701 142289801 n/a 8 142289751 142289851 n/a 8 142289801 142289901 n/a 8 142289851 142289951 n/a 9 5756751 5756851 RIC1 9 38437251 38437351 n/a 9 38437301 38437401 n/a 9 92291201 92291301 UNQ6494 9 92291251 92291351 UNQ6494 9 128307501 128307601 MAPKAP1 9 128307551 128307651 MAPKAP1 9 138192201 138192301 n/a 9 138192251 138192351 n/a X 47662501 47662601 n/a X 47662551 47662651 n/a X 52683851 52683951 SSX7

TABLE-US-00004 TABLE 4 Hypomethylated region, prostate cancer specific genomic locations Chromosome start end gene 1 2013951 2014051 PRKCZ 1 4079101 4079201 n/a 1 143907551 143907651 FAM72C 1 148903551 148903651 NBPF25P 1 152648651 152648751 LCE2C 1 152648701 152648801 LCE2C 1 153174651 153174751 n/a 1 153174701 153174801 n/a 1 153174751 153174851 n/a 1 153175201 153175301 LELP1 1 153175251 153175351 LELP1 1 153283751 153283851 PGLYRP3 1 153283801 153283901 PGLYRP3 1 153352051 153352151 n/a 1 153352101 153352201 n/a 1 153353201 153353301 n/a 1 153389951 153390051 S100A7A 1 158465801 158465901 n/a 1 158465851 158465951 n/a 1 159236001 159236101 n/a 1 159236051 159236151 n/a 1 175490551 175490651 TNR 1 175490601 175490701 TNR 1 182021701 182021801 n/a 1 182021751 182021851 n/a 1 209105801 209105901 n/a 1 248308901 248309001 OR2M5 1 248366251 248366351 OR2M3 10 2699351 2699451 n/a 10 6664951 6665051 LOC101928150 10 6665001 6665101 LOC101928150 10 6807101 6807201 n/a 10 7567801 7567901 n/a 10 7567851 7567951 n/a 10 26226851 26226951 MYO3A 10 26226901 26227001 MYO3A 10 26226951 26227051 MYO3A 11 5957801 5957901 n/a 11 6865801 6865901 n/a 11 7961051 7961151 OR10A3 11 7961101 7961201 OR10A3 11 22219001 22219101 ANO5 11 50220151 50220251 n/a 11 55579301 55579401 OR5L1 11 59949051 59949151 MS4A6A 11 121762851 121762951 n/a 11 121762901 121763001 n/a 11 121986951 121987051 MIR100HG 11 122100851 122100951 n/a 11 123900601 123900701 OR10G8 12 3053501 3053601 n/a 12 4361001 4361101 CCND2-AS1 12 4361051 4361151 CCND2-AS1 12 4361101 4361201 CCND2-AS1 12 124397801 124397901 DNAH10 12 124397851 124397951 DNAH10 12 127348201 127348301 n/a 12 127348251 127348351 n/a 12 127348301 127348401 n/a 12 127944451 127944551 n/a 12 127980601 127980701 n/a 12 127980651 127980751 n/a 12 128869901 128870001 TMEM132C 12 129595351 129595451 TMEM132D 12 130411151 130411251 n/a 12 130411201 130411301 n/a 12 130411251 130411351 n/a 12 130494601 130494701 n/a 12 130591301 130591401 n/a 12 130683451 130683551 n/a 12 130750301 130750401 n/a 12 131402501 131402601 n/a 12 131402551 131402651 n/a 12 131418151 131418251 n/a 12 131512551 131512651 ADGRD1 12 131512601 131512701 ADGRD1 12 131769201 131769301 n/a 12 131769251 131769351 n/a 12 131862601 131862701 n/a 12 131941451 131941551 n/a 12 132102101 132102201 n/a 12 132142001 132142101 n/a 12 132142051 132142151 n/a 12 132663851 132663951 n/a 12 132663901 132664001 n/a 14 22315001 22315101 n/a 14 47669951 47670051 MDGA2 14 47670001 47670101 MDGA2 14 47670051 47670151 MDGA2 14 47670101 47670201 MDGA2 14 47670151 47670251 MDGA2 14 47670201 47670301 MDGA2 14 47670251 47670351 MDGA2 14 97497101 97497201 n/a 14 97853651 97853751 n/a 14 97853701 97853801 n/a 14 97924251 97924351 LOC101929241 14 97924301 97924401 LOC101929241 14 97924401 97924501 LOC101929241 14 97924451 97924551 LOC101929241 14 97924501 97924601 LOC101929241 14 98101651 98101751 LOC100129345 14 98101701 98101801 LOC100129345 14 99181901 99182001 C14orf177 14 101495751 101495851 MIR494 14 101495801 101495901 MIR494 15 95287801 95287901 n/a 15 95287851 95287951 n/a 15 95287901 95288001 n/a 15 98646401 98646501 n/a 16 8337501 8337601 n/a 16 8337551 8337651 n/a 16 9855201 9855301 GRIN2A 16 9855251 9855351 GRIN2A 16 9857601 9857701 GRIN2A 16 10206551 10206651 GRIN2A 16 10271751 10271851 GRIN2A 16 10272701 10272801 GRIN2A 16 10272751 10272851 GRIN2A 16 23938951 23939051 PRKCB 16 23939001 23939101 PRKCB 16 23939051 23939151 PRKCB 16 24151201 24151301 PRKCB 16 24151251 24151351 PRKCB 16 24266251 24266351 CACNG3 16 24266301 24266401 CACNG3 16 29322201 29322301 SNX29P2 16 29322251 29322351 SNX29P2 16 32488401 32488501 n/a 16 46391151 46391251 n/a 16 65102901 65103001 CDH11 16 86327651 86327751 n/a 16 86421751 86421851 n/a 16 86666101 86666201 n/a 16 87645651 87645751 JPH3 17 3030101 3030201 OR1G1 17 3030151 3030251 OR1G1 17 21911401 21911501 FLI36000 17 21911451 21911551 FLI36000 17 22016951 22017051 n/a 17 22017001 22017101 n/a 17 22023751 22023851 MTRNR2L1 17 77386201 77386301 RBFOX3 17 77386351 77386451 RBFOX3 17 77386401 77386501 RBFOX3 17 77390001 77390101 RBFOX3 18 5392951 5393051 EPB41L3 18 5393001 5393101 EPB41L3 18 11153851 11153951 n/a 18 11153901 11154001 n/a 19 2715051 2715151 DIRAS1 19 15067451 15067551 SLC1A6 19 29281801 29281901 n/a 19 29281851 29281951 n/a 19 29281901 29282001 n/a 19 43271101 43271201 n/a 19 54568251 54568351 n/a 19 54903901 54904001 n/a 19 54903951 54904051 n/a 19 55036651 55036751 n/a 19 55042251 55042351 n/a 19 55042301 55042401 n/a 19 55692651 55692751 PTPRH 19 55692701 55692801 PTPRH 19 56274351 56274451 RFPL4A 19 56274401 56274501 RFPL4A 19 56346551 56346651 NLRP11 19 56346601 56346701 NLRP11 19 57646101 57646201 ZIM3 2 3633351 3633451 n/a 2 3633401 3633501 n/a 2 44513801 44513901 SLC3A1 2 44513851 44513951 SLC3A1 2 59470151 59470251 LOC101927285 2 59470201 59470301 LOC101927285 2 60880201 60880301 n/a 2 89215101 89215201 n/a 2 89215151 89215251 n/a 2 91910551 91910651 n/a 2 91910601 91910701 n/a 2 91936151 91936251 n/a 2 117006251 117006351 n/a 2 119134001 119134101 n/a 2 119134051 119134151 n/a 2 119471301 119471401 n/a 2 127401201 127401301 n/a 2 127529551 127529651 n/a 2 127529601 127529701 n/a 2 203636951 203637051 n/a 2 228336101 228336201 MIR5703 2 242190801 242190901 HDLBP 20 5282851 5282951 PROKR2 20 5282901 5283001 PROKR2 20 5284401 5284501 PROKR2 20 5450901 5451001 LOC643406 20 44876301 44876401 CDH22 20 59543701 59543801 n/a 20 59543751 59543851 n/a 20 59888501 59888601 CDH4 20 59888551 59888651 CDH4 20 59888601 59888701 CDH4 20 61715801 61715901 LOC63930 20 61754951 61755051 n/a 20 61755001 61755101 n/a 22 17073551 17073651 CCT8L2 22 22902051 22902151 PRAME 22 49635751 49635851 n/a 22 50482001 50482101 n/a 3 13860351 13860451 WNT7A 3 38835151 38835251 SCN10A 3 38835201 38835301 SCN10A 3 38835251 38835351 SCN10A 3 46245401 46245501 CCR1 3 100690851 100690951 ABI3BP 3 100690901 100691001 ABI3BP 3 192769601 192769701 n/a 3 192960401 192960501 MGC2889 3 192960451 192960551 MGC2889 3 192960551 192960651 MGC2889 3 192973501 192973601 HRASLS 3 193096401 193096501 ATP13A5 3 193097751 193097851 n/a 4 9453051 9453151 n/a 4 9523701 9523801 n/a 4 157059251 157059351 n/a 4 157059301 157059401 n/a 4 157059351 157059451 n/a 4 190462551 190462651 n/a 4 190751051 190751151 n/a 5 471201 471301 PP7080 5 2866801 2866901 n/a 5 3002401 3002501 n/a 5 3002451 3002551 n/a 5 3339501 3339601 n/a 5 3454101 3454201 LINC01019 5 3454151 3454251 LINC01019 5 4116251 4116351 n/a 5 5492351 5492451 n/a 5 152949001 152949101 GRIA1 5 153039151 153039251 GRIA1 5 153039201 153039301 GRIA1 6 133932801 133932901 TARID 6 133932851 133932951 TARID 6 153066801 153066901 n/a 6 154330951 154331051 OPRM1 6 155777801 155777901 NOX3 7 57218251 57218351 n/a 7 57247701 57247801 GUSBP10 7 57324901 57325001 n/a 7 57509951 57510051 ZNF716 7 57510151 57510251 ZNF716 7 57510201 57510301 ZNF716 7 57714301 57714401 n/a 7 127256601 127256701 PAX4 7 144967601 144967701 n/a 8 32732251 32732351 n/a 8 56106751 56106851 XKR4 8 56361951 56362051 SBF1P1 8 64314301 64314401 n/a 8 67085751 67085851 TRIM55 8 73849101 73849201 KCNB2 8 73849151 73849251 KCNB2 8 105988151 105988251 n/a 8 107139701 107139801 n/a 8 107139751 107139851 n/a 8 111906551 111906651 n/a 8 114390751 114390851 CSMD3 8 139784401 139784501 COL22A1 8 140624701 140624801 KCNK9 9 27949501 27949601 LINGO2 X 150944951 150945051 n/a X 150945001 150945101 n/a

[0170] In Tables 1 to 4, where the gene indicated is “n/a” this means that the genomic location defined in the table is a non-coding region of DNA or not within the location of a known gene. In certain embodiments, the set of genomic locations listed in Table 1 does not include the genomic locations listed in Table 1b below:

TABLE-US-00005 TABLE 1b Genomic locations that may be excluded from Table 1 Chromosome start end gene 2 26521751 26521851 n/a 2 63282651 63282751 OTX1 2 63283901 63284001 OTX1 2 201450501 201450601 AOX1 2 201450551 201450651 AOX1 2 201450601 201450701 AOX1 2 201450651 201450751 AOX1 3 33701201 33701301 CLASP2 3 170746251 170746351 n/a 4 20256801 20256901 SLIT2 4 54959951 54960051 n/a 4 54960001 54960101 n/a 4 74809851 74809951 n/a 4 87281351 87281451 MAPK10 4 87281401 87281501 MAPK10 5 134880301 134880401 n/a 5 170735101 170735201 n/a 5 172673051 172673151 n/a 6 27858551 27858651 HIST1H3J 6 139795501 139795601 LINC01625 7 129425301 129425401 n/a 9 971451 971551 n/a 12 81471601 81471701 ACSS3 12 81471651 81471751 ACSS3 12 95941801 95941901 USP44 17 80944051 80944151 B3GNTL1 19 46917001 46917101 CCDC8 19 46917051 46917151 CCDC8 19 55592451 55592551 EPS8L1

[0171] The method is for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer. The prostate cancer may be any type of prostate cancer. Suitably, it may be acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer. For example, it may be acinar adenocarcinoma prostate cancer or ductal adenocarcinoma prostate cancer. Alternatively, or additionally, the prostate cancer may be castration sensitive prostate cancer or castration resistant prostate cancer. Alternatively, or additionally, the prostate cancer may be metastatic prostate cancer, or it may be non-metastatic prostate cancer. In certain embodiments, it may be metastatic prostate cancer. In certain embodiments, the prostate cancer may be metastatic castration resistant prostate cancer or non-metastatic castration resistant prostate cancer. For example, it may be metastatic castration resistant prostate cancer.

[0172] The method is especially suitable for the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of metastatic prostate cancer.

[0173] The method is also especially suitable for the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of castration resistant prostate cancer prostate cancer.

[0174] The sample is a sample that comprises cfDNA. The sample may suitably be a blood sample, a plasma sample, or a urine sample. Preferably, the sample is a blood sample or a plasma sample. More preferably, the sample is a plasma sample.

[0175] The method may further comprise isolating the cfDNA from the sample. cfDNA can be isolated from the sample using a variety of techniques known in the art. For example, DNA (e.g., cfDNA) can be isolated by a column-based approach and/or a bead-based approach. In some embodiments, DNA (e.g., cfDNA) is isolated by means of a column-based approach, for example using a commercially available kit such as QIAamp circulating nucleic acid kit (Qiagen qiagen.com/ch/products/discovery-and-translational-research/dna-rna-purification/dna-purification/cell-free-dna/qiaamp-circulating-nucleic-acid-kit/#orderinginformation). In some embodiments, DNA (e.g., cfDNA) is isolated by means of a bead-based approach, for example an automated cf-DNA extraction system using a commercially available kit such as Maxwell RSC ccfDNA Plasma Kit (Promega (https://www.promega.co.uk/resources/protocols/technical-manuals/101/maxwell-rsc-ccfdna-plasma-kit-protocol/)).

[0176] The isolated cfDNA may be amplified before analysis. Thus the method may further comprise amplification of the isolated cfDNA. Amplification techniques are known to those of ordinary skill in the art and include, but are not limited to, cloning, polymerase chain reaction (PCR), polymerase chain reaction of specific alleles (PASA), polymerase chain ligation, nested polymerase chain reaction, and so forth.

[0177] The method comprises characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule. The methylome sequence of a cfDNA molecule may be characterised by using methylation aware sequencing, by genome sequencing followed by methylation profiling, or by targeted approaches that capture specific DNA sequences (for example using DNA probes). Examples of methylation aware sequencing include bisulfite sequencing, bisulfite-free methylation-aware sequencing, methylation arrays (for example methylation microarrays), enzymatic methylation sequencing, methylation-sensitive restriction enzyme digestion, methylation-specific PCR, methylation aware PCR based assays, methylation-dependent DNA precipitation, and methylated DNA binding proteins/peptides. In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using bisulfite sequencing, methylation microarrays, enzymatic methylation sequencing, bisulfite-free methylation-aware sequencing, or methylation aware PCR based assays.

[0178] Examples of targeted approaches that capture specific DNA sequences (for example using DNA probes) include cell-free methylated DNA immunoprecipitation and high-throughput sequencing (cfMeDIP-seq), methylation-dependent DNA precipitation, and methylated DNA binding proteins/peptides.

[0179] Bisulfite sequencing may comprise massive parallel sequencing with bisulfite conversion, for example treating the DNA molecule with sodium bisulfite and performing sequencing of the treated DNA molecule. Methylation assay sequencing may comprise treating the DNA molecule with sodium bisulfite, whole genome amplification, and hybridisation to a methylation-specific probe or a non-methylation probe, for example attached to a bead or chip.

[0180] Enzymatic methylation sequencing may comprise enzymatic treatment of the DNA molecule to convert methylated cytosine sites, followed by sequencing of the treated DNA. For example enzymatic methylation sequencing may comprise enzymatic treatment of the DNA molecule to convert methylated cytosine sites into a form protected from deamination, followed by deamination to convert unprotected cytosine to uracils, and sequencing of the treated DNA. An example of an enzymatic methylation sequencing kit includes NEBNext® Enzymatic Methyl-seq Kit (https://www.neb.com/products/e7120-nebnext-enzymatic-methyl-seq-kit#).

[0181] Examples of methylation aware PCR based assays include digital droplet PCR and qPCR (quantitative PCR).

[0182] An example of bisulfite-free methylation-aware sequencing is Oxford Nanopore seqeuencing (Oxford Nanopore Technologies, https://nanoporetech.com/))

[0183] In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using whole genome bisulfite sequencing, for example low pass whole genome bisulfite sequencing. In another embodiment, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using reduced representation bisulfite treatments. In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using methylation arrays, for example methylation microarrays, such as an Illumina Methylation Assay.

[0184] A variety of genome sequencing procedures are known in the art and may be used to practice the methods disclosed herein. For example, Sanger sequencing, Polony sequencing, 454 pyrosequencing, Combinatorial probe anchor synthesis, SOLiD sequencing, Ion Torrent semiconductor sequencing, DNA nanoball sequencing, Heliscope single molecule sequencing, Single molecule real time (SMRT) sequencing, Nanopore DNA sequencing, Microfluidic Sanger sequencing and Illumina dye sequencing.

[0185] A plurality of cfDNA molecules may be, for example, at least 100, at least 1000, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). Preferably, a plurality of cfDNA molecules may be, for example, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). More preferably, a plurality of cfDNA molecules may be, for example, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9).

[0186] The method may further comprise aligning the methylome sequences with a reference genome for the subject, for example by aligning the methylome sequences with hg38, hg19, hg18, hg17 or hg16. The alignment can, for example, be carried out using a variety of techniques known in the art. For example, a DNA sequence alignment tool, (e.g., BSMAP (PMID: 19635165), Bismark (PMID: 21493656), gemBS (PMID: 30137223), Arioc (PMID: 29554207), BS-Seeker2 (PMID: 24206606), MethylCoder (PMID: 21724594) or BatMeth2 (PMID: 30669962)) can be used to align the reads to the reference genome (for example hg38, hg19, hg18, hg17 or hg16).

[0187] The genomic location assigned to each methylome sequence in the alignment is based on the reference genome adopted. The genomic locations listed in Tables 1, 1b, 2 to 9 disclosed herein correspond to reference genome hg19. The corresponding locations in a different reference genome can be found using public available tools known in the art. An example of these tools is LiftOver (http://genome.ucsc.edu/).

[0188] In certain embodiments, the method comprises removing duplications of reads of the same DNA molecule (i.e. duplications of reads of the same cfDNA molecule). In this step, sequence reads having exactly the same sequence and start and end base pairs (i.e. the same unclipped alignment start and unclipped alignment end of the sequence) are removed, as they are likely to be duplicate sequence reads of the same sequence (i.e. duplicate of reads of the same cfDNA molecule). For example, PCR duplications can be removed as part of the aligning step, such as using Picard tools v2.1.0 (http://broadinstitute.github.io/picard).

[0189] The method comprises determining the average methylation ratio at 10 or more of the genomic regions for which the average methylation ratio has been determined, each genomic region being selected from the group consisting of: [0190] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0191] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence.

[0192] In certain embodiments, each genomic region for which the average methylation ratio has been determined is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 50, 100, 1000, 10,000 characterized methylome sequences. Preferably each genomic region is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 50, 100, or 1000 characterized methylome sequences. In certain preferred embodiments, each genomic region is covered by at least one sequence read of at least 10 characterized methylome sequences, for example at least one sequence read of at least 10, at least 15, at least 20, at least 25, at least 50, at least 100, or at least 1000 characterized methylome sequences.

[0193] In certain embodiments, each genomic region for which the average methylation ratio has been determined is covered by at least 2 sequence reads, for example at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads. Preferably, each genomic region is covered by at least 5 sequence reads, for example at least 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads. More preferably, each genomic region is covered by at least 10 sequence reads, for example at least 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads.

[0194] In embodiments wherein each genomic region for which the average methylation ratio has been determined is covered by at least 2 sequence reads (for example at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads) preferably each sequence read or the majority of the sequence reads (for example at least 50%, 60%, 70%, 80% or 90% of the sequence reads) are from different characterized methylome sequences. More preferably, each sequence read or at least 60%, 70%, 80% or 90% of the sequence reads are from different characterized methylome sequences.

[0195] In certain embodiments the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, 400 or more genomic regions, 500 or more genomic regions, 600 or more genomic regions, 700 or more genomic regions, 800 or more genomic regions, 900 or more genomic regions, or 1000 genomic regions. Each genomic region may be selected from the group consisting of: [0196] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0197] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0198] The genomic regions are preferably each different from each other. In certain preferred embodiments, the method comprises determining the average methylation ratio at 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, 400 or more genomic regions, 500 or more genomic regions, 600 or more genomic regions, 700 or more genomic regions, 800 or more genomic regions, 900 or more genomic regions, or 1000 genomic regions. Each genomic region may be selected from the group consisting of: [0199] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0200] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0201] In certain preferred embodiments, the method comprises determining the average methylation ratio at 500 or more genomic regions, 600 or more genomic regions, 700 or more genomic regions, 800 or more genomic regions, 900 or more genomic regions, or 1000 genomic regions. Each genomic region may be selected from the group consisting of: [0202] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0203] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0204] In certain preferred embodiments, the method comprises determining the average methylation ratio at 800 or more genomic regions, 900 or more genomic regions, or 1000 genomic regions. Each genomic region may be selected from the group consisting of: [0205] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0206] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0207] In one embodiment, each genomic region is selected from the group consisting of: [0208] a 100 to 200 bp region comprising or having a genomic location defined in Tables 3 and 4, and [0209] a 2 to 99 bp region within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus.

[0210] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0211] In certain embodiments, each genomic region is selected from the group consisting of: [0212] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 and 3, and a 2 to 99 bp region within a genomic location defined in Tables 1 and 3 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Tables 1 and 3, and 10 to 99 bp region within a genomic location defined in Tables 1 and 3 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 1 and 3, and 50 to 99 bp region within a genomic location defined in Tables 1 and 3 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 1 and 3, and 80 to 99 bp region within a genomic location defined in Tables 1 and 3 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Tables 1 and 3.

[0213] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0214] In certain embodiments, each genomic region is selected from the group consisting of: [0215] a 100 to 200 bp region comprising or having a genomic location defined in Tables 2 and 4, and a 2 to 99 bp region within a genomic location defined in Tables 2 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Tables 2 and 4, and 10 to 99 bp region within a genomic location defined in Tables 2 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 2 and 4, and 50 to 99 bp region within a genomic location defined in Tables 2 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 2 and 4, and 80 to 99 bp region within a genomic location defined in Tables 2 and 4 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Tables 2 and 4.

[0216] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0217] In certain preferred embodiments, each genomic region is selected from the group consisting of: [0218] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 and 2, and a 2 to 99 bp region within a genomic location defined in Tables 1 and 2 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Tables 1 and 2, and 10 to 99 bp region within a genomic location defined in Tables 1 and 2 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 1 and 2, and 50 to 99 bp region within a genomic location defined in Tables 1 and 2 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 1 and 2, and 80 to 99 bp region within a genomic location defined in Tables 1 and 2 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Tables 1 and 2.

[0219] In such preferred embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0220] In certain embodiments, each genomic region is selected from the group consisting of: [0221] a 100 to 200 bp region comprising or having a genomic location defined in Tables 3 and 4, and a 2 to 99 bp region within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Tables 3 and 4, and 10 to 99 bp region within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 3 and 4, and 50 to 99 bp region within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Tables 3 and 4, and 80 to 99 bp region within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Tables 3 and 4.

[0222] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0223] In one preferred embodiment, each genomic region is selected from the group consisting of: [0224] a 100 to 200 bp region comprising or having a genomic location defined in Table 5, and a 2 to 99 bp region within a genomic location defined in Table 5 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Table 5, and 10 to 99 bp region within a genomic location defined in Table 5 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 5, and 50 to 99 bp region within a genomic location defined in Table 5 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 5, and 80 to 99 bp region within a genomic location defined in Table 5 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Table 5.

TABLE-US-00006 TABLE 5 A preferred subset of hypermethylated and hypomethylated region genomic locations (The genomic locations are locations with reference to hg19) Hyper- or hypomethylated Chromosome start end region chr4 9104451 9104550 hypo chr12 54441001 54441100 hyper chr1 153174701 153174800 hypo chr4 9104401 9104500 hypo chr1 248308901 248309000 hypo chr12 4030001 4030100 hypo chr2 91936151 91936250 hypo chr2 198063601 198063700 hyper chr10 6779901 6780000 hypo chr19 56346551 56346650 hypo chr14 47670001 47670100 hypo chr17 77386351 77386450 hypo chr8 105988201 105988300 hypo chr2 54901001 54901100 hyper chr14 97924301 97924400 hypo chr14 95237401 95237500 hyper chr17 79422551 79422650 hyper chr14 97924251 97924350 hypo chr1 119526251 119526350 hyper chr14 37125801 37125900 hyper chr2 177012701 177012800 hyper chr14 47670201 47670300 hypo chr17 3030101 3030200 hypo chr4 77226351 77226450 hyper chr3 38835251 38835350 hypo chr5 87439351 87439450 hyper chr9 22005201 22005300 hyper chr2 198063651 198063750 hyper chr12 131512601 131512700 hypo chr2 879451 879550 hypo chr5 87439401 87439500 hyper chr1 204165251 204165350 hypo chr9 132650551 132650650 hyper chr20 1975351 1975450 hypo chr17 79422601 79422700 hyper chr9 110399201 110399300 hyper chr6 170531301 170531400 hyper chr9 132650601 132650700 hyper chr7 45066701 45066800 hyper chr8 139124351 139124450 hypo chr1 207103601 207103700 hyper chr8 99950901 99951000 hyper chr8 99950851 99950950 hyper chr7 45066651 45066750 hyper chr9 38437301 38437400 hypo chr12 4361001 4361100 hypo chr17 72776151 72776250 hyper chr12 4361051 4361150 hypo chr2 204571201 204571300 hyper chr1 162527451 162527550 hypo chr1 207103651 207103750 hyper chr4 108814501 108814600 hyper chr14 37125851 37125950 hyper chr8 139124401 139124500 hypo chr4 77226301 77226400 hyper chr20 1975301 1975400 hypo chr2 232186901 232187000 hyper chr20 5282901 5283000 hypo chr20 1975401 1975500 hypo chr6 152804751 152804850 hypo chr19 55042751 55042850 hypo chr12 132102101 132102200 hypo chr17 77386401 77386500 hypo chr14 47670051 47670150 hypo chr9 140586201 140586300 hyper chr5 179344551 179344650 hyper chr1 143907501 143907600 hypo chr1 143907451 143907550 hypo chr1 119526201 119526300 hyper chr6 152804701 152804800 hypo chr2 228324901 228325000 hyper chr19 55042801 55042900 hypo chr3 160475151 160475250 hyper chr1 182021751 182021850 hypo chr1 182021701 182021800 hypo chr8 111906551 111906650 hypo chr6 170531351 170531450 hyper chr2 232186851 232186950 hyper chr8 130365301 130365400 hypo chr2 117006251 117006350 hypo chr3 194868701 194868800 hyper chr18 11153901 11154000 hypo chr18 11153851 11153950 hypo chr1 175490551 175490650 hypo chr3 160475201 160475300 hyper chr19 2776001 2776100 hyper chr3 193096401 193096500 hypo chr2 228324851 228324950 hyper chr8 120779851 120779950 hypo chr12 131512551 131512650 hypo chr9 125796751 125796850 hyper chr3 194868651 194868750 hyper chr10 7567801 7567900 hypo chr1 175490601 175490700 hypo chr1 68154601 68154700 hyper chr17 3030151 3030250 hypo chr2 91910601 91910700 hypo chr2 91910551 91910650 hypo chr14 97924451 97924550 hypo chr9 22005551 22005650 hyper chr11 121986951 121987050 hypo chr14 97497051 97497150 hypo chr1 95172951 95173050 hypo chr3 38835201 38835300 hypo chr14 37124151 37124250 hyper chr4 13544651 13544750 hyper chrX 150944951 150945050 hypo chr3 46448751 46448850 hyper chr1 248153651 248153750 hypo chr20 19866701 19866800 hypo chr2 3633351 3633450 hypo chr14 104742101 104742200 hypo chr20 5450901 5451000 hypo chr1 153175251 153175350 hypo chr9 22005501 22005600 hyper chr12 116997101 116997200 hyper chr15 98646401 98646500 hypo chr12 130494601 130494700 hypo chr4 120502101 120502200 hypo chr7 5518851 5518950 hyper chr17 55562951 55563050 hyper chr7 57510201 57510300 hypo chr5 3002401 3002500 hypo chr3 100690901 100691000 hypo chr3 100690851 100690950 hypo chr14 97924501 97924600 hypo chr2 206551451 206551550 hyper chr1 2876551 2876650 hypo chr12 4030051 4030150 hypo chr12 132663901 132664000 hypo chr1 153174651 153174750 hypo chr6 34252801 34252900 hyper chr2 177012651 177012750 hyper chr6 45296101 45296200 hyper chr12 8438301 8438400 hypo chr2 177012551 177012650 hyper chr1 2876501 2876600 hypo chr3 194868751 194868850 hyper chr7 6476051 6476150 hyper chr3 127453801 127453900 hyper chr3 127453851 127453950 hyper chr12 7062101 7062200 hyper chr14 59296601 59296700 hypo chr9 91006701 91006800 hyper chr9 110399251 110399350 hyper chr2 71116551 71116650 hyper chr3 72227251 72227350 hyper chr2 60880201 60880300 hypo chr7 129411101 129411200 hyper chr12 111536901 111537000 hyper chr17 55562901 55563000 hyper chr4 101438801 101438900 hyper chr17 21911401 21911500 hypo chr11 47939651 47939750 hyper chr2 54900951 54901050 hyper chr14 59296651 59296750 hypo chr16 10206551 10206650 hypo chr1 143907551 143907650 hypo chr14 47669951 47670050 hypo chr19 33162851 33162950 hyper chr14 93412701 93412800 hypo chr12 130711351 130711450 hypo chr2 100426951 100427050 hypo chr2 100426901 100427000 hypo chr9 22005601 22005700 hyper chr2 2581301 2581400 hypo chr17 59534651 59534750 hyper chr10 6779951 6780050 hypo chr5 176758401 176758500 hyper chr9 96080351 96080450 hyper chr7 129411151 129411250 hyper chr17 79422501 79422600 hyper chr15 86098601 86098700 hyper chr22 50618601 50618700 hyper chr19 55104601 55104700 hypo chr10 94450951 94451050 hyper chr14 47670101 47670200 hypo chr8 130365251 130365350 hypo chr1 2876451 2876550 hypo chr1 204165301 204165400 hypo chr2 172974201 172974300 hyper chr2 172974151 172974250 hyper chr17 72776201 72776300 hyper chr19 55036701 55036800 hypo chr1 95173001 95173100 hypo chr12 4361101 4361200 hypo chr7 5518901 5519000 hyper chr12 6665301 6665400 hyper chr1 169696601 169696700 hypo chr12 132142051 132142150 hypo chr12 132142001 132142100 hypo chr8 56361951 56362050 hypo chr16 23988951 23989050 hypo chr9 91006751 91006850 hyper chr2 228324951 228325050 hyper chr5 134826351 134826450 hyper chr2 879501 879600 hypo chr4 53862451 53862550 hyper chr14 37124201 37124300 hyper chr10 6664951 6665050 hypo chr8 56106801 56106900 hypo chr8 142289801 142289900 hypo chr14 104742051 104742150 hypo chr5 5492401 5492500 hypo chr20 31123201 31123300 hyper chr2 89215151 89215250 hypo chr2 89215101 89215200 hypo chr2 232186951 232187050 hyper chr5 10445501 10445600 hyper chr3 177397701 177397800 hyper chr11 47939701 47939800 hyper chr6 34252751 34252850 hyper chr19 57646101 57646200 hypo chr4 74809851 74809950 hyper chr19 33162801 33162900 hyper chr1 64937351 64937450 hyper chr1 68154551 68154650 hyper chr2 172945201 172945300 hyper chr17 22023751 22023850 hypo chr1 65399451 65399550 hyper chr19 46526251 46526350 hyper chr2 171569151 171569250 hyper chr10 31423501 31423600 hyper chr14 37125901 37126000 hyper chr11 57519401 57519500 hypo chr16 23939051 23939150 hypo chr19 29281851 29281950 hypo chr19 29281801 29281900 hypo chr10 94450901 94451000 hyper chr1l 6865801 6865900 hypo chr9 140586151 140586250 hyper chr6 41394801 41394900 hyper chr4 108814551 108814650 hyper chrX 150945001 150945100 hypo chr19 18508551 18508650 hyper chr9 96080301 96080400 hyper chr14 95237351 95237450 hyper chr17 59532801 59532900 hyper chr20 5282851 5282950 hypo chr8 142289851 142289950 hypo chr5 72676801 72676900 hyper chr17 22017001 22017100 hypo chr2 71126251 71126350 hyper chr2 59477151 59477250 hypo chr7 149112151 149112250 hyper chr1 4079101 4079200 hypo chr17 78724151 78724250 hyper chr14 60976901 60977000 hyper chr9 5756751 5756850 hypo chr22 17073551 17073650 hypo

[0225] In such embodiments, preferably the method comprises determining the average methylation ratio at 10 or more genomic regions, 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, or 250 genomic regions. For example, the method comprises determining the average methylation ratio at 10 or more genomic regions, 50 or more genomic regions or 100 or more genomic regions.

[0226] In another preferred embodiments, each genomic region is selected from the group consisting of: [0227] a 100 to 200 bp region comprising or having a genomic location defined in Table 6, and a 2 to 99 bp region within a genomic location defined in Table 6 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Table 6, and 10 to 99 bp region within a genomic location defined in Table 6 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 6, and 50 to 99 bp region within a genomic location defined in Table 6 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 6, and 80 to 99 bp region within a genomic location defined in Table 6 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Table 6.

TABLE-US-00007 TABLE 6 A preferred subset of hypomethylated region genomic locations (The genomic locations are locations with reference to hg19) Hyper- or hypomethylated Chromosome start end region chr4 9104451 9104550 hypo chr1 153174701 153174800 hypo chr4 9104401 9104500 hypo chr1 248308901 248309000 hypo chr12 4030001 4030100 hypo chr2 91936151 91936250 hypo chr10 6779901 6780000 hypo chr19 56346551 56346650 hypo chr14 47670001 47670100 hypo chr17 77386351 77386450 hypo chr8 105988201 105988300 hypo chr14 97924301 97924400 hypo chr14 97924251 97924350 hypo chr14 47670201 47670300 hypo chr17 3030101 3030200 hypo chr3 38835251 38835350 hypo chr12 131512601 131512700 hypo chr2 879451 879550 hypo chr1 204165251 204165350 hypo chr20 1975351 1975450 hypo chr8 139124351 139124450 hypo chr9 38437301 38437400 hypo chr12 4361001 4361100 hypo chr12 4361051 4361150 hypo chr1 162527451 162527550 hypo chr8 139124401 139124500 hypo chr20 1975301 1975400 hypo chr20 5282901 5283000 hypo chr20 1975401 1975500 hypo chr6 152804751 152804850 hypo chr19 55042751 55042850 hypo chr12 132102101 132102200 hypo chr17 77386401 77386500 hypo chr14 47670051 47670150 hypo chr1 143907501 143907600 hypo chr1 143907451 143907550 hypo chr6 152804701 152804800 hypo chr19 55042801 55042900 hypo chr1 182021751 182021850 hypo chr1 182021701 182021800 hypo chr8 111906551 111906650 hypo chr8 130365301 130365400 hypo chr2 117006251 117006350 hypo chr18 11153901 11154000 hypo chr18 11153851 11153950 hypo chr1 175490551 175490650 hypo chr3 193096401 193096500 hypo chr8 120779851 120779950 hypo chr12 131512551 131512650 hypo chr10 7567801 7567900 hypo chr1 175490601 175490700 hypo chr17 3030151 3030250 hypo chr2 91910601 91910700 hypo chr2 91910551 91910650 hypo chr14 97924451 97924550 hypo chr11 121986951 121987050 hypo chr14 97497051 97497150 hypo chr1 95172951 95173050 hypo chr3 38835201 38835300 hypo chrX 150944951 150945050 hypo chr1 248153651 248153750 hypo chr20 19866701 19866800 hypo chr2 3633351 3633450 hypo chr14 104742101 104742200 hypo chr20 5450901 5451000 hypo chr1 153175251 153175350 hypo chr15 98646401 98646500 hypo chr12 130494601 130494700 hypo chr4 120502101 120502200 hypo chr7 57510201 57510300 hypo chr5 3002401 3002500 hypo chr3 100690901 100691000 hypo chr3 100690851 100690950 hypo chr14 97924501 97924600 hypo chr1 2876551 2876650 hypo chr12 4030051 4030150 hypo chr12 132663901 132664000 hypo chr1 153174651 153174750 hypo chr12 8438301 8438400 hypo chr1 2876501 2876600 hypo chr14 59296601 59296700 hypo chr2 60880201 60880300 hypo chr17 21911401 21911500 hypo chr14 59296651 59296750 hypo chr16 10206551 10206650 hypo chr1 143907551 143907650 hypo chr14 47669951 47670050 hypo chr14 93412701 93412800 hypo chr12 130711351 130711450 hypo chr2 100426951 100427050 hypo chr2 100426901 100427000 hypo chr2 2581301 2581400 hypo chr10 6779951 6780050 hypo chr19 55104601 55104700 hypo chr14 47670101 47670200 hypo chr8 130365251 130365350 hypo chr1 2876451 2876550 hypo chr1 204165301 204165400 hypo chr19 55036701 55036800 hypo chr1 95173001 95173100 hypo chr12 4361101 4361200 hypo chr1 169696601 169696700 hypo chr12 132142051 132142150 hypo chr12 132142001 132142100 hypo chr8 56361951 56362050 hypo chr16 23988951 23989050 hypo chr2 879501 879600 hypo chr10 6664951 6665050 hypo chr8 56106801 56106900 hypo chr8 142289801 142289900 hypo chr14 104742051 104742150 hypo chr5 5492401 5492500 hypo chr2 89215151 89215250 hypo chr2 89215101 89215200 hypo chr19 57646101 57646200 hypo chr17 22023751 22023850 hypo chr11 57519401 57519500 hypo chr16 23939051 23939150 hypo chr19 29281851 29281950 hypo chr19 29281801 29281900 hypo chr11 6865801 6865900 hypo chrX 150945001 150945100 hypo chr20 5282851 5282950 hypo chr8 142289851 142289950 hypo chr17 22017001 22017100 hypo chr2 59477151 59477250 hypo chr1 4079101 4079200 hypo chr9 5756751 5756850 hypo chr22 17073551 17073650 hypo chr22 24979551 24979650 hypo chr11 7961101 7961200 hypo chr11 7961051 7961150 hypo chr5 19531551 19531650 hypo chr1 175490501 175490600 hypo chr5 19531501 19531600 hypo chr21 44375551 44375650 hypo chr7 39056351 39056450 hypo chr14 47670251 47670350 hypo chr1 148903551 148903650 hypo chr3 192960551 192960650 hypo chr19 55042301 55042400 hypo chr14 104742151 104742250 hypo chr4 157059301 157059400 hypo chr3 33757901 33758000 hypo chr4 3895151 3895250 hypo chr14 97924401 97924500 hypo chr7 39056301 39056400 hypo chr2 242190801 242190900 hypo chr19 55042251 55042350 hypo chr6 159872251 159872350 hypo

[0228] In such embodiments, preferably the method comprises determining the average methylation ratio at 10 or more genomic regions, at 12 or more genomic regions, for example at 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, or 150 genomic regions. For example, the method comprises determining the average methylation ratio at 10 or more genomic regions, 50 or more genomic regions or 100 or more genomic regions.

[0229] In another preferred embodiment, each genomic region is selected from the group consisting of:

a 100 to 200 bp region comprising or having a genomic location defined in Table 7, and a 2 to 99 bp region within a genomic location defined in Table 7 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Table 7, and 10 to 99 bp region within a genomic location defined in Table 7 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 7, and 50 to 99 bp region within a genomic location defined in Table 7 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 7, and 80 to 99 bp region within a genomic location defined in Table 7 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Table 7.

TABLE-US-00008 TABLE 7 A preferred subset of hypermethylated and hypomethylated region genomic locations (The genomic locations are locations with reference to hg19) Hyper- or hypomethylated Chromosome start end region chr4 9104451 9104550 hypo chr12 54441001 54441100 hyper chr1 153174701 153174800 hypo chr4 9104401 9104500 hypo chr1 248308901 248309000 hypo chr12 4030001 4030100 hypo chr2 91936151 91936250 hypo chr2 198063601 198063700 hyper chr10 6779901 6780000 hypo chr19 56346551 56346650 hypo chr14 47670001 47670100 hypo chr17 77386351 77386450 hypo chr8 105988201 105988300 hypo chr2 54901001 54901100 hyper chr14 97924301 97924400 hypo chr14 95237401 95237500 hyper chr17 79422551 79422650 hyper chr14 97924251 97924350 hypo chr1 119526251 119526350 hyper chr14 37125801 37125900 hyper chr2 177012701 177012800 hyper chr14 47670201 47670300 hypo chr17 3030101 3030200 hypo chr4 77226351 77226450 hyper chr3 38835251 38835350 hypo chr5 87439351 87439450 hyper chr9 22005201 22005300 hyper chr2 198063651 198063750 hyper chr12 131512601 131512700 hypo chr2 879451 879550 hypo chr5 87439401 87439500 hyper chr1 204165251 204165350 hypo chr9 132650551 132650650 hyper chr20 1975351 1975450 hypo chr17 79422601 79422700 hyper chr9 110399201 110399300 hyper chr6 170531301 170531400 hyper chr9 132650601 132650700 hyper chr7 45066701 45066800 hyper chr8 139124351 139124450 hypo chr1 207103601 207103700 hyper chr8 99950901 99951000 hyper chr8 99950851 99950950 hyper chr7 45066651 45066750 hyper chr9 38437301 38437400 hypo chr12 4361001 4361100 hypo chr17 72776151 72776250 hyper chr12 4361051 4361150 hypo chr2 204571201 204571300 hyper chr1 162527451 162527550 hypo chr1 207103651 207103750 hyper chr4 108814501 108814600 hyper chr14 37125851 37125950 hyper chr8 139124401 139124500 hypo chr4 77226301 77226400 hyper chr20 1975301 1975400 hypo chr2 232186901 232187000 hyper chr20 5282901 5283000 hypo chr20 1975401 1975500 hypo chr6 152804751 152804850 hypo chr19 55042751 55042850 hypo chr12 132102101 132102200 hypo chr17 77386401 77386500 hypo chr14 47670051 47670150 hypo chr9 140586201 140586300 hyper chr5 179344551 179344650 hyper chr1 143907501 143907600 hypo chr1 143907451 143907550 hypo chr1 119526201 119526300 hyper chr6 152804701 152804800 hypo chr2 228324901 228325000 hyper chr19 55042801 55042900 hypo chr3 160475151 160475250 hyper chr1 182021751 182021850 hypo chr1 182021701 182021800 hypo chr8 111906551 111906650 hypo chr6 170531351 170531450 hyper chr2 232186851 232186950 hyper chr8 130365301 130365400 hypo chr2 117006251 117006350 hypo chr3 194868701 194868800 hyper chr18 11153901 11154000 hypo chr18 11153851 11153950 hypo chr1 175490551 175490650 hypo chr3 160475201 160475300 hyper chr19 2776001 2776100 hyper chr3 193096401 193096500 hypo chr2 228324851 228324950 hyper chr8 120779851 120779950 hypo chr12 131512551 131512650 hypo chr9 125796751 125796850 hyper chr3 194868651 194868750 hyper chr10 7567801 7567900 hypo chr1 175490601 175490700 hypo chr1 68154601 68154700 hyper chr17 3030151 3030250 hypo chr2 91910601 91910700 hypo chr2 91910551 91910650 hypo chr14 97924451 97924550 hypo chr9 22005551 22005650 hyper

[0230] In such embodiments, preferably the method comprises determining the average methylation ratio at 10 or more genomic regions, at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, or 100 genomic regions. For example, the method comprises determining the average methylation ratio at 10 or more genomic regions, 50 or more genomic regions or 100 genomic regions.

[0231] In certain preferred embodiments, at least 25% of the genomic regions comprise, have or are within a genomic location defined in Tables 1 and/or 2. For example, at least 25% of the genomic regions comprise or have a genomic location defined in Tables 1 and/or 2.

[0232] In certain preferred embodiments, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise, have or are within a genomic location defined in Tables 1 and/or 2. For example, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise or have a genomic location defined in Tables 1 and/or 2.

[0233] In certain embodiments, at least 25% of the genomic regions comprise, have or are within a genomic location defined in Tables 3 and/or 4. For example, at least 25% of the genomic regions comprise or have a genomic location defined in Tables 3 and/or 4.

[0234] In certain embodiments, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise, have or are within a genomic location defined in Tables 3 and/or 4. For example, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise or have a genomic location defined in Tables 3 and/or 4.

[0235] In certain embodiments, at least 25% of the genomic regions comprise, have or are within a genomic location defined in Tables 1 and/or 3. For example, at least 25% of the genomic regions comprise or have a genomic location defined in Tables 1 and/or 3.

[0236] In certain embodiments, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise, have or are within a genomic location defined in Tables 1 and/or 3. For example, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise or have a genomic location defined in Tables 1 and/or 3.

[0237] In certain embodiments, at least 25% of the genomic regions comprise, have or are within a genomic location defined in Tables 2 and/or 4. For example, at least 25% of the genomic regions comprise or have a genomic location defined in Tables 2 and/or 4.

[0238] In certain embodiments, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise, have or are within a genomic location defined in Tables 3 and/or 4. For example, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or all of the genomic regions comprise or have a genomic location defined in Tables 3 and/or 4.

[0239] In certain preferred embodiments, determining the average methylation ratio for a genomic region comprises calculating the sum of the methylation ratios of all CpGs within the genomic region and dividing the sum by the number of CpGs within the genomic region. In such embodiments, the average methylation ratio may also be referred to as the mean methylation ratio. For the avoidance of doubt, if a genomic region has only one CpG locus, the average methylation ratio for the genomic region is the same as the methylation ratio for the single CpG locus in the genomic region.

[0240] The method of the present invention comprises calculating a methylation score using the average methylation ratio for each genomic region for which the average methylation ratio has been determined.

[0241] In certain embodiments, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0242] determining the median or the mean of the average methylation ratios for all genomic regions (i.e. all genomic regions for which an average methylation ratio has been determined in the method); or [0243] determining the median or the mean of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median or the mean of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or [0244] comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0245] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0246] determining the median of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or [0247] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or [0248] comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0249] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0250] determining the median of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or [0251] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score.

[0252] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0253] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score.

[0254] In very preferred embodiments wherein calculating a methylation score using the average methylation ratio for each genomic region comprises determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score, the first group of genomic regions are all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 1 or 2), and the second group of genomic regions are all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 3 or 4, or Table 6).

[0255] In another embodiment, the first group of genomic regions are all of the genomic regions (for which the average methylation ratio has been determined) having a methylation pattern specific to prostate tissue (i.e. selected from those comprising, having or within a genomic location defined in Table 1 or 3), and the second group of genomic regions are all of the genomic regions (for which the average methylation ratio has been determined) having a methylation pattern specific to prostate cancer (i.e. selected from those comprising, having or within a genomic location defined in Table 2 or 4).

[0256] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0257] determining the median of the average methylation ratios for all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 1 or 2) to obtain a first methylation score and determining the median of the average methylation ratios for all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 3 or 4, or Table 6) to obtain a second methylation score.

[0258] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0259] determining the median of the average methylation ratios for all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those those comprising, having or within a genomic location defined in Table 1 or 2) to obtain a first methylation score.

[0260] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0261] determining the median of the average methylation ratios for all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those those comprising, having or within a genomic location defined in Table 3 or 4, or Table 6) to obtain a second methylation score.

[0262] In one embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region. In such embodiments, preferably the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by: [0263] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0264] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0265] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0266] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0267] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0268] a cfDNA sample from a different subject having prostate cancer, wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50 or 100 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of prostate cancer fraction); [0269] a characterized methylome sequence of a white blood cell; [0270] a characterized methylome sequence of a prostate cancer cell line; [0271] a characterized methylome sequence of a cancerous prostate cell; and/or [0272] a characterized methylome sequence of a non-cancerous prostate cell.

[0273] In one preferred embodiment, the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by [0274] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0275] a sample of white blood cells from a subject, for example the subject or a healthy subject; and/or [0276] a characterized methylome sequence of a white blood cell.

[0277] The method of the present invention comprises analyzing the methylation ratio scores to determine the level of prostate cancer fraction in the cfDNA sample. For example, no level (for example no detectable level) of prostate cancer fraction in the cfDNA sample may be determined. Alternatively, a level of cancer fraction in the cfDNA sample may be determined. The minimum percentage level of prostate cancer fraction in the cfDNA sample that may be determined may be 0.01% of cancer fraction in the cfDNA sample. In certain embodiments, the minimum percentage level of prostate cancer fraction in the cfDNA sample that may be determined may be 0.02%, 0.03%, 0.04%, 0.06%, 0.07%, 0.08%, 0.05%, 0.09%, 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, 0.9%, or 1% of cancer fraction in the cfDNA sample. For example, the minimum percentage level of prostate cancer fraction in the cfDNA sample that may be determined may be 0.01%, 0.05%, 0.1% or 0.5% of cancer fraction in the cfDNA sample. Preferably, the minimum percentage level of prostate cancer fraction in the cfDNA is 0.01%.

[0278] The method comprises analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0279] Preferably, analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores. For example, the method may comprise comparing the methylation score to one reference methylation scores. In certain embodiments, the method comprises comparing the methylation score to two or more reference methylation scores, for example 2, 3, 4, 5, 6, 8, 9, 10, 12, 15, 20, 30, 50, 100, 200, 300, 400, 500 or 1000 reference methylation scores. In certain embodiments, the method comprises comparing the methylation score to 5 or more reference methylation scores, for example 10 or more, 15 or more, 20 or more, 30, or more 50, or more 100, or more 200, or more 300, or more 400, or more 500 or 1000 or more reference methylation scores.

[0280] In embodiments wherein the method comprises comparing the methylation score to two or more reference methylation scores, the reference methylation scores may come from different types of reference samples and/or reference methylomes (for example a cfDNA sample from a healthy subject and a cancer cell line sample) and/or the same type of reference samples or reference methylomes but from different sources (for example, two or more cfDNA samples each from a different a healthy subject).

[0281] A reference methylation score is a methylation score calculated for the same genomic regions (for example, calculated using the average methylation ratio for the same genomic regions) in a reference sample or reference methylome. A reference sample or reference methylome may be selected from the group consisting of:

a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject;
a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a sample of white blood cells from a subject, for example the subject or a healthy subject;
a cfDNA sample from a different subject having prostate cancer, wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50 or 100 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of prostate cancer fraction);
a characterized methylome sequence of a white blood cell;
a characterized methylome sequence of a prostate cancer cell line;
a characterized methylome sequence of a cancerous prostate cell; and/or
a characterized methylome sequence of a non-cancerous prostate cell.

[0282] A reference sample or reference methylome may be one that can be used to represent a sample having 0% tumour fraction, for example a reference sample or reference methylome selected from one or more of the following

a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a sample of white blood cells from a subject, for example the subject or a healthy subject; and/or
a characterized methylome sequence of a white blood cell.

[0283] A reference sample or reference methylome may be one that can be used to represent a sample having 100% tumour fraction, for example a reference sample or reference methylome selected from one or more of the following

a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a characterized methylome sequence of a prostate cancer cell line; and/or
a characterized methylome sequence of a cancerous prostate cell.

[0284] A reference sample or reference methylome may be one that can be used to represent a sample having 10 to 90% tumour fraction, for example one or more cfDNA samples from different subjects having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is/are known. A level of prostate cancer fraction in each cfDNA sample can be determined by looking at genomic markers.

[0285] Preferably, analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores that can be used to represent a sample having 100% tumour fraction, and can be used to represent a sample having 0% tumour fraction, and optionally can be used to represent a sample having 10-90% tumour fraction. For example, analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises:

comparing the methylation score to one or more reference methylation scores for a reference sample or reference methylome selected from the group consisting of: [0286] a cfDNA sample from a healthy subject, for example a healthy age-matched subject, [0287] a sample of white blood cells from a subject, for example the subject or a healthy subject, and/or [0288] a characterized methylome sequence of a white blood cell;
and
comparing the methylation score to one or more reference methylation scores for a reference sample or reference methylome selected from the group consisting of: [0289] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0290] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0291] a characterized methylome sequence of a prostate cancer cell line; and/or [0292] a characterized methylome sequence of a cancerous prostate cell.
and optionally comparing the methylation score to one or more reference methylation scores for one or more cfDNA samples from different subjects having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is/are known.

[0293] Preferably, the reference methylation score for a reference sample or reference methylome that a methylation ratio score is compared to is calculated in the same way as the methylation score for the sample obtained from the subject (i.e. the sample that the method of the invention is being carried out in respect of). For example, if the methylation ratio for the selected genomic regions of the sample obtained from the subject is calculated by determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score, the reference methylation score for a reference sample or reference methylome is calculated by determining the median (or the mean) of the average methylation ratios for the same first group of genomic regions to obtain a first reference methylation score and/or determining the median (or the mean) of the average methylation ratios for the same second group of genomic regions to obtain a second reference methylation score.

[0294] Or, for example, if the methylation ratio for the selected genomic regions of the sample obtained from the subject is calculated by determining the median (or the mean) of the average methylation ratios for all genomic regions, the reference methylation score for a reference sample or reference methylome is calculated by determining the median (or the mean) of the average methylation ratios for the same genomic regions.

[0295] In embodiments wherein the method comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region, analyzing the methylation ratio scores to determine the level of prostate cancer fraction in the cfDNA sample may comprise determining how many methylation ratio scores are indicative of prostate cancer fraction in the cfDNA sample.

[0296] In certain embodiments, analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises using a mathematical model, such as a linear regression model or another linear model (for example, a general linear model, a heteroscedastic model, a generalised linear model, or a hierarchical linear model).

[0297] In certain embodiments, analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises using a mathematical model that compares the methylation score for the sample to reference methylation scores that can be used to represent a sample having 100% tumour fraction, and can be used to represent a sample having 0% tumour fraction, and optionally can be used to represent a sample having 10-90% tumour fraction. For example, the method comprises using mathematical model that compares the methylation score for the sample to reference methylation scores for a cfDNA sample from a healthy subject, for example a healthy age-matched subject (0% tumour fraction) and/or a characterized methylome sequence of a white blood cell (0% tumour fraction) and/or a sample of white blood cells from a subject, for example the subject or a healthy subject, (0% tumour fraction) and/or a characterized methylome sequence of a prostate cancer cell line (100% tumour fraction) and/or a prostate cancer biopsy sample from a prostate cancer patient (100% tumour fraction) and/or one or more cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of prostate cancer fraction (10-90% tumour fraction).

[0298] In one embodiment, the method comprises using mathematical model that compares the methylation score for the sample to reference methylation scores for a cfDNA sample from a healthy subject, for example a healthy age-matched subject (0% tumour fraction) and/or a characterized methylome sequence of a prostate cancer cell line (100% tumour fraction) and/or a prostate cancer biopsy sample from a prostate cancer patient (100% tumour fraction) and/or one or more cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of prostate cancer fraction (10-90% tumour fraction).

[0299] The method may further comprise measuring the level of prostate-specific antigen (PSA) in a sample of blood from the subject. It may also comprise determining if the subject has an abnormal level of PSA in the blood (for example a level of PSA in the blood of at least 4.0 ng/mL). An abnormal level of PSA in the blood may be, for example, a level of PSA in the blood of at least 4.0 ng/mL). A normal level of PSA in the blood may, for example, be a level of PSA in the blood of 4.0 ng/mL or less.

[0300] In one preferred embodiment, the method is for screening, monitoring, and/or prognostication of prostate cancer, wherein prostate cancer with a poor prognosis is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%. For example, a prostate cancer with a poor prognosis is predicted when at least 0.01% prostate cancer fraction is determined, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction is determined.

[0301] In some instances, a “poor” prognosis refers to a low likelihood that a subject will likely respond favorably to a drug or set of drugs, is in complete or partial remission, or there is a decrease and/or a stop in the progression of prostate cancer. In some instances, a “poor” prognosis refers to a survival of a subject that is expected to be from less than 5 years to less than 1 month. In some instances, a “poor” prognosis refers to a survival of a subject in which the survival of the subject upon treatment is expected to be from less than 5 years to less than 1 month.

[0302] In one preferred embodiment, the method is for detection of prostate cancer, wherein prostate cancer is detected when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction.

[0303] In one preferred embodiment, the method is for screening, monitoring, and/or prognostication of prostate cancer, wherein prostate cancer with a poor prognosis is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%, for example at least 0.01% prostate cancer fraction, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction.

[0304] In one preferred embodiment, the method is for detecting, screening and/or prognostication of metastatic prostate cancer, wherein metastatic prostate cancer is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction.

[0305] In one preferred embodiment, the method is for selecting treatment of prostate cancer or ascertaining whether treatment is working in prostate cancer, wherein a new treatment is selected when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction.

[0306] In one preferred embodiment, the method is for ascertaining whether treatment of prostate cancer is working, wherein it is determined that the treatment is not working when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% prostate cancer fraction.

[0307] The method may further comprise repeating the method on second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of prostate cancer fraction in each sample. Preferably, the second sample is of the same type as the first sample, for example if the first sample is a plasma sample then the second sample is a plasma sample. The invention may further comprise repeating the method on a third, and optionally a 4.sup.th, 5.sup.th, 6.sup.th 7.sup.th, 8.sup.th, 9.sup.th and/or 10.sup.th, sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the third, and optionally the 4.sup.th, 5.sup.th, 6.sup.th, 7.sup.th, 8.sup.th, 9.sup.th and/or 10.sup.th, sample comprises circulating free DNA (cfDNA), and comparing the level of prostate cancer fraction in each sample. Preferably, all samples are of the same type as the first sample, for example if the first sample is a plasma sample the all other samples are plasma samples.

[0308] In one preferred embodiment, the method is for monitoring of prostate cancer, wherein the method comprises repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of prostate cancer fraction in each sample.

[0309] In one preferred embodiment, the method is for selecting treatment of prostate cancer, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of prostate cancer fraction in each sample, wherein a new treatment is selected if the level of prostate cancer is increased in the second sample, for example an increase of at least 0.01%.

[0310] In one preferred embodiment, the method is for ascertaining whether treatment of prostate cancer is working, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the treatment is not working if the level of prostate cancer is increased in the second sample, for example an increase of at least 0.01%.

[0311] In one preferred embodiment, the method is for prognostication of prostate cancer, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the prognosis is poor if the level of prostate cancer is increased in the second sample, for example an increase of at least 0.01%. In one preferred embodiment, the method is for prognostication of prostate cancer, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the prognosis is good if the level of prostate cancer is decreased in the second sample, for example a decrease of at least 0.01%. In some instances, a “good” prognosis refers to the likelihood that a subject will likely respond favorably to a drug or set of drugs, leading to a complete or partial remission, or a decrease and/or a stop in the progression of prostate cancer. In some instances, a “good” prognosis refers to the survival of a subject of from at least 1 month to at least 90 years. In some instances, a “good” prognosis refers to the survival of a subject in which the survival of the subject upon treatment is from at least 1 month to at least 90 years.

[0312] In certain preferred embodiments, the method of present invention comprises the additional step of obtaining a biological sample from a subject.

[0313] The methods of the invention can be used with the kits, methods of treatment, therapeutic agents for the treatment of prostate cancer, methods of determining one or more suitable therapeutic agents for the treatment of prostate cancer, methods for determining a treatment regimen, computerized (or computer implemented) methods, computer-assisted methods, computer products and/or computer implemented software described herein. Embodiments and preferred embodiments for the methods of the invention are equally applicable to the kits, methods of treatment, therapeutic agents for the treatment of prostate cancer, methods of determining one or more suitable therapeutic agents for the treatment of prostate cancer, methods for determining a treatment regimen, computerized (or computer implemented) methods, computer-assisted methods, computer products and/or computer implemented software described herein.

Methods of the Invention to Determine Whether a Sample Comprises cfDNA Derived from a Prostate Cancer Subtype

[0314] The present invention also provides a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA, the method comprising: [0315] characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0316] determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0317] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and [0318] a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus, [0319] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0320] calculating a methylation score using the average methylation ratio for each of the genomic regions; [0321] analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0322] Tables 8 is provided below. The genomic locations of Table 8 are locations with reference to hg19.

TABLE-US-00009 Chromosome start end gene chr12 52240301 52240400 n/a chr8 143535751 143535850 n/a chr17 81036151 81036250 n/a chr8 143535801 143535900 n/a chr5 142005201 142005300 FGF1 chr17 81036101 81036200 n/a chr12 52240351 52240450 n/a chr19 47736001 47736100 BBC3 chr10 3480051 3480150 LOC105376360 chr14 101123351 101123450 LINC00523 chr8 144303301 144303400 n/a chr7 95155001 95155100 ASB4 chr8 143535501 143535600 n/a chr15 41219401 41219500 n/a chr15 41219451 41219550 n/a chr7 1251201 1251300 n/a chr8 143535851 143535950 n/a chr2 189191651 189191750 GULP1 chr8 144303251 144303350 n/a chr8 143535601 143535700 n/a chr3 23782851 23782950 n/a chr1 1936451 1936550 n/a chr7 158800951 158801050 LINC00689 chr12 322251 322350 SLC6A12 chr1 15655951 15656050 FHAD1 chr8 143535701 143535800 n/a chr20 36037701 36037800 n/a chr20 36037751 36037850 n/a chr17 7083051 7083150 ASGR1 chr7 5319551 5319650 n/a chr17 7083001 7083100 ASGR1 chr10 131650451 131650550 EBF3 chr1 1936501 1936600 n/a chr19 35818801 35818900 n/a chr10 3479951 3480050 LOC105376360 chr4 1160801 1160900 SPON2 chr19 47735751 47735850 BBC3 chr10 3494301 3494400 LOC105376360 chr17 78982051 78982150 n/a chr10 4331801 4331900 n/a chr1 1920801 1920900 CFAP74 chr9 132482351 132482450 PRRX2 chr8 1923051 1923150 KBTBD11 chr16 1159851 1159950 n/a chr2 189191701 189191800 GULP1 chr1 200707101 200707200 n/a chr20 48124151 48124250 PTGIS chr19 35818851 35818950 n/a chr10 131650701 131650800 EBF3 chr10 3379051 3379150 LOC105376360 chr10 3449001 3449100 LOC105376360 chr12 107297051 107297150 n/a chr19 35981501 35981600 KRTDAP chr13 106063151 106063250 n/a chr5 2207051 2207150 n/a chr8 54164751 54164850 OPRK1 chr3 129326701 129326800 n/a chr1 223435701 223435800 SUSD4 chr2 11294551 11294650 PQLC3 chr17 25798951 25799050 KSR1 chr22 37215901 37216000 PVALB chr11 45392501 45392600 LOC399886 chr11 45392551 45392650 LOC399886 chr17 35277351 35277450 n/a chr9 89410901 89411000 n/a chr9 89410951 89411050 n/a chr8 103572851 103572950 ODF1 chr6 168629801 168629900 n/a chr3 129326651 129326750 n/a chr1 204655151 204655250 LRRN2 chr1 204655201 204655300 LRRN2 chr1 88108801 88108900 n/a chr10 4386801 4386900 n/a chr2 11294501 11294600 PQLC3 chr16 49530551 49530650 ZNF423 chr16 49530601 49530700 ZNF423 chr7 95155051 95155150 ASB4 chr10 73324401 73324500 CDH23 chr5 150538351 150538450 ANXA6 chr7 1388201 1388300 n/a chr3 186170701 186170800 n/a chr8 1923101 1923200 KBTBD11 chr8 54164651 54164750 OPRK1 chr16 1316401 1316500 n/a chr10 4386851 4386950 n/a chr4 1535701 1535800 n/a chr8 144213001 144213100 n/a chr10 131650651 131650750 EBF3 chr10 3480001 3480100 LOC105376360 chr3 64305701 64305800 n/a chr3 64305751 64305850 n/a chr1 1936551 1936650 n/a chr10 3480101 3480200 LOC105376360 chr10 3277051 3277150 n/a chr4 24796601 24796700 SOD3 chr3 46622551 46622650 TDGF1 chr14 104688501 104688600 n/a chr1 55504701 55504800 PCSK9 chr22 37215951 37216050 PVALB chr1 172291651 172291750 DNM3 chr1 2527501 2527600 MMEL1 chr15 27210251 27210350 n/a chr8 54164601 54164700 OPRK1 chr7 3019151 3019250 CARD11 chr11 71010451 71010550 n/a chr19 35981451 35981550 KRTDAP chr16 876151 876250 n/a chr8 1923001 1923100 KBTBD11 chr7 1251251 1251350 n/a chr1 38606051 38606150 n/a chr10 131650501 131650600 EBF3 chr4 140201651 140201750 MGARP chr14 105052601 105052700 C14orf180 chr10 3378851 3378950 LOC105376360 chr14 106095451 106095550 n/a chr12 6933201 6933300 GPR162 chr8 54164801 54164900 OPRK1 chr13 106063101 106063200 n/a chr10 94448551 94448650 n/a chr8 54164701 54164800 OPRK1 chr17 79459401 79459500 n/a chr7 158818151 158818250 LINC00689 chr6 25727351 25727450 HIST1H2AA chr5 1010951 1011050 NKD2 chr1 2424651 2424750 PLCH2 chr3 128724951 128725050 EFCC1 chr12 322951 323050 SLC6A12 chr10 3591201 3591300 LOC105376360 chr10 3591251 3591350 LOC105376360 chr1 2424701 2424800 PLCH2 chr7 1687001 1687100 n/a chr17 27396901 27397000 n/a chr4 7252451 7252550 SORCS2 chr10 134610401 134610500 n/a chr7 1388151 1388250 n/a chr5 2207001 2207100 n/a chr6 37503051 37503150 LOC100505530 chr10 131752851 131752950 EBF3 chr8 143546801 143546900 ADGRB1 chr15 102094651 102094750 n/a chr14 101128351 101128450 LINC00523 chr3 64338501 64338600 n/a chr3 64338551 64338650 n/a chr2 209271151 209271250 PTH2R chr1 15655901 15656000 FHAD1 chr16 29267801 29267900 n/a chr12 107297101 107297200 n/a chr22 43621801 43621900 SCUBE1 chr10 5406551 5406650 UCN3 chr17 79109751 79109850 AATK chr14 105052451 105052550 C14orf180 chr3 55931401 55931500 ERC2 chr3 55931451 55931550 ERC2 chr16 1316351 1316450 n/a chr10 3708501 3708600 LOC105376360 chr16 57317701 57317800 PLLP chr10 118084351 118084450 CCDC172 chr10 3572301 3572400 LOC105376360 chr1 3507101 3507200 MEGF6 chr8 700101 700200 ERICH1-AS1 chr9 6716301 6716400 n/a chr6 112132901 112133000 FYN chr8 143535651 143535750 n/a chr14 103691501 103691600 n/a chr4 1564101 1564200 n/a chr12 322151 322250 SLC6A12 chr12 322201 322300 SLC6A12 chr7 1686751 1686850 n/a chr3 128725001 128725100 EFCC1 chr10 4414951 4415050 n/a chr14 105052551 105052650 C14orf180 chr9 129282651 129282750 n/a chr9 129282701 129282800 n/a chr5 137225301 137225400 PKD2L2 chr1 7569301 7569400 CAMTA1 chr12 44858051 44858150 n/a chr20 43377501 43377600 KCNK15 chr20 43377551 43377650 KCNK15 chr1 1974801 1974900 n/a chr16 89009501 89009600 CBFA2T3 chr3 72704651 72704750 n/a chr14 70037601 70037700 CCDC177 chr6 25727301 25727400 HIST1H2AA chr15 27210301 27210400 n/a chr15 62543151 62543250 n/a chr10 3300501 3300600 n/a chr7 99067201 99067300 n/a chr6 168617401 168617500 n/a chr1 210501351 210501450 HHAT chr5 1207401 1207500 SLC6A19 chr10 131650401 131650500 EBF3 chr17 35277401 35277500 n/a chr5 173097501 173097600 n/a chr5 173097551 173097650 n/a chr17 76522851 76522950 DNAH17 chr4 3288751 3288850 n/a chr19 49528551 49528650 CGB chr19 49528601 49528700 CGB chr10 130844201 130844300 n/a chr1 172291701 172291800 DNM3 chr2 209271101 209271200 PTH2R chr1 6531301 6531400 PLEKHG5 chr22 40051501 40051600 CACNA1I chr16 876201 876300 n/a chr17 25798651 25798750 KSR1 chr17 25798701 25798800 KSR1 chr14 106174351 106174450 n/a chr16 22776051 22776150 MIR548D2 chr14 106174301 106174400 n/a chr2 3697501 3697600 n/a chr16 29267751 29267850 n/a chr7 1459051 1459150 n/a chr9 122734551 122734650 n/a chr10 4386751 4386850 n/a chr6 37527301 37527400 n/a chr17 21278901 21279000 KCNJ12 chr1 3347951 3348050 PRDM16 chr8 1707251 1707350 n/a chr10 3708451 3708550 LOC105376360 chr1 223435751 223435850 SUSD4 chr4 24796551 24796650 SOD3 chr6 45500901 45501000 RUNX2 chr1 38513551 38513650 n/a chr10 135054951 135055050 VENTX chr10 103326701 103326800 n/a chr16 4673901 4674000 MGRN1 chr19 44146901 44147000 n/a chr7 1686701 1686800 n/a chr3 14862901 14863000 FGD5 chr16 1159801 1159900 n/a chr1 210612301 210612400 HHAT chr8 142452401 142452500 MROH5 chr15 99088101 99088200 n/a chr21 43547751 43547850 UMODL1 chr10 130959651 130959750 n/a chr1 1974751 1974850 n/a chr20 61162201 61162300 MIR133A2 chr12 52238951 52239050 n/a chr12 52239001 52239100 n/a chr7 1251151 1251250 n/a chr19 17138801 17138900 n/a chr19 17138851 17138950 n/a chr15 68699651 68699750 ITGA11 chr10 3797401 3797500 n/a chr10 3797451 3797550 n/a chr5 149683251 149683350 ARSI chr5 149683301 149683400 ARSI chr2 159705601 159705700 n/a chr1 2424601 2424700 PLCH2 chr14 103691451 103691550 n/a chr5 1010901 1011000 NKD2 chr12 133178951 133179050 n/a chr12 107297251 107297350 n/a chr12 107297301 107297400 n/a chr22 43827801 43827900 MPPED1 chr11 72974051 72974150 n/a chr10 135054901 135055000 VENTX chr14 101128401 101128500 LINC00523 chr9 132482401 132482500 PRRX2 chr17 60214601 60214700 n/a chr16 57317651 57317750 PLLP chr5 162997901 162998000 n/a chr9 140127301 140127400 SLC34A3 chr17 78982001 78982100 n/a chr10 131650551 131650650 EBF3 chr20 61979401 61979500 CHRNA4 chr14 106095501 106095600 n/a chr3 72704701 72704800 n/a chr1 14220301 14220400 n/a chr5 2207101 2207200 n/a chr9 137660401 137660500 COL5A1 chr11 64739801 64739900 C11orf85 chr7 1329401 1329500 n/a chr13 106063051 106063150 n/a chr4 1535651 1535750 n/a chr17 14206951 14207050 HS3ST3B1 chr16 22776101 22776200 MIR548D2 chr4 6575801 6575900 MAN2B2 chr1 200707051 200707150 n/a chr14 103569401 103569500 EXOC3L4 chr1 7408801 7408900 CAMTA1 chr1 1920751 1920850 CFAP74 chr16 876101 876200 n/a chr16 474251 474350 n/a chr4 3288901 3289000 n/a chr1 3534401 3534500 n/a chr7 4678651 4678750 n/a chr19 36004901 36005000 DMKN chr5 131350101 131350200 n/a chr6 134350851 134350950 SLC2A12 chr9 132383101 132383200 NTMT1 chr10 131744351 131744450 EBF3 chr1 64197451 64197550 n/a chr20 61979301 61979400 CHRNA4 chr20 44934651 44934750 CDH22 chr1 9341951 9342050 n/a chr10 94448501 94448600 n/a chr4 3288851 3288950 n/a chr12 118312351 118312450 KSR2 chr20 21483901 21484000 n/a chr7 1329351 1329450 n/a chr3 185420301 185420400 IGF2BP2 chr3 185420351 185420450 IGF2BP2 chr10 131650751 131650850 EBF3 chr16 14380701 14380800 n/a chr11 57364951 57365050 SERPING1 chr17 25583251 25583350 n/a chr15 62543101 62543200 n/a chr19 47735951 47736050 BBC3 chr14 104639751 104639850 KIF26A chr5 1856051 1856150 LOC101929034 chr20 44934701 44934800 CDH22 chr10 134610451 134610550 n/a chr21 47398651 47398750 n/a chr10 3343101 3343200 n/a chr7 3019101 3019200 CARD11 chr21 44494701 44494800 CBS chr16 89009451 89009550 CBFA2T3 chr17 79109801 79109900 AATK chr9 139587851 139587950 n/a chr1 2527451 2527550 MMEL1 chr21 46973201 46973300 n/a chr2 202753101 202753200 CDK15 chr1 157140751 157140850 n/a chr5 2207151 2207250 n/a chr1 1097301 1097400 n/a chr17 63134151 63134250 RGS9 chr9 136500151 136500250 n/a chr3 194097051 194097150 n/a chr3 129326751 129326850 n/a chr7 2728901 2729000 AMZ1 chr5 137225001 137225100 PKD2L2 chr15 102094601 102094700 n/a chr10 4230551 4230650 n/a chr5 2205701 2205800 n/a chr16 14380651 14380750 n/a chr1 25298701 25298800 n/a chr11 1102501 1102600 MUC2 chr11 14994301 14994400 CALCA chr11 14994351 14994450 CALCA chr14 106438051 106438150 ADAM6 chr22 43829751 43829850 MPPED1 chr8 22018451 22018550 SFTPC chr21 34351051 34351150 n/a chr10 3544651 3544750 LOC105376360 chr11 60482451 60482550 MS4A8 chr11 2190101 2190200 TH chr20 4705201 4705300 PRND chr17 1811301 1811400 n/a chr5 141993201 141993300 FGF1 chr14 23290301 23290400 n/a chr17 60214651 60214750 n/a chr4 140201601 140201700 MGARP chr20 61979451 61979550 CHRNA4 chr11 64739751 64739850 C11orf85 chr16 1111151 1111250 n/a chr4 3288801 3288900 n/a chr1 38513601 38513700 n/a chr7 73466051 73466150 ELN chr16 24697401 24697500 n/a chr16 85201801 85201900 n/a chr9 137859601 137859700 n/a chr1 1936751 1936850 n/a chr1 22975601 22975700 n/a chr1 22975651 22975750 n/a chr5 1207451 1207550 SLC6A19 chr4 3865101 3865200 n/a chr21 46799851 46799950 n/a chr3 13058851 13058950 IQSEC1 chr1 7130401 7130500 CAMTA1 chr14 104852051 104852150 n/a chr5 1923501 1923600 n/a chr16 2863801 2863900 n/a chr11 120592101 120592200 GRIK4 chr11 120592151 120592250 GRIK4 chr1 17022551 17022650 ESPNP chr11 128796401 128796500 n/a chr15 75019301 75019400 n/a chr2 3697451 3697550 n/a chr11 120590051 120590150 GRIK4 chr11 120590101 120590200 GRIK4 chr12 49366151 49366250 WNT10B chr10 131650351 131650450 EBF3 chr8 144472051 144472150 n/a chr5 493301 493400 SLC9A3 chr1 234039901 234040000 SLC35F3 chr4 1564151 1564250 n/a chr14 103691301 103691400 n/a chr8 142452451 142452550 MROH5 chr7 1329451 1329550 n/a chr22 43805251 43805350 n/a chr22 43805301 43805400 n/a chr22 37771301 37771400 ELFN2 chr3 194090601 194090700 LRRC15 chr8 125249851 125249950 LOC101927588 chr7 2728851 2728950 AMZ1 chr7 1388251 1388350 n/a chr6 168629951 168630050 n/a chr19 36004951 36005050 DMKN chr11 63996751 63996850 DNAJC4 chr20 4705251 4705350 PRND chr3 196515551 196515650 PAK2 chr3 196515601 196515700 PAK2 chr17 65527651 65527750 PITPNC1 chr20 23969801 23969900 GGTLC1 chr7 23471801 23471900 IGF2BP3 chr6 134350801 134350900 SLC2A12 chr2 121279851 121279950 n/a chr4 184244751 184244850 n/a chr12 124607901 124608000 ZNF664 - FAM101A chr15 68699601 68699700 ITGA11 chr2 242151551 242151650 ANO7 chr5 2205751 2205850 n/a chr5 172924801 172924900 n/a chr5 137225351 137225450 PKD2L2 chr5 493251 493350 SLC9A3 chr8 144367251 144367350 n/a chr19 554951 555050 n/a chr12 1675901 1676000 FBXL14 chr5 74532301 74532400 ANKRD31 chr15 78186501 78186600 n/a chr16 24697451 24697550 n/a chr9 137859551 137859650 n/a chr1 21913451 21913550 n/a chr4 1537251 1537350 n/a chr11 69706801 69706900 n/a chr22 37771251 37771350 ELFN2 chr10 3526751 3526850 LOC105376360 chr2 219487501 219487600 PLCD4 chr2 219487551 219487650 PLCD4 chr16 876251 876350 n/a chr14 104639801 104639900 KIF26A chr8 700151 700250 ERICH1-AS1 chr6 18990551 18990650 n/a chr20 1164951 1165050 TMEM74B chr4 26493401 26493500 n/a chr6 168617451 168617550 n/a chr1 7408851 7408950 CAMTA1 chr10 131650301 131650400 EBF3 chr22 37771201 37771300 ELFN2 chr16 474301 474400 n/a chr17 66288801 66288900 ARSG chr21 41027851 41027950 B3GALT5 chr10 131706751 131706850 EBF3 chr7 1748001 1748100 ELFN1 chr12 52238601 52238700 n/a chr12 52238651 52238750 n/a chr7 158828251 158828350 VIPR2 chr5 137225401 137225500 PKD2L2 chr21 43547801 43547900 UMODL1 chr1 57718951 57719050 DAB1 chr1 57719001 57719100 DAB1 chr15 99974801 99974900 n/a chr14 104688551 104688650 n/a chr16 14380751 14380850 n/a chr21 44494751 44494850 CBS chr9 89411001 89411100 n/a chr19 14313651 14313750 ADGRL1 chr17 74237601 74237700 n/a chr19 3821051 3821150 MIR1268A chr3 66139601 66139700 SLC25A26 chr10 4482651 4482750 n/a chr10 3602701 3602800 LOC105376360 chr10 3602751 3602850 LOC105376360 chr15 29825301 29825400 FAM189A1 chr20 61979351 61979450 CHRNA4 chr12 322901 323000 SLC6A12 chr7 73466001 73466100 ELN chr17 79109701 79109800 AATK chr10 5407001 5407100 UCN3 chr11 67462801 67462900 n/a chr7 45188151 45188250 n/a chr1 87994651 87994750 n/a chr11 64780701 64780800 ARL2 chr7 73790751 73790850 CLIP2 chr5 532951 533050 n/a chr2 242797901 242798000 PDCD1 chr15 23894801 23894900 n/a chr15 23894751 23894850 n/a chr5 2206751 2206850 n/a chr7 1407501 1407600 n/a chr20 23970051 23970150 GGTLC1 chr19 554851 554950 n/a chr5 2205951 2206050 n/a chr15 101807351 101807450 n/a chr4 1160751 1160850 SPON2 chr14 104768501 104768600 n/a chr9 6716351 6716450 n/a chr2 66743751 66743850 MEIS1 chr17 25798401 25798500 KSR1 chr11 102216851 102216950 BIRC2 chr10 4358501 4358600 n/a chr12 116008051 116008150 n/a chr14 70476801 70476900 SMOC1 chr9 139587901 139588000 n/a chr7 131831551 131831650 PLXNA4 chr5 141993251 141993350 FGF1 chr3 194097101 194097200 n/a chr16 88963701 88963800 CBFA2T3 chr15 29037851 29037950 PDCD6IPP2 chr6 134350901 134351000 SLC2A12 chr8 143546851 143546950 ADGRB1 chr9 129387201 129387300 LMX1B chr14 104617951 104618050 KIF26A chr4 3288401 3288500 n/a chr8 81963301 81963400 PAG1 chr8 81963351 81963450 PAG1 chr3 126080301 126080400 n/a chr9 136567001 136567100 SARDH chr7 1329001 1329100 n/a chr6 37014501 37014600 n/a chr6 37014551 37014650 n/a chr10 3544601 3544700 LOC105376360 chr4 3776451 3776550 n/a chr11 72980801 72980900 P2RY6 chr14 76877951 76878050 ESRRB chr11 120044501 120044600 n/a chr2 159705351 159705450 n/a chr12 86230751 86230850 RASSF9 chr12 86230801 86230900 RASSF9 chr14 94406501 94406600 ASB2 chr14 106438101 106438200 ADAM6 chr7 29186301 29186400 CPVL chr16 29242051 29242150 n/a chr4 187071151 187071250 FAM149A chr19 40032701 40032800 n/a chr17 77536201 77536300 n/a chr3 97542301 97542400 CRYBG3 chr6 25761601 25761700 SLC17A4 chr1 9342001 9342100 n/a chr17 60828151 60828250 Mar-10 chr19 5455301 5455400 ZNRF4 chr7 44279601 44279700 CAMK2B chr14 106174251 106174350 n/a chr1 156831151 156831250 NTRK1 chr5 150538301 150538400 ANXA6 chr2 239695751 239695850 n/a chr21 46816651 46816750 n/a chr5 162997951 162998050 n/a chr10 3457351 3457450 LOC105376360 chr1 7539101 7539200 CAMTA1 chr7 1137351 1137450 C7orf50 chr5 180597551 180597650 n/a chr12 52240401 52240500 n/a chr2 71099251 71099350 n/a chr11 62100751 62100850 n/a chr14 101928001 101928100 n/a chr14 94463751 94463850 LINC00521 chr14 94463801 94463900 LINC00521 chr14 101123451 101123550 LINC00523 chr7 3488801 3488900 SDK1 chr5 132944101 132944200 FSTL4 chr10 131034801 131034900 n/a chr1 38517201 38517300 n/a chr20 62004751 62004850 n/a chr5 1217751 1217850 SLC6A19 chr15 60919451 60919550 RORA-AS1 chr16 88963651 88963750 CBFA2T3 chr2 159705401 159705500 n/a chr9 135033201 135033300 n/a chr17 7082951 7083050 ASGR1 chr19 18902651 18902750 COMP chr19 18902701 18902800 COMP chr1 6531151 6531250 PLEKHG5 chr1 1084501 1084600 n/a chr1 1084551 1084650 n/a chr10 3708401 3708500 LOC105376360 chr10 131691251 131691350 EBF3 chr5 2205901 2206000 n/a chr13 113807851 113807950 n/a chr7 127881551 127881650 LEP chr5 2335601 2335700 n/a chr21 42219751 42219850 DSCAM chr10 130959601 130959700 n/a chr10 4697351 4697450 LINC00704 chr10 4697401 4697500 LINC00705 chr7 1407301 1407400 n/a chr5 137224951 137225050 PKD2L2 chr1 226756401 226756500 C1orf95 chr1 226756451 226756550 C1orf95 chr1 200143101 200143200 NR5A2 chr11 67219501 67219600 CABP4 chr6 168629851 168629950 n/a chr17 14207001 14207100 HS3ST3B1 chr4 74847801 74847900 PF4 chr11 67619801 67619900 n/a chr9 138171701 138171800 n/a chr2 54560551 54560650 C2orf73 chr1 15655851 15655950 FHAD1 chr22 32750851 32750950 RFPL3 chr1 156828651 156828750 INSRR chr14 103691351 103691450 n/a chr2 27938101 27938200 n/a chr10 118084301 118084400 CCDC172 chr16 85198551 85198650 n/a chr22 37499451 37499550 TMPRSS6 chr3 139258301 139258400 RBP1 chr22 50457151 50457250 n/a chr11 75222401 75222500 GDPD5 chr6 169351351 169351450 n/a chr5 532901 533000 n/a chr14 93154751 93154850 RIN3 chr14 104623601 104623700 KIF26A chr11 63996801 63996900 DNAJC4 chr6 112132951 112133050 FYN chr4 3691301 3691400 n/a chr7 4870201 4870300 RADIL chr15 66543901 66544000 MEGF11 chr14 105105101 105105200 n/a chr7 564251 564350 HRAT92 chr1 14220251 14220350 n/a chr16 1316151 1316250 n/a chr1 21044901 21045000 KIF17 chr3 169540251 169540350 LRRIQ4 chr1 64197401 64197500 n/a chr1 231761601 231761700 DISC1 chr3 54353651 54353750 CACNA2D3 chr10 3500151 3500250 LOC105376360 chr1 23521351 23521450 HTR1D chr9 139925801 139925900 C9orf139 chr8 1644901 1645000 DLGAP2 chr8 1644951 1645050 DLGAP2 chr5 150538401 150538500 ANXA6 chr19 47735701 47735800 BBC3 chr1 22889251 22889350 EPHA8 chr14 106229551 106229650 n/a chr22 43621751 43621850 SCUBE1 chr14 89881701 89881800 FOXN3 chr20 30618851 30618950 CCM2L chr3 14595751 14595850 n/a chr16 84336251 84336350 WFDC1 chr17 26795251 26795350 n/a chr14 104770801 104770900 n/a chr11 102216901 102217000 BIRC2 chr9 122734601 122734700 n/a chr3 169540101 169540200 LRRIQ4 chr16 14380601 14380700 n/a chr21 46420501 46420600 LINC00162 chr11 68781901 68782000 MRGPRF-AS1 chr16 22776001 22776100 MIR548D2 chr7 30718001 30718100 CRHR2 chr5 137225251 137225350 PKD2L2 chr4 3690751 3690850 n/a chr10 4194451 4194550 n/a chr1 205913951 205914050 n/a chr5 114514651 114514750 TRIM36 chr17 75789551 75789650 n/a chr9 33448251 33448350 AQP3 chr11 4843051 4843150 OR51F2 chr17 41739251 41739350 MEOX1 chr16 1295551 1295650 n/a chr2 159705551 159705650 n/a chr4 7652101 7652200 SORCS2 chr10 134662251 134662350 CFAP46 chr7 1329301 1329400 n/a chr12 47219951 47220050 SLC38A4 chr10 13039651 13039750 CCDC3 chr1 226791451 226791550 C1orf95 chr8 143261951 143262050 n/a chr17 81036051 81036150 n/a chr10 28971201 28971300 BAMBI chr17 34996051 34996150 n/a chr14 105052501 105052600 C14orf180 chr7 44279651 44279750 CAMK2B chr7 3018401 3018500 CARD11 chr10 131650601 131650700 EBF3 chr17 1811351 1811450 n/a chr21 47399551 47399650 n/a chr2 121279801 121279900 n/a chr10 3568801 3568900 LOC105376360 chr19 15585451 15585550 PGLYRP2 chr8 42009151 42009250 n/a chr11 2293051 2293150 ASCL2 chr10 3250701 3250800 n/a chr2 86037151 86037250 n/a chr1 1936601 1936700 n/a chr7 3018601 3018700 CARD11 chr17 78456401 78456500 n/a chr10 134303901 134304000 n/a chr8 144303201 144303300 n/a chr13 28562501 28562600 URAD chr13 28562551 28562650 URAD chr9 132482451 132482550 PRRX2 chr1 48360401 48360500 TRABD2B chr1 48360451 48360550 TRABD2B chr14 100625001 100625100 DEGS2 chr5 180597601 180597700 n/a chr14 70348401 70348500 SMOC1 chr14 70348451 70348550 SMOC1 chr11 62100701 62100800 n/a chr9 136567051 136567150 SARDH chr14 37075451 37075550 n/a chr10 4194501 4194600 n/a chr21 46799901 46800000 n/a chr16 57916851 57916950 CNGB1 chr10 3343001 3343100 n/a chr10 1602501 1602600 ADARB2 chr1 226791351 226791450 C1orf95 chr6 41435651 41435750 n/a chr2 26788701 26788800 C2orf70 chr20 62004701 62004800 n/a chr7 24328551 24328650 NPY chr19 1505901 1506000 ADAMTSL5 chr9 34588501 34588600 CNTFR chr10 3343051 3343150 n/a chr9 132383301 132383400 NTMT1 chr1 205913901 205914000 n/a chr2 242797851 242797950 PDCD1 chr9 132383351 132383450 NTMT1 chr4 8158251 8158350 ABLIM2 chr10 3281051 3281150 n/a chr15 62358751 62358850 C2CD4A chr15 33437351 33437450 FMN1 chr15 78114851 78114950 n/a chr7 99987501 99987600 PILRA chr4 1504551 1504650 n/a chr5 140710351 140710450 PCDHGA1 chr6 33561351 33561450 LINC00336 chr6 33561401 33561500 LINC00336 chr3 169540301 169540400 LRRIQ4 chr8 143570901 143571000 ADGRB1 chr14 101123301 101123400 LINC00523 chr15 99088051 99088150 n/a chr19 36195351 36195450 ZBTB32 chr16 67336051 67336150 KCTD19 chr1 63798301 63798400 n/a chr1 63798351 63798450 n/a chr7 36013301 36013400 n/a chr5 2204551 2204650 n/a chr3 139258251 139258350 RBP1 chr11 67462851 67462950 n/a chr19 36195401 36195500 ZBTB32 chr17 1202251 1202350 TUSC5 chr16 281351 281450 n/a chr15 75019351 75019450 n/a chr10 4446051 4446150 LINC00703 chr17 60214551 60214650 n/a chr1 200175551 200175650 n/a chr1 154843201 154843300 KCNN3 chr7 1747951 1748050 ELFN1 chr16 29242101 29242200 n/a chr8 143868151 143868250 LY6D chr4 3752251 3752350 n/a chr6 130992701 130992800 n/a chr7 1684601 1684700 n/a chr11 2210201 2210300 n/a chr17 79109601 79109700 AATK chr14 103569351 103569450 EXOC3L4 chr8 136510551 136510650 KHDRBS3 chr7 1358201 1358300 n/a chr10 3373301 3373400 LOC105376360 chr6 46455901 46456000 RCAN2 chr6 46455951 46456050 RCAN2 chr5 73969151 73969250 HEXB chr1 203525601 203525700 n/a chr22 37771351 37771450 ELFN2 chr19 17571601 17571700 NXNL1 chr2 202753251 202753350 CDK15 chr13 50703451 50703550 DLEU1 chr3 185866551 185866650 DGKG chr12 116008101 116008200 n/a chr11 62100801 62100900 n/a chr4 3690901 3691000 n/a chr9 140127251 140127350 SLC34A3 chr7 3018451 3018550 CARD11 chr7 99987601 99987700 PILRA chr5 2537751 2537850 n/a chr16 30034801 30034900 C16orf92 chr22 37500701 37500800 TMPRSS6 chr9 132315801 132315900 n/a chr10 2978801 2978900 n/a chr1 61408051 61408150 NFIA-AS2 chr11 62100651 62100750 n/a chr17 66288751 66288850 ARSG chr7 2959101 2959200 CARD11 chr22 25160851 25160950 PIWIL3 chr20 23970101 23970200 GGTLC1 chr4 1537551 1537650 n/a chr2 27938151 27938250 n/a chr1 226791401 226791500 C1orf95 chr14 104768451 104768550 n/a chr10 3250751 3250850 n/a chr1 218537401 218537500 TGFB2 chr1 229480101 229480200 n/a chr7 30029851 30029950 SCRN1 chr7 30029901 30030000 SCRN1 chr16 2863851 2863950 n/a chr3 64225051 64225150 n/a chr3 64225101 64225200 n/a chr22 25160451 25160550 PIWIL3 chr14 65289701 65289800 SPTB chr7 4843901 4844000 RADIL chr16 90115051 90115150 URAHP chr16 90115101 90115200 URAHP chr19 3030301 3030400 TLE2 chr4 3677601 3677700 LOC100133461 chr5 140710501 140710600 PCDHGA1 chr2 242797751 242797850 PDCD1 chr14 93154701 93154800 RIN3 chr15 29611951 29612050 FAM189A1 chr14 106208351 106208450 n/a chr11 120561251 120561350 GRIK4 chr17 27396951 27397050 n/a chr6 17988951 17989050 n/a chr19 45720101 45720200 EXOC3L2 chr10 4296351 4296450 n/a chr4 187729101 187729200 n/a chr4 187729151 187729250 n/a chr1 94270151 94270250 BCAR3 chr3 127173651 127173750 n/a chr16 84336301 84336400 WFDC1 chr7 89747951 89748050 DPY19L2P4 chr2 239048601 239048700 KLHL30 chr5 1010851 1010950 NKD2 chr1 87994701 87994800 n/a chr19 51538151 51538250 KLK12 chr17 41739201 41739300 MEOX1 chr10 112834851 112834950 n/a chr19 41062001 41062100 SPTBN4 chr16 281401 281500 n/a chr7 99987551 99987650 PILRA chr10 3313151 3313250 n/a chr20 61371501 61371600 NTSR1 chr22 26877601 26877700 HPS4 chr22 26877651 26877750 HPS4 chr22 18508301 18508400 MICAL3 chr16 3142651 3142750 ZSCAN10 chr6 170585851 170585950 LOC285804 chr9 122800851 122800950 n/a chr12 299701 299800 SLC6A12 chr15 33437301 33437400 FMN1 chr10 4378551 4378650 n/a chr10 4378601 4378700 n/a chr12 111137051 111137150 n/a chr7 2728751 2728850 AMZ1 chr11 72980851 72980950 P2RY6 chr19 3030251 3030350 TLE2 chr15 29825351 29825450 FAM189A1 chr1 210612251 210612350 HHAT chr16 88880801 88880900 GALNS chr15 60919401 60919500 RORA chr7 1137301 1137400 C7orf50 chr5 180597651 180597750 n/a chr2 42077601 42077700 n/a chr10 134610351 134610450 n/a chr14 104852001 104852100 n/a chr8 144854651 144854750 n/a chr10 94448451 94448550 n/a chr1 15685251 15685350 FHAD1 chr13 28563651 28563750 URAD chr6 25727151 25727250 HIST1H2AA chr17 75848751 75848850 n/a chr5 137225101 137225200 PKD2L2 chr19 56914751 56914850 ZNF583 chr7 23471751 23471850 IGF2BP3 chr14 104627851 104627950 KIF26A chr1 4794901 4795000 AJAP1 chr19 46651201 46651300 IGFL2 chr17 21278851 21278950 KCNJ12 chr12 58736301 58736400 n/a chr5 73969201 73969300 HEXB chr17 77644501 77644600 n/a chr12 322601 322700 SLC6A12 chr2 189191601 189191700 GULP1 chr1 14220201 14220300 n/a chr6 168629901 168630000 n/a chr1 861751 861850 SAMD11 chr7 3018351 3018450 CARD11 chr7 2728801 2728900 AMZ1 chr12 116944101 116944200 n/a chr7 89747901 89748000 STEAP2-AS1 chr6 168630001 168630100 n/a chr16 29242001 29242100 n/a chr7 1329051 1329150 n/a chr5 170743851 170743950 n/a chr1 65362451 65362550 JAK1 chr7 1407351 1407450 n/a chr10 4358551 4358650 n/a chr11 92806401 92806500 n/a chr14 101123501 101123600 LINC00523 chr8 914451 914550 ERICH1-AS1 chr7 1407251 1407350 n/a chr2 113379951 113380050 n/a chr14 100631751 100631850 n/a chr12 44858001 44858100 n/a chr14 104865801 104865900 n/a chr8 94508451 94508550 LINC00535 chr6 25727251 25727350 HIST1H2AA chr19 4566501 4566600 n/a chr21 44724701 44724800 n/a chr7 158800601 158800700 LINC00689 chr9 138109251 138109350 n/a chr11 69706751 69706850 n/a chr6 25727001 25727100 HIST1H2BA chr9 137731801 137731900 COL5A1 chr19 56914801 56914900 ZNF583 chr14 23290351 23290450 n/a chr5 137225151 137225250 PKD2L2 chr10 3300451 3300550 n/a chr10 130959701 130959800 n/a chr17 27347151 27347250 n/a chr4 1535601 1535700 n/a chr10 34496301 34496400 PARD3 chr3 14595851 14595950 n/a chr7 3018301 3018400 CARD11 chr6 168533451 168533550 n/a chr16 1198651 1198750 n/a chr11 2293201 2293300 n/a chr14 105044951 105045050 C14orf180 chr11 2293251 2293350 n/a chr10 131357151 131357250 MGMT chr5 497501 497600 SLC9A3 chr2 242797801 242797900 PDCD1 chr1 1920701 1920800 CFAP74 chr14 106320501 106320600 n/a chr14 105045001 105045100 C14orf180 chr3 185788701 185788800 ETV5 chr14 94451401 94451500 n/a chr11 118042701 118042800 SCN2B chr7 1266101 1266200 n/a chr1 2527401 2527500 MMEL1 chr6 17988901 17989000 n/a chr5 10653251 10653350 ANKRD33B chr5 10653301 10653400 ANKRD33B chr16 1198601 1198700 n/a chr5 140710451 140710550 PCDHGA1 chr14 104617901 104618000 KIF26A chr15 100016301 100016400 n/a chr1 33391451 33391550 n/a chr5 137225201 137225300 PKD2L2 chr3 97542151 97542250 CRYBG3 chr6 156954501 156954600 n/a chr11 2293301 2293400 n/a chr3 13058901 13059000 IQSEC1 chr17 74581301 74581400 ST6GALNAC2 chr12 107297151 107297250 n/a chr1 7602001 7602100 CAMTA1 chr14 104768351 104768450 n/a chr1 121260701 121260800 EMBP1 chr7 1686651 1686750 n/a chr14 100624951 100625050 DEGS2 chr7 72788001 72788100 n/a chr5 2205801 2205900 n/a chr17 74581401 74581500 ST6GALNAC2 chr10 134610301 134610400 n/a chr19 554901 555000 n/a chr21 46816601 46816700 n/a chr10 4230501 4230600 n/a chr7 1251301 1251400 n/a chr22 19744001 19744100 TBX1 chr8 143545151 143545250 ADGRB1 chr19 45003801 45003900 ZNF180 chr7 2959151 2959250 CARD11 chr3 169540351 169540450 LRRIQ4 chr2 209271201 209271300 PTH2R chr13 31620351 31620450 n/a chr1 200003301 200003400 NR5A2 chr11 67462751 67462850 n/a chr20 47278501 47278600 PREX1 chr22 37499801 37499900 TMPRSS6 chr7 73465951 73466050 ELN chr19 17571551 17571650 NXNL1 chr1 1936801 1936900 n/a chr11 2206101 2206200 n/a chr14 100631801 100631900 n/a chr2 75136551 75136650 LINC01291 chr10 12543351 12543450 CAMK1D chr4 3677551 3677650 LOC100133461 chr22 19744051 19744150 TBX1 chr14 106208401 106208500 n/a chr14 105044901 105045000 n/a chr22 37500651 37500750 TMPRSS6 chr6 168630051 168630150 n/a chr4 1537601 1537700 n/a chr7 104897151 104897250 SRPK2 chr14 106174201 106174300 n/a chr21 42219701 42219800 DSCAM chr10 79270701 79270800 KCNMA1 chr14 104623551 104623650 KIF26A chr1 7601951 7602050 CAMTA1 chr2 121279901 121280000 n/a chr7 120967801 120967900 WNT16 chr7 120967851 120967950 WNT16 chr7 65970101 65970200 n/a chr16 474201 474300 n/a chr1 1957751 1957850 GABRD chr1 3534351 3534450 n/a chr5 173738051 173738150 n/a chr11 120764501 120764600 LOC101929227 chr9 122800901 122801000 n/a chr9 129387151 129387250 LMX1B chr6 18990501 18990600 n/a chr3 72704601 72704700 n/a chr10* 26502051 26502150 n/a chr5* 111090051 111090150 NREP chr5* 111090101 111090200 NREP chr10* 26502101 26502200 n/a chr15* 67841351 67841450 MAP2K5 chr15* 67841401 67841500 MAP2K5 chr8* 25902201 25902300 EBF2 All regions including, having, or within a genomic location of Table 8 are hypomethylated regions except for the 7 locations indicated with a *, which are hypermethylated regions In Table 8, where the gene indicated is “n/a” this means that the genomic location defined in the table is a non-coding region of DNA or not within the location of a known gene.

[0323] The prostate cancer subtype is one that has an aggressive clinical course and/or androgen receptor (AR) copy number gain, for example an androgen-insensitive prostate cancer subtype. The prostate cancer subtype may be a subtype (i.e. one having an aggressive clinical course and/or androgen receptor (AR) copy number gain) of acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer. For example, it may be a subtype (i.e. one having an aggressive clinical course and/or androgen receptor (AR) copy number gain) of acinar adenocarcinoma prostate cancer or ductal adenocarcinoma prostate cancer. Alternatively, or additionally, the prostate cancer may be castration sensitive prostate cancer or castration resistant prostate cancer. Alternatively, or additionally, the prostate cancer may be metastatic prostate cancer, or it may be non-metastatic prostate cancer. In certain embodiments, it may be metastatic prostate cancer. In certain embodiments, the prostate cancer may be metastatic castration resistant prostate cancer or non-metastatic castration resistant prostate cancer. For example, it may be metastatic castration resistant prostate cancer.

[0324] The method is especially suitable for the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of metastatic prostate cancer and/or castration resistant prostate cancer, and particularly prostate cancers subtypes that have an aggressive clinical course and androgen receptor (AR) copy number gain, for example an androgen-insensitive prostate cancer subtype.

[0325] The sample is a sample that comprises cfDNA. The sample may suitably be a blood sample, a plasma sample, or a urine sample. Preferably, the sample is a blood sample or a plasma sample. More preferably, the sample is a plasma sample.

[0326] The method may further comprise isolating the cfDNA from the sample. cfDNA can be isolated from the sample using a variety of techniques known in the art. For example, DNA (e.g., cfDNA) can be isolated by a column-based approach and/or a bead-based approach. In some embodiments, DNA (e.g., cfDNA) is isolated by means of a column-based approach, for example using a commercially available kit such as QIAamp circulating nucleic acid kit (Qiagen qiagen.com/ch/products/discovery-and-translational-research/dna-rna-purification/dna-purification/cell-free-dna/qiaamp-circulating-nucleic-acid-kit/#orderinginformation). In some embodiments, DNA (e.g., cfDNA) is isolated by means of a bead-based approach, for example an automated cf-DNA extraction system using a commercially available kit such as Maxwell RSC ccfDNA Plasma Kit (Promega (https://www.promega.co.uk/resources/protocols/technical-manuals/101/maxwell-rsc-ccfdna-plasma-kit-protocol/)).

[0327] The isolated cfDNA may be amplified before analysis. Thus the method may further comprise amplification of the isolated cfDNA. Amplification techniques are known to those of ordinary skill in the art and include, but are not limited to, cloning, polymerase chain reaction (PCR), polymerase chain reaction of specific alleles (PASA), polymerase chain ligation, nested polymerase chain reaction, and so forth.

[0328] The method comprises characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule. The methylome sequence of a cfDNA molecule may be characterised by using methylation aware sequencing, by genome sequencing followed by methylation profiling, or by targeted approaches that capture specific DNA sequences (for example using DNA probes). Examples of methylation aware sequencing include bisulfite sequencing, bisulfite-free methylation-aware sequencing, methylation arrays (for example methylation microarrays), enzymatic methylation sequencing, methylation-sensitive restriction enzyme digestion, methylation-specific PCR, methylation aware PCR based assays, methylation-dependent DNA precipitation, methylated DNA binding proteins/peptides, single molecule sequences without sodium bisulfite treatment. In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using bisulfite sequencing, methylation microarrays, enzymatic methylation sequencing, bisulfite-free methylation-aware sequencing, or methylation aware PCR based assays.

[0329] Examples of targeted approaches that capture specific DNA sequences (for example using DNA probes) include cell-free methylated DNA immunoprecipitation and high-throughput sequencing (cfMeDIP-seq), methylation-dependent DNA precipitation, and methylated DNA binding proteins/peptides.

[0330] Bisulfite sequencing may comprise massive parallel sequencing with bisulfite conversion, for example treating the DNA molecule with sodium bisulfite and performing sequencing of the treated DNA molecule. Methylation assay sequencing may comprise treating the DNA molecule with sodium bisulfite, whole genome amplification, and hybridisation to a methylation-specific probe or a non-methylation probe, for example attached to a bead or chip.

[0331] Enzymatic methylation sequencing may comprise enzymatic treatment of the DNA molecule to convert methylated cytosine sites, followed by sequencing of the treated DNA. For example enzymatic methylation sequencing may comprise enzymatic treatment of the DNA molecule to convert methylated cytosine sites into a form protected from deamination, followed by deamination to convert unprotected cytosine to uracils, and sequencing of the treated DNA. An example of an enzymatic methylation sequencing kit includes NEBNext® Enzymatic Methyl-seq Kit (https://www.neb.com/products/e7120-nebnext-enzymatic-methyl-seq-kit#).

[0332] Examples of methylation aware PCR based assays include digital droplet PCR and qPCR (quantitative PCR).

[0333] An example of bisulfite-free methylation-aware sequencing is Oxford Nanopore seqeuencing (Oxford Nanopore Technologies, https://nanoporetech.com/))

[0334] In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using whole genome bisulfite sequencing, for example low pass whole genome bisulfite sequencing. In another embodiment, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using reduced representation bisulfite treatments. In certain embodiments, the methylome sequence of a plurality of cfDNA molecules in the sample is characterised using methylation arrays, for example methylation microarrays, such as a Illumina Methylation Assay.

[0335] A variety of genome sequencing procedures are known in the art and may be used to practice the methods disclosed herein. For example, Sanger sequencing, Polony sequencing, 454 pyrosequencing, Combinatorial probe anchor synthesis, SOLiD sequencing, Ion Torrent semiconductor sequencing, DNA nanoball sequencing, Heliscope single molecule sequencing, Single molecule real time (SMRT) sequencing, Nanopore DNA sequencing, Microfluidic Sanger sequencing and Illumina dye sequencing.

[0336] A plurality of cfDNA molecules may be, for example, at least 100, at least 1000, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). Preferably, a plurality of cfDNA molecules may be, for example, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). More preferably, a plurality of cfDNA molecules may be, for example, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9).

[0337] The method may further comprise aligning the methylome sequences with a reference genome for the subject, for example by aligning the methylome sequences with hg38, hg19, hg18, hg17 or hg16. The alignment can, for example, be carried out using a variety of techniques known in the art. For example, a DNA sequence alignment tool, (e.g., BSMAP (PMID: 19635165), Bismark (PMID: 21493656), gemBS (PMID: 30137223), Arioc (PMID: 29554207), BS-Seeker2 (PMID: 24206606), MethylCoder (PMID: 21724594) or BatMeth2 (PMID: 30669962)) can be used to align the reads to the reference genome (for example hg38, hg19, hg18, hg17 or hg16).

[0338] The genomic location assigned to each methylome sequence in the alignment is based on the reference genome adopted. The genomic locations listed in Tables 1, 1b, 2 to 9 disclosed herein correspond to reference genome hg19. The corresponding locations in a different reference genome can be found using public available tools known in the art. An example of these tools is LiftOver (http://genome.ucsc.edu/).

[0339] In certain embodiments, the method comprises removing duplications of reads of the same DNA molecule (i.e. duplications of reads of the same cfDNA molecule). In this step, sequence reads having exactly the same sequence and start and end base pairs (for example the same unclipped alignment start and unclipped alignment end of the sequence) are removed, as they are likely to be duplicate sequence reads of the same sequence (i.e. duplicate of reads of the same cfDNA molecule). For example, PCR duplications can be removed as part of the aligning step, such as using Picard tools v2.1.0 (http://broadinstitute.github.io/picard).

[0340] The method comprises determining the average methylation ratio at 10 or more of the genomic regions for which the average methylation ratio has been determined, each genomic region being selected from the group consisting of: [0341] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and [0342] a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence.

[0343] In one preferred embodiment, the method comprises determining the average methylation ratio at 10 or more of the genomic regions for which the average methylation ratio has been determined, each genomic region being selected from the group consisting of:

a 100 to 200 bp region comprising or having a genomic location defined in Table 9, and
a 2 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence.

[0344] In certain embodiments, each genomic region for which the average methylation ratio has been determined is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 50, 100, 1000, 10,000 characterized methylome sequences. Preferably each genomic region is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 50, 100, or 1000 characterized methylome sequences. In certain preferred embodiments, each genomic region is covered by at least one sequence read of at least 10 characterized methylome sequences, for example at least one sequence read of at least 10, at least 15, at least 20, at least 25, at least 50, at least 100, or at least 1000 characterized methylome sequences.

[0345] In certain embodiments, each genomic region for which the average methylation ratio has been determined is covered by at least 2 sequence reads, for example at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads. Preferably, each genomic region is covered by at least 5 sequence reads, for example at least 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads. More preferably, each genomic region is covered by at least 10 sequence reads, for example at least 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads.

[0346] In embodiments wherein each genomic region for which the average methylation ratio has been determined is covered by at least 2 sequence reads (for example at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, or 10,000 sequence reads) preferably each sequence read or the majority of the sequence reads (for example at least 50%, 60%, 70%, 80% or 90% of the sequence reads) are from different characterized methylome sequences. More preferably, each sequence read or at least 60%, 70%, 80% or 90% of the sequence reads are from different characterized methylome sequences.

[0347] In certain embodiments the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, 400 or more genomic regions, or 500 or more genomic regions. Each genomic region may be selected from the group consisting of: [0348] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and [0349] a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus.

[0350] The genomic regions are preferably each different from each other. In certain preferred embodiments, the method comprises determining the average methylation ratio at 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, 400 or more genomic regions, or 500 or more genomic regions. Each genomic region may be selected from the group consisting of: [0351] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and [0352] a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus.

[0353] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, 400 or more genomic regions, or 500 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0354] In certain embodiments the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, or 150 genomic regions. Each genomic region may be selected from the group consisting of: [0355] a 100 to 200 bp region comprising or having a genomic location defined in Table 9, and [0356] a 2 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus.

[0357] The genomic regions are preferably each different from each other.

[0358] In certain embodiments, each genomic region is selected from the group consisting of: [0359] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus.

[0360] More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Table 8, and 10 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 8, and 50 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 8, and 80 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Table 8.

[0361] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, 150 or more genomic regions, 200 or more genomic regions, 300 or more genomic regions, or 400 or more genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0362] In certain embodiments, each genomic region is selected from the group consisting of: [0363] a 100 to 200 bp region comprising or having a genomic location defined in Table 9, and a 2 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus.

[0364] More suitably, each genomic region is selected from the group consisting of: a 100 to 150 bp region comprising or having a genomic location defined in Table 9, and 10 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 9, and 50 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus. More suitably, each genomic region is selected from the group consisting of: a 100 to 120 bp region comprising or having a genomic location defined in Table 9, and 80 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus. For example, each genomic region is selected from a 100 bp region having a genomic location defined in Table 9.

[0365] In such embodiments, preferably the method comprises determining the average methylation ratio at 12 or more genomic regions, for example 15 or more genomic regions, 20 or more genomic regions, 25 or more genomic regions, 30 or more genomic regions, 50 or more genomic regions, 75 or more genomic regions, 100 or more genomic regions, 125 or more genomic regions, or 150 genomic regions. For example, the method comprises determining the average methylation ratio at 100 or more genomic regions.

[0366] In certain preferred embodiments, determining the average methylation ratio for a genomic region comprises calculating the sum of the methylation ratios of all CpGs within the genomic region and dividing the sum by the number of CpGs within the genomic region. In such embodiments, the average methylation ratio may also be referred to as the mean methylation ratio. For the avoidance of doubt, if a genomic region has only one CpG locus, the average methylation ratio for the genomic region is the same as the methylation ratio for the single CpG locus in the genomic region.

[0367] The method of the present invention comprises calculating a methylation score using the average methylation ratio for each genomic region for which the average methylation ratio has been determined.

[0368] In certain embodiments, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0369] determining the median or the mean of the average methylation ratios for all genomic regions (i.e. all genomic regions for which an average methylation ratio has been determined in the method); or [0370] determining the median or the mean of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median or the mean of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or [0371] comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0372] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0373] determining the median of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or [0374] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or [0375] comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0376] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises: [0377] determining the median of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or [0378] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score.

[0379] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0380] determining the median of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median of the average methylation ratios for second group of genomic regions to obtain a second methylation score.

[0381] In embodiments wherein calculating a methylation score using the average methylation ratio for each genomic region comprises determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for a second group of genomic regions to obtain a second methylation score, the first group of genomic regions are all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 8), and the second group of genomic regions are all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method, i.e. selected from those comprising, having or within a genomic location defined in Table 8).

[0382] In one preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0383] determining the median of the average methylation ratios for all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method) to obtain a first methylation score and determining the median of the average methylation ratios for all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method to obtain a second methylation score.

[0384] In one embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0385] determining the median of the average methylation ratios for all of the hypermethylated genomic regions (i.e. all hypermethylated genomic regions for which an average methylation ratio has been determined in the method) to obtain a first methylation score.

[0386] In one especially preferred embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises [0387] determining the median of the average methylation ratios for all of the hypomethylated genomic regions (i.e. all hypomethylated genomic regions for which an average methylation ratio has been determined in the method) to obtain a second methylation score.

[0388] In one embodiment, calculating a methylation score using the average methylation ratio for each genomic region comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region. In such embodiments, preferably the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by: [0389] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0390] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0391] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient, for example a prostate cancer patient with a known subtype; [0392] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0393] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0394] a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype); [0395] a characterized methylome sequence of a white blood cell; [0396] a characterized methylome sequence of a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0397] a characterized methylome sequence of a cancerous prostate cell, for example a cancerous prostate cell of a known subtype; and/or [0398] a characterized methylome sequence of a non-cancerous prostate cell.

[0399] In one preferred embodiment, the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by [0400] a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype).

[0401] In one preferred embodiment, the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by [0402] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient, for example a prostate cancer patient with a known subtype; [0403] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0404] a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype); [0405] a characterized methylome sequence of a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; and/or [0406] a characterized methylome sequence of a cancerous prostate cell, for example a cancerous prostate cell of a known subtype.

[0407] The method of the present invention comprises analyzing the methylation ratio scores to determine whether the sample comprises cfDNA derived from a prostate cancer subtype and/or determine the level of cfDNA in the sample that is derived from a prostate cancer subtype. For example, no level (for example no detectable level) of cfDNA derived from a prostate cancer subtype in the cfDNA sample may be determined. Alternatively, a level of cfDNA derived from a prostate cancer subtype in the cfDNA sample may be determined. The minimum percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA sample that may be determined may be 0.01% of cfDNA derived from a prostate cancer subtype in the cfDNA sample. In certain embodiments, the minimum percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA sample that may be determined may be 0.02%, 0.03%, 0.04%, 0.06%, 0.07%, 0.08%, 0.05%, 0.09%, 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, 0.9%, 1%, 2%, 3% 4%, 5%, 10%, 15%, 20%, 30%, 40%, 50% of cfDNA derived from a prostate cancer subtype in the cfDNA sample. For example, the minimum percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA sample that may be determined may be 0.01%, 0.05%, 0.1% or 0.5%. Preferably, the minimum percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA is 0.01%.

[0408] The method comprises analyzing the methylation score to determine the level of cfDNA derived from a prostate cancer subtype in the cfDNA sample.

[0409] If level of cfDNA derived from a prostate cancer subtype in the cfDNA sample is determined, the subject can be classed as having the subtype. As such, analyzing the methylation score to determine whether there is a level of cfDNA derived from a prostate cancer subtype in the cfDNA sample may also be referred as analyzing the methylation score to determine whether a subject has a prostate cancer subtype.

[0410] Preferably, analyzing the methylation score to determine the level of cfDNA derived from a prostate cancer subtype in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores. For example, the method may comprise comparing the methylation score to one reference methylation scores. In certain embodiments, the method comprises comparing the methylation score to two or more reference methylation scores, for example 2, 3, 4, 5, 6, 8, 9, 10, 12, 15, 20, 30, 50, 100, 200, 300, 400, 500 or 1000 reference methylation scores. In certain embodiments, the method comprises comparing the methylation score to 5 or more reference methylation scores, for example 10 or more, 15 or more, 20 or more, 30, or more 50, or more 100, or more 200, or more 300, or more 400, or more 500 or 1000 or more reference methylation scores.

[0411] In embodiments wherein the method comprises comparing the methylation score to two or more reference methylation scores, the reference methylation scores may come from different types of reference samples and/or reference methylomes (for example a cfDNA sample from a healthy subject and a cancer cell line sample) and/or the same type of reference samples or reference methylomes but from different sources (for example, two or more cfDNA samples each from a different healthy subject).

[0412] A reference methylation score is a methylation score calculated for the same genomic regions (for example, calculated using the average methylation ratio for the same genomic regions) in a reference sample or reference methylome. A reference sample or reference methylome may be selected from the group consisting of: [0413] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0414] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0415] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient, for example a prostate cancer patient with a known subtype; [0416] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0417] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0418] a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype); [0419] a characterized methylome sequence of a white blood cell; [0420] a characterized methylome sequence of a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0421] a characterized methylome sequence of a cancerous prostate cell, for example a cancerous prostate cell of a known subtype; and/or [0422] a characterized methylome sequence of a non-cancerous prostate cell.

[0423] A reference sample or reference methylome may be one that can be used to represent a sample having no cfDNA derived from the prostate cancer subtype (for example an undetectable level of cfDNA in the prostate cancer subtype in the cfDNA sample), for example a reference sample or reference methylome selected from one or more of the following [0424] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0425] a sample of white blood cells from a subject, for example the subject or a healthy subject; and/or [0426] a characterized methylome sequence of a white blood cell.

[0427] A reference sample or reference methylome may be one that can be used to represent a sample having 100% cfDNA derived from the prostate cancer subtype, for example a reference sample or reference methylome selected from one or more of the following [0428] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient, for example a prostate cancer patient with a known subtype; [0429] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0430] a characterized methylome sequence of a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0431] a characterized methylome sequence of a cancerous prostate cell, for example a cancerous prostate cell of a known subtype; and/or

[0432] A reference sample or reference methylome may be one that can be used to represent a sample having 10 to 90% cfDNA derived from a prostate cancer subtype, for example one or more cfDNA samples from different subjects having prostate cancer known to have the prostate cancer subtype, wherein the level of cfDNA derived from the prostate cancer subtype in each cfDNA sample from the different subjects is/are known. A level of cfDNA derived from the prostate cancer subtype in each cfDNA sample can be determined by looking at genomic markers.

[0433] Preferably, analyzing the methylation score to determine the level of cfDNA derived from the prostate cancer subtype in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores that can be used to represent a sample having 100% cfDNA derived from the prostate cancer subtype, and can be used to represent a sample having 0% cfDNA derived from the prostate cancer subtype, and optionally can be used to represent a sample having 10-90% cfDNA derived from the prostate cancer subtype. For example, analyzing the methylation score to determine the level of cfDNA derived from a prostate cancer subtype in the cfDNA sample comprises:

comparing the methylation score to one or more reference methylation scores for a reference sample or reference methylome selected from the group consisting of: [0434] a cfDNA sample from a healthy subject, for example a healthy age-matched subject, [0435] a sample of white blood cells from a subject, for example the subject or a healthy subject, and/or [0436] a characterized methylome sequence of a white blood cell;
and
comparing the methylation score to one or more reference methylation scores for a reference sample or reference methylome selected from the group consisting of: [0437] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient, for example a prostate cancer patient with a known subtype; [0438] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0439] a characterized methylome sequence of a prostate cancer cell line, for example a prostate cancer cell line of a known subtype; [0440] a characterized methylome sequence of a cancerous prostate cell, for example a cancerous prostate cell of a known subtype
and optionally comparing the methylation score to one or more reference methylation scores for one or more cfDNA samples from different subjects having prostate cancer, wherein the level of cfDNA derived from the prostate cancer subtype in each cfDNA sample from the different subjects is/are known.

[0441] Preferably, the reference methylation score for a reference sample or reference methylome that a methylation ratio score methylation ratio score is compared to is calculated in the same way as the methylation score for the sample obtained from the subject (i.e. the sample that the method of the invention is being carried out in respect of). For example, if the methylation ratio for the selected genomic regions of the sample obtained from the subject is calculated by determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score, the reference methylation score for a reference sample or reference methylome is calculated by determining the median (or the mean) of the average methylation ratios for the same first group of genomic regions to obtain a first reference methylation score and/or determining the median (or the mean) of the average methylation ratios for the same second group of genomic regions to obtain a second reference methylation score.

[0442] Or, for example, if the methylation ratio for the selected genomic regions of the sample obtained from the subject is calculated by determining the median (or the mean) of the average methylation ratios for all genomic regions, the reference methylation score for a reference sample or reference methylome is calculated by determining the median (or the mean) of the average methylation ratios for the same genomic regions.

[0443] In embodiments wherein the method comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region, analyzing the methylation ratio scores to determine the level of cfDNA derived from the prostate cancer subtype in the cfDNA sample may comprise determining how many methylation ratio scores are indicative of the prostate cancer subtype.

[0444] In certain embodiments, analyzing the methylation score to determine the level of cfDNA derived from the prostate cancer subtype in the cfDNA sample comprises using a mathematical model, such as a linear regression model or another linear model (for example, a general linear model, a heteroscedastic model, a generalised linear model, or a hierarchical linear model).

[0445] In certain embodiments, analyzing the methylation score to determine the level of level of cfDNA derived from the prostate cancer subtype in the cfDNA sample comprises using a mathematical model that compares the methylation score for the sample to reference methylation scores that can be used to represent a sample having 100% cfDNA derived from the prostate cancer subtype in the cfDNA, and can be used to represent a sample having 0% cfDNA derived from the prostate cancer subtype in the cfDNA, and optionally can be used to represent a sample having 10-90% cfDNA derived from the prostate cancer subtype in the cfDNA. For example, the method comprises using mathematical model that compares the methylation score for the sample to reference methylation scores for a cfDNA sample from a healthy subject, for example a healthy age-matched subject (0% cfDNA derived from the prostate cancer subtype in the cfDNA) and/or a characterized methylome sequence of a white blood cell (0% cfDNA derived from the prostate cancer subtype in the cfDNA) and/or a sample of white blood cells from a subject, for example the subject or a healthy subject, (0% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or a characterized methylome sequence of a prostate cancer cell line (100% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or a prostate cancer biopsy sample from a prostate cancer patient (100% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or one or more cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of cfDNA derived from the prostate cancer subtype in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype (10-90% cfDNA derived from the prostate cancer subtype in the cfDNA sample).

[0446] In one embodiment, the method comprises using mathematical model that compares the methylation score for the sample to reference methylation scores for a cfDNA sample from a healthy subject, for example a healthy age-matched subject (0% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or a characterized methylome sequence of a prostate cancer cell line (100% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or a prostate cancer biopsy sample from a prostate cancer patient (100% cfDNA derived from the prostate cancer subtype in the cfDNA sample) and/or one or more cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein the level of cfDNA derived from the prostate cancer subtype in the cfDNA in each cfDNA sample from the different subjects is known, and preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype in the cfDNA sample (10-90% cfDNA derived from the prostate cancer subtype in the cfDNA sample).

[0447] The method may further comprise measuring the level of prostate-specific antigen (PSA) in a sample of blood from the subject. It may also comprise determining if the subject has an abnormal level of PSA in the blood (for example a level of PSA in the blood of at least 4.0 ng/mL). An abnormal level of PSA in the blood may be, for example, a level of PSA in the blood of at least 4.0 ng/mL). A normal level of PSA in the blood may, for example, be a level of PSA in the blood of 4.0 ng/mL or less.

[0448] In one preferred embodiment, the method is for screening, monitoring, and/or prognostication of prostate cancer, wherein prostate cancer with a poor prognosis is predicted when a level of cfDNA derived from the prostate cancer subtype in the cfDNA sample is determined, for example a detectable level of cfDNA derived from the prostate cancer subtype in the sample, for example a percentage level of cfDNA derived from the prostate cancer subtype in the sample of at least 0.01%. For example, a prostate cancer with a poor prognosis is predicted when at least 0.01% cfDNA derived from the prostate cancer subtype in the sample is determined, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the sample is determined.

[0449] In some instances, a “poor” prognosis refers to a low likelihood that a subject will likely respond favorably to a drug or set of drugs, is in complete or partial remission, or there is a decrease and/or a stop in the progression of prostate cancer. In some instances, a “poor” prognosis refers to a survival of a subject that is expected to be from less than 5 years to less than 1 month (for example less than 3 years to less than 1 month, or less than 3 years to less than 6 months). In some instances, a “poor” prognosis refers to a survival of a subject in which the survival of the subject upon treatment is expected to be from less than 5 years to less than 1 month.

[0450] In one preferred embodiment, the method is for detection of prostate cancer, wherein the prostate cancer subtype is detected when a level of cfDNA derived from a prostate cancer subtype in the cfDNA sample is determined, for example a detectable level of cfDNA derived from the prostate cancer subtype in the cfDNA sample, for example a percentage level of cfDNA derived from the prostate cancer subtype in the cfDNA sample of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the cfDNA sample.

[0451] In one preferred embodiment, the method is for screening, monitoring, and/or prognostication of prostate cancer, wherein prostate cancer with a poor prognosis is predicted when a level of cfDNA derived from the prostate cancer subtype in the cfDNA sample is determined, for example a detectable level of prostate cancer, for example a percentage level of cfDNA derived from the prostate cancer subtype in the cfDNA sample of at least 0.01%, for example at least 0.01% cfDNA derived from the prostate cancer subtype in the cfDNA sample, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the cfDNA sample.

[0452] In one preferred embodiment, the method is for detecting, screening and/or prognostication of metastatic prostate cancer, wherein metastatic prostate cancer is predicted when a level of cfDNA derived from the prostate cancer subtype in the cfDNA sample is determined, for example a detectable level of cfDNA derived from the prostate cancer subtype in the cfDNA sample, for example a percentage level of cfDNA derived from the prostate cancer subtype in the cfDNA sample of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the cfDNA sample.

[0453] In one preferred embodiment, the method is for selecting treatment of prostate cancer or ascertaining whether treatment is working in prostate cancer, wherein a new treatment is selected when a level of cfDNA derived from the prostate cancer subtype in the cfDNA sample is determined, for example a detectable level of cfDNA derived from the prostate cancer subtype in the cfDNA sample, for example a percentage level of cfDNA derived from the prostate cancer subtype in the cfDNA sample of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the cfDNA sample.

[0454] In one preferred embodiment, the method is for ascertaining whether treatment of prostate cancer is working, wherein it is determined that the treatment is not working when a level of prostate cancer is determined, for example a detectable level of cfDNA derived from the prostate cancer subtype in the cfDNA sample, for example a percentage level of cfDNA derived from the prostate cancer subtype in the cfDNA sample of at least 0.01%, or for example, at least 0.02%, at least 0.03%, at least 0.04%, at least 0.05%, at least 0.1%, at least 0.5% or at least 1% cfDNA derived from the prostate cancer subtype in the cfDNA sample.

[0455] The method may further comprising repeating the method on second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of cfDNA derived from the prostate cancer subtype in each sample. Preferably, the second sample is of the same type as the first sample, for example if the first sample is a plasma sample then the second sample is a plasma sample. The invention may further comprise repeating the method on a third, and optionally a 4.sup.th, 5.sup.th, 6.sup.th, 7.sup.th, 8th, 9.sup.th and/or 10.sup.th, sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the third, and optionally the 4.sup.th, 5.sup.th, 6.sup.th, 7.sup.th, 8.sup.th, 9.sup.th and/or 10.sup.th, sample comprises cfDNA, and comparing the level of cfDNA derived from the prostate cancer subtype in each sample. Preferably, all samples are of the same type as the first sample, for example if the first sample is a plasma sample the all other samples are plasma samples.

[0456] In one preferred embodiment, the method is for monitoring of prostate cancer, wherein the method comprises repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of cfDNA derived from the prostate cancer subtype in each cfDNA sample.

[0457] In one preferred embodiment, the method is for selecting treatment of prostate cancer, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of cfDNA derived from the prostate cancer subtype in each cfDNA sample, wherein a new treatment is selected if the level of prostate cancer is increased in the second sample, for example an increase of at least 0.01%.

[0458] In one preferred embodiment, the method is for ascertaining whether treatment of prostate cancer is working, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the treatment is not working if the level of cfDNA derived from the prostate cancer subtype is increased in the second sample, for example an increase of at least 0.01%.

[0459] In one preferred embodiment, the method is for prognostication of prostate cancer, comprising repeating the method on second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the prognosis is poor if the level of cfDNA derived from the prostate cancer subtype is increased in the second sample, for example an increase of at least 0.01%. In one preferred embodiment, the method is for prognostication of prostate cancer, comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, wherein it is determined that the prognosis is good if the level of cfDNA derived from the prostate cancer subtype is decreased in the second sample, for example a decrease of at least 0.01%. In some instances, a “good” prognosis refers to the likelihood that a subject will likely respond favorably to a drug or set of drugs, leading to a complete or partial remission, or a decrease and/or a stop in the progression of prostate cancer. In some instances, a “good” prognosis refers to the survival of a subject of from at least 1 month to at least 90 years. In some instances, a “good” prognosis refers to the survival of a subject in which the survival of the subject upon treatment is from at least 1 month to at least 90 years.

[0460] In certain preferred embodiments, the method of present invention comprises the additional step of obtaining a biological sample from a subject.

[0461] The methods can be used with the kits, methods of treatment, therapeutic agents for the treatment of prostate cancer, methods of determining one or more suitable therapeutic agents for the treatment of prostate cancer, methods for determining a treatment regimen, computerized (or computer implemented) methods, computer-assisted methods, computer products and/or computer implemented software described herein. Embodiments and preferred embodiments for the methods are equally applicable to the kits, methods of treatment, therapeutic agents for the treatment of prostate cancer, methods of determining one or more suitable therapeutic agents for the treatment of prostate cancer, methods for determining a treatment regimen, computerized (or computer implemented) methods, computer-assisted methods, computer products and/or computer implemented software described herein.

Kits

[0462] A further aspect, the invention provides an in-vitro diagnostic kit for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA. Preferably, the kits of the invention comprise one or more reagents for detecting the presence or absence of at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4.

[0463] In certain embodiments, the kit comprises DNA sampling reagents and, preferably, methylome analysis reagents, such as bisulfate reagents. In certain embodiments, the kit comprises DNA amplification agents, for example primers for amplification of specific DNA molecules, for example for amplification of at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4.

[0464] In one preferred embodiment, the kit comprises instructions for use. In certain embodiments, the kit comprises instructions for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample using the kit. For example the kit comprises instructions for use which define how to determine the level of prostate cancer fraction in a sample comprising cfDNA from a subject, for example by following a method of the invention defined herein.

[0465] In one preferred embodiment, the kit comprises a computer product or a computer-executable software for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample using the kit. In certain embodiments, the computer product comprises a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform a method of the invention. In certain embodiments, the computer-executable software comprises software for performing a method of the invention.

[0466] In certain embodiments the kit comprises of one or more containers and may also include sampling equipment, for example, bottles, bags (such as intravenous fluid bags), vials, syringes, and test tubes. Other components may include needles, diluents, wash reagents and buffers. Usefully, the kit may include at least one container comprising a pharmaceutically-acceptable buffer, such as phosphate-buffered saline, Ringer's solution and dextrose solution.

[0467] If a reagent is for detecting the presence or absence of a DNA molecule having a DNA sequence corresponding to all of a genomic location defined in Tables 1 to 4, the reagent is able to detect the presence of a DNA sequence having or comprising a genomic location defined in Tables 1 to 4. For example, the reagent is able to detect the presence of the a DNA sequence having a genomic location defined in Tables 1 to 4 or comprising a genomic location defined in Tables 1 to 4 and having a sequence length of 101 to 200 bp, for example having a sequence length of 101 to 180, a sequence length of 101 to 150, a sequence length of 101 to 140, a sequence length of 101 to 130, a sequence length of 101 to 120, or a sequence length of 101 to 110 bp.

[0468] If a reagent is for detecting the presence or absence of a DNA molecule having a DNA sequence corresponding to a part of a genomic location defined in Tables 1 to 4, the reagent is able to detect the presence of a DNA sequence comprising at least a 10 bp continuous sequence within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus. Preferably, if a reagent is for detecting the presence or absence of a DNA molecule having a DNA sequence corresponding to a part of a genomic location defined in Tables 1 to 4, the reagent is able to detect the presence of a DNA sequence comprising at least a 15 bp continuous sequence within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, for example at least a 20, 25, 30, 35, 40, 45, 50, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 or 99 bp continuous sequence within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus. In certain preferred embodiments, if a reagent is for detecting the presence or absence of a DNA molecule having a DNA sequence corresponding to a part of a genomic location defined in Tables 1 to 4, the reagent is able to detect the presence of a DNA sequence comprising (or consisting of) a 20, 25, 30, 35, 40, 45, 50, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 or 99 bp continuous sequence within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0469] In certain embodiments, the kit comprises one or more reagents for detecting the presence or absence of at least 15 DNA molecules. For example, the kit comprises one or more reagents for detecting the presence or absence of 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules.

[0470] In certain embodiments, the kit comprises one or more reagents for detecting the presence or absence of at least 50 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 75 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 100 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 150 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 250 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 500 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 500, 600, 700, 800, 900 or 1000 DNA molecules), at least 700 DNA molecules or at least 900 DNA molecules (for example, the kit comprises one or more reagents for detecting the presence or absence of 900 or 1000 DNA molecules).

[0471] In certain preferred embodiments, the genomic location is a location defined in Tables 1 and 2. In certain embodiments, the genomic location is a location defined in Tables 3 and 4. In certain embodiments, the genomic location is a location defined in Tables 1 and 3. In certain embodiments, the genomic location is a location defined in Tables 2 and 4.

[0472] In certain preferred embodiments, the genomic location is a location defined in Table 5. In certain preferred embodiments, the genomic location is a location defined in Table 6. In certain preferred embodiments, the genomic location is a location defined in Table 7.

[0473] In certain embodiments, the kit comprises oligonucleotides for specifically hybridizing to at least a section of the at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4. An oligonucleotide for specifically hybridizing to at least a section of a DNA molecules may be for hybridizing to at least a 10 bp section, at least a 12 bp section, at least a 14 bp section, at least a 15 bp section, at least a 18 bp section, at least a 20 bp section of a DNA molecule, at least a 25 bp section of a DNA molecule, at least a 30 bp section of a DNA molecule or at least a 40 bp section of a DNA molecule. In certain embodiments, an oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may be for hybridizing to a 10 bp section, 12 bp section, 14 bp section, 15 bp section, 18 bp section, 20 bp section, 25 bp section or 30 bp section.

[0474] An oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of at least 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90 or 95 bp. An oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may comprise not more than 100, 90, 80, or 70 bp. An oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90 or 95 bp. Preferably, an oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of 15, 18, 19, 20, 21, 22, 23, 24, 25, 30, 40, 50, 60 or 70 bp. In certain embodiments, an oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of 20 to 90 bp, for example 30 to 80 bp, 50 to 80 bp. In certain embodiments, an oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of 55 to 95 bp. In certain embodiments, an oligonucleotide for specifically hybridizing to at least a section of a DNA molecule may have a sequence of 60 to 80 bp, for example a sequence of 70 bp.

[0475] In certain embodiments, the kit comprises oligonucleotides for specifically hybridizing to at least a section of at least 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 200, 250, 300, 400, 500, 600, 700, 800, or 900 DNA molecules corresponding to a genomic region having or comprising a genomic location defined in Tables 1 to 4. In certain embodiments, the kit comprises oligonucleotides for specifically hybridizing to at least a section of 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 200, 250, 300, 400, 500, 600, 700, 800, 900, or 1000 DNA molecules corresponding to a genomic region having a genomic location defined in Tables 1 to 4.

[0476] In the kits of the invention comprising oligonucleotides, preferably at least one of the oligonucleotides for specifically hybridizing to at least a section of the DNA molecules is an amplification primer. Even more preferably, each oligonucleotide for specifically hybridizing to at least a section of the DNA molecules is an amplification primer.

Method of Treatment and Uses of Therapeutic Agents for the Treatment of a Subject Having Prostate Cancer

[0477] As the methods of the invention of the present invention are for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer, a method of the invention may be used in a method of treatment of a subject having prostate cancer and/or used with a therapeutic agent for use in the treatment of a subject having prostate cancer.

[0478] A therapeutic agent for the treatment of prostate cancer for use in the methods of treatment and uses of the present invention, as well as in the methods, kits, and other aspects of the present invention, is selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, a chemotherapy agent and a radionuclide agent.

[0479] A hormonal agent for the treatment of prostate cancer is selected from the group consisting of LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), estrogens and steroids (for example prednisone or dexamethasone).

[0480] A targeted agent for the treatment of prostate cancer is selected from the group consisting of poly(ADP-ribose) polymerase (PARP) inhibitors (for example olaparib, rucaparib, niraparib or talazoparib), epidermal growth factor receptor (EGFR) inhibitors (for example gefitinib, erlotinib, afatinib, brigatinib, icotinib, cetuximab, osimertinib, adavosertib, or lapatinib), and tyrosine kinase inhibitors (for example imatinib, gefitinib, erlotinib, or sunitinib).

[0481] A biologic agent for the treatment of prostate cancer is selected from the group consisting of monoclonal antibodies (for example pertuzumab, trastuzumab or solitomab), hormones (for example a hormonal agent selected from LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), and estrogens), interferons (for example interferons-α, -β, -γ), and interleukin-based products (for example interleukin-2).

[0482] An immunotherapy agent for the treatment of prostate cancer is selected from the group consisting of cancer vaccines (for example sipuleucel-T), T-cell therapies, monoclonal antibody therapies, immune checkpoint therapies (for example a PD-1 inhibitor (e.g. pembrolizumab, nivolumab, cemiplimab, or spartalizumab), PD-L1 inhibitors (e.g. atezolizumab, avelumab or durvalumab), or a CTLA-4 (e.g. ipilimumab)), and non-specific immunotherapies (for example interferons or inerleukins).

[0483] A chemotherapy agent for the treatment of prostate cancer is selected from the group consisting selected from docetaxel, cabazitaxel, and c-Met inhibitors (for example cabozantinib).

[0484] A radionuclide agent for the treatment of prostate cancer is selected from Radium223 and PSMA-labelled radionuclide (for example .sup.225Ac-Labeled PSMA-617 or .sup.177Lu-Labeled PSMA-617).

[0485] A therapeutic agent for the treatment of prostate cancer may be administered in amounts indicated in the Physicians' Desk Reference (PDR) or as otherwise determined by one of ordinary skill in the art.

[0486] In certain preferred embodiments, a therapeutic agent for the treatment of prostate cancer for use in the methods of treatment and uses of the present invention, as well as in the methods, kits, and other aspects of the present invention, is a hormonal agent and optionally a chemotherapy agent and/or optionally a further hormonal agent and/or optionally a targeted agent and/or optionally a radionuclide agent and/or an immunotherapy agent. For example, a hormonal agent selected from a LHRH agonist (for example leuprolide, goserelin, triptorelin, or histrelin) and a LHRH antagonist (for example degarelix), and optionally docetaxel and/or optionally a PARP inhibitor (for example olaparib, rucaparib, niraparib or talazoparib). Or, for example, a LHRH agonist (for example leuprolide, goserelin, triptorelin, or histrelin) or a LHRH antagonist (for example degarelix), and optionally a chemotherapy agent (for example docetaxel, cabazitaxel, carboplatin) and/or optionally a further hormonal treatment (for example enzalutamide, abiraterone, darolutamide) and/or optionally a radionuclide agent (Radium223 or PSMA-labelled radionuclide) and/or optionally a PARP inhibitor (for example olaparib, rucaparib, niraparib or talazoparib) and/or an immunotherapy agent (for example nivolumab, pembroluzimab, ipilumimab, durvalumab).

[0487] In embodiments wherein the prostate cancer is castration sensitive prostate cancer, preferably the therapeutic agent is a LHRH agonist (for example leuprolide, goserelin, triptorelin, or histrelin) or a LHRH antagonist (for example degarelix) and optionally a chemotherapy agent (for example docetaxel, cabazitaxel, carboplatin) and/or optionally a further hormonal treatment (for example enzalutamide, abiraterone, darolutamide) and/or optionally a radionuclide agent (Radium223 or PSMA-labelled radionuclide (for example .sup.225Ac-Labeled PSMA-617 or .sup.177Lu-Labeled PSMA-617)) and/or optionally a PARP inhibitor (for example olaparib, rucaparib, niraparib or talazoparib) and/or immunotherapy (for example nivolumab, pembroluzimab, ipilumimab, durvalumab).

[0488] In embodiments wherein the prostate cancer is castration resistant prostate cancer, preferably the therapeutic agent for the treatment of prostate cancer is a LHRH agonist (for example leuprolide, goserelin, triptorelin, or histrelin) or a LHRH antagonist (for example degarelix), and optionally a chemotherapy agent (for example docetaxel, cabazitaxel, carboplatin) and/or optionally a further hormonal treatment (for example enzalutamide, abiraterone, darolutamide) and/or optionally a radionuclide agent (Radium223 or a PSMA-labelled radionuclide (for example .sup.225Ac-Labeled PSMA-617 or .sup.177Lu-Labeled PSMA-617)) and/or optionally a PARP inhibitor (for example olaparib, rucaparib, niraparib or talazoparib) and/or immunotherapy agent (for example nivolumab, pembroluzimab, ipilumimab, durvalumab).

[0489] A non-therapeutic treatment for the treatment of prostate cancer is selected from surgery and radiotherapy. A surgical treatment of prostate cancer is selected from the group consisting of radical prostatectomy, a trans-urethral resection of the prostate, and an orchidectomy. A radiotherapy treatment of prostate cancer is selected from external beam localized radiotherapy of the prostate, external beam radiotherapy of metastatic sites.

[0490] In certain embodiments, methods of treatment of the present invention comprise treating the subject using a therapeutic agent for the treatment of prostate cancer, surgery, and/or radiotherapy. In certain embodiments, methods of treatment of the present invention comprise administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer, and/or radiotherapy, and/or performing surgery. In certain embodiments, methods of treatment of the present invention comprise starting, ceasing or altering treatment with a therapeutic agent, or initiating a non-therapeutic treatment (e.g., surgery or radiation).

[0491] The present invention provides a method for treating prostate cancer in a subject comprising a method defined herein (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) and further comprising treating the subject using a therapeutic agent for the treatment of prostate cancer, surgery, and/or radiotherapy.

[0492] The present invention also provides a method for treating prostate cancer in a subject comprising a method defined herein (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) and further comprising administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer, and/or radiotherapy, and/or performing surgery.

[0493] A method of treatment of the present invention is performed before and/or after a method of the invention defined herein (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein).

[0494] Preferably, a method for treating prostate cancer of the present invention comprises administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer surgery, and/or radiotherapy after a method of the invention defined herein, for example after the subject has been determined to have a level of prostate cancer fraction, or determined to have cfDNA derived from a prostate cancer subtype, based on a method as described herein. In another preferred embodiment, a method for treating prostate cancer of the present invention comprises administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer after a method of the invention defined herein, for example after the subject has been determined to have a level of prostate cancer fraction, or determined to have cfDNA derived from a prostate cancer subtype, based on a method as described herein.

[0495] In one embodiment, a method for treating prostate cancer of the present invention comprises administering a therapeutic agent for the treatment of prostate cancer to the subject for at least 1 week, 2 weeks, 3 weeks, 4 weeks, 1 month, 2 months, 3 months, 6 months, 9 months, 12 months, 24 months or 36 months. A therapeutic agent for the treatment of prostate cancer may be administered, for example, daily, every second day, twice per week, weekly or monthly.

[0496] In one embodiment, a method for treating prostate cancer of the present invention comprises treating a subject using a therapeutic agent for the treatment of prostate cancer for at least 1 week, 2 weeks, 3 weeks, 4 weeks, 1 month, 2 months, 3 months, 6 months, 9 months, 12 months, 24 months or 36 months.

[0497] A therapeutic agent for the treatment of prostate cancer may be administered in amounts and at frequencies indicated in the Physicians' Desk Reference (PDR) or as otherwise determined by one of ordinary skill in the art.

[0498] In one preferred embodiment, a method for treating prostate cancer of the present invention comprises performing the method of the invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) before treating the subject, and subsequently repeating the method of the invention, for example at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months or at least 36 months after starting or finishing the treatment, for example after administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer, and/or radiotherapy, and/or performing surgery.

[0499] In another preferred embodiment, a method for treating prostate cancer of the present invention comprises performing the method (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) before treating the subject, and subsequently repeating the method, for example at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months or at least 36 months after performing the first method of the invention.

[0500] In embodiments comprising repeating the method (for example, repeating the method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein), the method may be repeated once, or it may be repeated multiple times, for examples 2, 3, 4, 5, 6 or more times.

[0501] In embodiments comprising repeating the method (for example, repeating the method for detecting, screening, monitoring, staging, classification, selecting treatment, ascertaining whether treatment is working, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein), after the subsequent method(s) is performed, the method may further comprise continuing to treat the subject with the therapeutic agent for the treatment of prostate cancer if the level of prostate cancer tumour fraction is the same or substantially the same in the initial and subsequent method(s) or lower in the subsequent method(s) than in the initial method.

[0502] In embodiments comprising repeating the method (for example, repeating the method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein), after the subsequent method(s) is performed, the method may further comprise:

ceasing or altering (e.g. changing the dose or frequency of the dosing) treatment with the therapeutic agent for the treatment of prostate cancer; and/or
initiating treatment with a second or further therapeutic agent for the treatment of prostate cancer; and/or
initiating a non-therapeutic agent treatment (e.g., surgery or radiation),
if the level of prostate cancer tumour fraction is substantially the same in the initial and subsequent method or higher in the subsequent method than in the initial method; or
if the sample comprises cfDNA derived from a prostate cancer subtype and/or the sample comprises a level of cfDNA derived from a prostate cancer subtype that is substantially the same in the initial and subsequent method or higher in the subsequent method than in the initial method.

[0503] The invention further provides a method of treating a subject in need of treatment with a therapeutic agent for the treatment of prostate cancer, comprising

i) performing a method of the invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) to determine the level of prostate cancer tumour fraction in the subject;
ii) administering a therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer tumour fraction or if the sample comprises cfDNA derived from a prostate cancer subtype and/or if the sample comprises a level of cfDNA derived from a prostate cancer subtype (for example 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more cfDNA derived from a prostate cancer subtype.

[0504] In certain embodiments, the method of treating a subject comprises administering a therapeutic agent for the treatment of prostate cancer if the subject has a detectable level of prostate cancer tumour DNA, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0505] In certain embodiments the method further comprises administering a second therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer fraction (for example a detectable level of prostate cancer fraction, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction). In one preferred embodiment, the method further comprises administering a second therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer fraction 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0506] In certain embodiments, the method of treating a subject comprises

(iii) at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months, or at least 36 months, after the administration of the therapeutic agent, a further sample comprising cfDNA is obtained from the subject, and the method of the invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) is performed to determine the level of prostate cancer fraction in the further sample.

[0507] The invention also provides a therapeutic agent for the treatment of prostate cancer, for use in the treatment of prostate cancer, wherein

i) a method of the invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) is performed to determine the level of prostate cancer prostate cancer fraction in a subject;
ii) the therapeutic agent is administered if the subject has a level of prostate cancer.

[0508] In certain embodiments, the therapeutic agent for use in the treatment of prostate cancer is one for use in a treatment that comprises administering a therapeutic agent for the treatment of prostate cancer if the subject has a detectable level of prostate cancer tumour DNA, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0509] In certain embodiments, the therapeutic agent for use in the treatment of prostate cancer is one for use in a treatment that comprises administering a second therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer fraction (for example a detectable level of prostate cancer fraction, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction). In one preferred embodiment, the therapeutic agent for use in the treatment of prostate cancer is one for use in a treatment that comprises administering a second therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer fraction 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0510] In certain embodiments, the therapeutic agent for use in the treatment of prostate cancer is one for use in a treatment in which

(iii) at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months, or at least 36 months, after the administration of the therapeutic agent, a further sample comprising cfDNA is obtained from the subject, and the method of the invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein) is performed to determine the level of prostate cancer fraction in the further sample.

Applications of Methods of the Invention

[0511] The present invention also provides a method of determining one or more suitable therapeutic agents for the treatment of prostate cancer in a subject having prostate cancer comprising [0512] performing a method of invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein, to determine the level of prostate cancer fraction in the cfDNA sample); [0513] determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to the level of prostate cancer, whereby one therapeutic agent is suitable for a subject with no level of prostate cancer tumour fraction or a percentage level of prostate cancer fraction of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of prostate cancer fraction or a percentage level of prostate cancer fraction of 0.01% or more;
or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with no level of prostate cancer fraction or a percentage level of prostate cancer fraction of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of prostate cancer fraction or a percentage level of prostate cancer fraction of 0.01% or more.

[0514] In certain embodiments, no level of prostate cancer tumour is no detectable level of prostate cancer. In certain embodiments, a level of prostate cancer tumour is a detectable level of prostate cancer, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction. In certain embodiments, a level of prostate cancer tumour is a detectable level of prostate cancer. In certain embodiments, a level of prostate cancer fraction 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0515] In certain embodiments, the method of determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer comprises [0516] performing a method of invention; [0517] determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to the level of prostate cancer, whereby one therapeutic agent is suitable for a subject with no level of prostate cancer tumour fraction, and two or more therapeutic agents are suitable for a subject with a level of prostate cancer fraction;
or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with no level of prostate cancer fraction, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of prostate cancer fraction.

[0518] In certain embodiments, the method of determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer comprises [0519] performing a method of invention; [0520] determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to the level of prostate cancer, whereby one therapeutic agent is suitable for a subject with a level of prostate cancer fraction of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of prostate cancer fraction of 0.01% or more;
or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with a level of prostate cancer fraction of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of prostate cancer fraction of 0.01% or more.

[0521] The present invention also provides a method of determining a suitable treatment regimen for a subject having prostate cancer comprising: [0522] performing a method of invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein, to determine the level of prostate cancer fraction in the cfDNA sample); [0523] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having no level of prostate cancer fraction or a level of prostate cancer fraction of less than 0.01%, and a non-standard treatment is suitable for a subject with a level of prostate cancer fraction or a level of prostate cancer fraction of 0.01% or more.

[0524] In certain embodiments, no level of prostate cancer tumour is no detectable level of prostate cancer. In certain embodiments, a percentage level of prostate cancer tumour is a detectable level of prostate cancer, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction. In certain embodiments, a level of prostate cancer tumour is a detectable level of prostate cancer. In certain embodiments, a percentage level of prostate cancer fraction 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0525] In certain embodiments, a standard treatment is a treatment with a therapeutic agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with two or more therapeutic agents for the treatment of prostate cancer.

[0526] In certain embodiments, a standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a chemotherapeutic agent for the treatment of prostate cancer and/or a immunotherapy treatment of prostate cancer and/or a targeted treatment of prostate cancer and/or a biologic agent treatment of prostate cancer and/or a radionuclide agent treatment.

[0527] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0528] performing a method of invention; [0529] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having no level of prostate cancer fraction, and a non-standard treatment is suitable for a subject with a level of prostate cancer fraction.

[0530] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0531] performing a method of invention; [0532] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 0.01%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 0.01% or more.

[0533] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0534] performing a method of invention; [0535] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 0.02%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 0.02% or more.

[0536] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0537] performing a method of invention; [0538] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 0.05%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 0.05% or more.

[0539] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0540] performing a method of invention; [0541] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 0.1%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 0.1% or more.

[0542] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0543] performing a method of invention; [0544] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 0.5%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 0.5% or more.

[0545] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising

performing a method of invention;
determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having a percentage level of prostate cancer fraction of less than 1%, and a non-standard treatment is suitable for a subject with a percentage level of prostate cancer fraction of 1% or more.

[0546] The present invention also provides a method of performing a method of invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein, to determine whether the sample comprises cfDNA derived from a prostate cancer subtype);

determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby one therapeutic agent is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a level of cfDNA derived from a prostate cancer subtype of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of cfDNA derived from a prostate cancer subtype (for example a percentage level of cfDNA derived from a prostate cancer subtype of at least 0.01%);
or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a level of cfDNA derived from a prostate cancer subtype of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of cfDNA derived from a prostate cancer subtype (for example a percentage level of cfDNA derived from a prostate cancer subtype of at least 0.01%).

[0547] In certain embodiments, no cfDNA derived from a prostate cancer subtype is no detectable cfDNA derived from a prostate cancer subtype. In certain embodiments, a percentage level of cfDNA derived from a prostate cancer subtype is a detectable level of cfDNA derived from a prostate cancer subtype, for example 0.01% or more, 0.02% or more, 0.03% or more, 0.04% or more, 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction. In certain embodiments, a level of cfDNA derived from a prostate cancer subtype is a detectable level of prostate cancer. In certain embodiments, a percentage level of cfDNA derived from a prostate cancer subtype is 0.01% or more, more preferably 0.02% or more, more preferably 0.03% or more, more preferably 0.04% or more, for example 0.05% or more, 0.1% or more, 0.5% or more, or 1% or more, prostate cancer fraction.

[0548] The present invention also provides a method of determining a suitable treatment regimen for a subject having prostate cancer comprising:

performing a method of invention (for example, a method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA as defined herein, to determine whether the sample comprises cfDNA derived from a prostate cancer subtype);
determining the treatment regimen by reference whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby a standard treatment is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a percentage level of cfDNA derived from a prostate cancer subtype of less than 0.01%, and a non-standard treatment is suitable for a subject with a level cfDNA derived from a prostate cancer subtype (for example a detectable level of prostate cancer fraction) or a percentage level of prostate cancer fraction of at least 0.01%.

[0549] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising

performing a method of invention;
determining the treatment regimen by reference whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby a standard treatment is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a percentage level of cfDNA derived from a prostate cancer subtype of less than 0.1%, and a non-standard treatment is suitable for a subject with a level cfDNA derived from a prostate cancer subtype (for example a detectable level of prostate cancer fraction) or a percentage level of prostate cancer fraction of at least 0.1%.

[0550] In certain embodiments, the method of determining a suitable treatment regimen for a subject having prostate cancer comprising

performing a method of invention;
determining the treatment regimen by reference whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby a standard treatment is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a percentage level of cfDNA derived from a prostate cancer subtype of less than 1%, and a non-standard treatment is suitable for a subject with a level cfDNA derived from a prostate cancer subtype (for example a detectable level of prostate cancer fraction) or a percentage level of prostate cancer fraction of at least 1%.

[0551] In certain embodiments, a standard treatment is a treatment with a therapeutic agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with two or more therapeutic agents for the treatment of prostate cancer.

[0552] In certain embodiments, a standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a chemotherapeutic agent for the treatment of prostate cancer and/or a immunotherapy treatment of prostate cancer and/or a targeted treatment of prostate cancer and/or a biologic agent treatment and/or a radionuclide agent treatment of prostate cancer.

Computer Implemented Methods and Software

[0553] The invention also provides a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product and/or a computer implemented software for performing or implementing the method defined herein, for example the method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject described herein, the methods of treatment and therapeutic agents for use described herein, the methods of determining one or more suitable therapeutic agents for the treatment of prostate cancer described herein, the methods for determining a treatment regimen described herein, and the methods of determining a solid cancer cfDNA methylome signature. A kit of the invention may comprise a computerized method and/or computer-assisted method and/or a computer product and/or a computer implemented software of the present invention.

[0554] A computerized method and/or computer-assisted method and/or a computer product and/or a computer implemented software for performing or implementing a method defined herein comprises performing one or more steps of the relevant method, or in certain embodiments, comprises performing the relevant method. A computerized (or computer implemented) method and/or computer-assisted method and/or a computer implemented software can control a computer product to execute, perform or implement one or more steps of the relevant method, or in certain embodiments, comprises performing the relevant method.

[0555] In certain embodiments, the present invention provides a computer product. A computer product of the present invention has the means for performing or implementing one or more method described herein.

[0556] In some embodiments, a computer product of the present invention comprises at least one memory containing at least one computer program or software adapted to control the operation of the computer system to perform or implement a method that includes receiving and characterizing DNA methylation data e.g., receiving and characterizing methylome sequences of a plurality of cfDNA molecules and determining the average methylation ratio at 10 or more genomic regions, and at least one processor for executing the computer program or software.

[0557] In some embodiments, a computer product of the present invention comprises a non-transitory computer readable medium storing a plurality of instructions that, when executed, control a computer system to perform one or more steps of a method described herein or comprises performing or implementing a method described herein.

[0558] In certain embodiments, a computer product is a product having a computer, where the computer comprises a computer-readable medium embodying software to operate the computer. In some cases, the computer system includes one or more general or special purpose processors and associated memory, including volatile and non-volatile memory devices. In some cases, the computer product memory stores software or computer programs for controlling the operation of the computer system to make a special purpose system according to the invention or to implement a system to perform the methods according to the invention. In some cases, the computer system includes a single or multi-core central processing unit (CPU), an ARM processor or similar computer processor for processing the data. In some cases, the CPU or microprocessor is any conventional general purpose single- or multi-chip microprocessor, a RISC or MISS processor, a Power PC processor, or an ALPHA processor. In some cases, the microprocessor is any conventional or special purpose microprocessor such as a digital signal processor or a graphics processor. The microprocessor typically has conventional address lines, conventional data lines, and one or more conventional control lines. The software or computer program may be executed on dedicated system or on a general purpose computer having, for example, a Windows, Unix, Linux or other operating system. In some instances, the system includes non-volatile memory, such as disk memory and solid state memory for storing computer programs, software and data and volatile memory, such as high speed ram for executing programs and software.

[0559] In certain embodiments, a computer product is a storage device used for storing data accessible by a computer, as well as any other means for providing access to data by a computer. Examples of a storage device-type computer-readable medium include: a magnetic hard disk; an optical disk, such as a CD-ROM and a DVD; a magnetic tape; a memory chip. Examples of a computer-readable physical storage media include any physical computer-readable storage medium, e.g., solid state memory (such as flash memory), magnetic and optical computer-readable storage media and devices, and memory that uses other persistent storage technologies. In certain embodiments, a computer product is computer readable media selected from the group consisting of RAM (random access memory), ROM (read only memory), EPROM (erasable programmable read only memory), EEPROM (electrically erasable programmable read only memory), flash memory or other memory technology, CD-ROM (compact disc read only memory), DVDs (digital versatile disks) or other optical storage media, magnetic cassettes, magnetic tape, and magnetic disk storage.

[0560] In one preferred embodiment, the present invention provides a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product and/or computerized (or computer implemented) software for detection, screening, monitoring, staging, classification and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA, the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in the sample; and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform or implement a method of the invention.

[0561] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and
a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus,
and wherein each genomic region is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each genomic region;
analyse the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0562] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and
a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each of the genomic regions;
analyze the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0563] In one preferred embodiment, the present invention provides a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product method and/or computerized (or computer implemented) software for classifying a prostate cancer patient into one or more of a plurality of treatment categories, the method comprising determining the level of prostate cancer fraction in a sample obtained from a subject, wherein the sample comprises cfDNA, the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample obtained from a subject, wherein the sample comprises cfDNA;
and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform or implement a method of the invention.

[0564] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and
a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus,
and wherein each genomic region is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each genomic region;
analyse the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0565] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and
a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each of the genomic regions;
analyze the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0566] In another preferred embodiment, the present invention provides a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product method and/or computerized (or computer implemented) software for classifying a prostate cancer patient into one or more of a plurality of treatment categories, the method comprising determining the subtype of prostate cancer a sample obtained from a subject, wherein the sample comprises cfDNA, the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample obtained from a subject, wherein the sample comprises cfDNA;
and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform or implement a method of the invention.

[0567] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and
a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus,
and wherein each genomic region is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each genomic region;
analyse the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0568] For example, in one embodiment it causes the computer to perform or implement a method comprising the following steps:

characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and
a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence;
calculate a methylation score using the average methylation ratio for each of the genomic regions;
analyze the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0569] In one embodiment, the plurality of treatment categories are selected from a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, and a chemotherapy agent.

[0570] In one embodiment, the plurality of treatment categories are selected from a treatment with a single agent (for example a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, and a chemotherapy agent); and treatment with a combination of agents (for example, a combination of two or more agents selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, and a chemotherapy agent).

[0571] In one preferred embodiment, the plurality of treatment categories are selected from a treatment with a single agent (for example a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, and a chemotherapy agent); and treatment with a combination of two, three, four of five agents (for example, a combination of two, three, four of five agents selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, and a chemotherapy agent).

[0572] For example, the plurality of treatment categories are selected from a hormonal agent; and a hormonal agent and a chemotherapeutic agent and/or a further hormonal agent.

[0573] In one preferred embodiment, the plurality of treatment categories are selected from a standard treatment (for example a treatment with a hormonal agent); and a non-standard treatment (for example a hormonal agent for the treatment of prostate cancer, and a chemotherapeutic agent for the treatment of prostate cancer and/or a immunotherapy treatment of prostate cancer and/or a targeted treatment of prostate cancer and/or a biologic agent treatment of prostate cancer).

[0574] In certain embodiments, a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product and/or a computer implemented software described herein further comprises treating the subject for prostate cancer using a therapeutic agent for the treatment of prostate cancer;

or ceasing or altering treatment with a therapeutic agent for the treatment of prostate cancer; or initiating a non-therapeutic agent treatment for prostate cancer (for example initiation of treatment by surgery or radiation).

[0575] In another preferred embodiment, the present invention provides a computerized (or computer implemented) method and/or computer-assisted method and/or a computer product and/or computerized (or computer implemented) software for determining a solid cancer cfDNA methylome signature for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of the solid cancer, the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample from a subject known to have the solid cancer;
and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to to perform or implement a method comprising the following steps:
(i) characterize the methylome sequence of a plurality of cfDNA molecules in a first sample comprising cfDNA from a subject known to have the solid cancer, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
(ii) determine the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample by aligning the methylome sequences;
(iii) determine the methylation ratio of each CpG locus and/or average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample;
repeating steps (i) to (iii) for one or more further samples comprising cfDNA each from subjects known to have the solid cancer;
perform a variance analysis of all or a selection of the methylation ratios of the CpG loci and/or all or a selection of average methylation ratios of the genomic regions of the samples;
select a group of CpG loci and/or genomic regions associated with a feature of the samples; and
select CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature.
Prostate Cancer cfDNA Methylome Signatures

[0576] The invention also provides a cfDNA methylome signature comprising a set of genomic locations defining 10 or more genomic regions.

[0577] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of: [0578] a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Tables 1 to 4, and [0579] a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0580] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions, a set of genomic locations defining 500 or more genomic regions, a set of genomic locations defining 600 or more genomic regions, a set of genomic locations defining 700 or more genomic regions, a set of genomic locations defining 800 or more genomic regions, a set of genomic locations defining 900 or more genomic regions, or a set of genomic locations defining 1000 genomic regions.

[0581] The signature is for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of prostate cancer. The methylation state (for example the average methylation ratio) of the genomic regions defined by the set of genomic locations of the signature may be used to indicate one or more of the following: the presence of prostate cancer cfDNA in the cfDNA sample, the level of prostate cancer fraction in the cfDNA sample, a subtype of prostate cancer (for example a genomic subtype or molecular subtype, such as castration resistant prostate cancer), if the prostate cancer is metastatic, the aggression of the prostate cancer, the prognosis of the prostate cancer, the tumour response to a treatment, the relapse of the prostate cancer, and/or the residual disease following curative treatment. The methylation state of the genomic regions defined by the set of genomic locations of the signature may be used to indicate the presence of prostate cancer cfDNA in the cfDNA sample and/or the level of prostate cancer fraction in the cfDNA sample.

[0582] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Tables 1 and 3, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Tables 1 and 3 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the the 100 bp genomic locations defined in Tables 1 and 3.

[0583] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions. In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Tables 2 and 4, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Tables 2 and 4 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Tables 2 and 4.

[0584] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions.

[0585] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Tables 1 and 2, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Tables 1 and 2 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Tables 1 and 2.

[0586] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions, or a set of genomic locations defining 500 or more genomic regions.

[0587] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Tables 3 and 4, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Tables 3 and 4 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Tables 3 and 4.

[0588] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions.

[0589] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Table 5, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Table 5 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Table 5.

[0590] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions.

[0591] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Table 6, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Table 6 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Table 6.

[0592] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions.

[0593] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Table 7, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Table 7 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Table 7.

[0594] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 genomic regions.

[0595] The invention also provides a cfDNA methylome signature comprising a set of genomic locations defining 10 or more genomic regions. [0596] a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Table 8, and [0597] a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Table 8 and comprising at least one CpG locus.

[0598] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 or more genomic regions, a set of genomic locations defining 200 or more genomic regions, a set of genomic locations defining 300 or more genomic regions, a set of genomic locations defining 400 or more genomic regions, a set of genomic locations defining 500 or more genomic regions, a set of genomic locations defining 600 or more genomic regions, a set of genomic locations defining 700 or more genomic regions, a set of genomic locations defining 800 or more genomic regions, a set of genomic locations defining 900 or more genomic regions, or a set of genomic locations defining 1000 genomic regions.

[0599] The signature is for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of prostate cancer.

[0600] The methylation state (for example the average methylation ratio) of the genomic regions defined by the set of genomic locations of the signature may be used to indicate one or more of the following: the presence of prostate cancer cfDNA in the cfDNA sample, a subtype of prostate cancer (for example a genomic subtype or molecular subtype, such as one that has an aggressive clinical course and/or a AR copy number gain), if the prostate cancer is metastatic, the aggression of the prostate cancer, the prognosis of the prostate cancer, the tumour response to a treatment, the relapse of the prostate cancer, and/or the residual disease following curative treatment. Preferably, the methylation state of the genomic regions defined by the set of genomic locations of the signature may be used to indicate the presence of prostate cancer cfDNA in the cfDNA sample and/or a subtype of prostate cancer, such as one that has an aggressive clinical course and/or a AR copy number gain.

[0601] In one embodiment, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the group consisting of:

a 100 to 200 bp (for example a 100 to 150 bp or 100 to 120 bp) genomic location comprising or having a genomic location defined in Table 9, and
a 2 to 99 bp (for example a 20 to 99 bp or 50 to 99 bp) genomic location within a genomic location defined in Table 9 and comprising at least one CpG locus. For example, the set of genomic locations defining 10 or more genomic regions are genomic locations selected from the 100 bp genomic locations defined in Table 9.

TABLE-US-00010 TABLE 9 A preferred subset of hypomethylated region genomic locations of Table 8 (The genomic locations are locations with reference to hg19; all regions including, having, or within the genomic locations of table 9 are hypomethylated regions) Chromosome start end chr12 52240301 52240400 chr8 143535751 143535850 chr17 81036151 81036250 chr8 143535801 143535900 chr5 142005201 142005300 chr17 81036101 81036200 chr12 52240351 52240450 chr19 47736001 47736100 chr10 3480051 3480150 chr14 101123351 101123450 chr8 144303301 144303400 chr7 95155001 95155100 chr8 143535501 143535600 chr15 41219401 41219500 chr15 41219451 41219550 chr7 1251201 1251300 chr8 143535851 143535950 chr2 189191651 189191750 chr8 144303251 144303350 chr8 143535601 143535700 chr3 23782851 23782950 chr1 1936451 1936550 chr7 158800951 158801050 chr12 322251 322350 chr1 15655951 15656050 chr8 143535701 143535800 chr20 36037701 36037800 chr20 36037751 36037850 chr17 7083051 7083150 chr7 5319551 5319650 chr17 7083001 7083100 chr10 131650451 131650550 chr1 1936501 1936600 chr19 35818801 35818900 chr10 3479951 3480050 chr4 1160801 1160900 chr19 47735751 47735850 chr10 3494301 3494400 chr17 78982051 78982150 chr10 4331801 4331900 chr1 1920801 1920900 chr9 132482351 132482450 chr8 1923051 1923150 chr16 1159851 1159950 chr2 189191701 189191800 chr1 200707101 200707200 chr20 48124151 48124250 chr19 35818851 35818950 chr10 131650701 131650800 chr10 3379051 3379150 chr10 3449001 3449100 chr12 107297051 107297150 chr19 35981501 35981600 chr13 106063151 106063250 chr5 2207051 2207150 chr8 54164751 54164850 chr3 129326701 129326800 chr1 223435701 223435800 chr2 11294551 11294650 chr17 25798951 25799050 chr22 37215901 37216000 chr11 45392501 45392600 chr11 45392551 45392650 chr17 35277351 35277450 chr9 89410901 89411000 chr9 89410951 89411050 chr8 103572851 103572950 chr6 168629801 168629900 chr3 129326651 129326750 chr1 204655151 204655250 chr1 204655201 204655300 chr1 88108801 88108900 chr10 4386801 4386900 chr2 11294501 11294600 chr16 49530551 49530650 chr16 49530601 49530700 chr7 95155051 95155150 chr10 73324401 73324500 chr5 150538351 150538450 chr7 1388201 1388300 chr3 186170701 186170800 chr8 1923101 1923200 chr8 54164651 54164750 chr16 1316401 1316500 chr10 4386851 4386950 chr4 1535701 1535800 chr8 144213001 144213100 chr10 131650651 131650750 chr10 3480001 3480100 chr3 64305701 64305800 chr3 64305751 64305850 chr1 1936551 1936650 chr10 3480101 3480200 chr10 3277051 3277150 chr4 24796601 24796700 chr3 46622551 46622650 chr14 104688501 104688600 chr1 55504701 55504800 chr22 37215951 37216050 chr1 172291651 172291750 chr1 2527501 2527600 chr15 27210251 27210350 chr8 54164601 54164700 chr7 3019151 3019250 chr11 71010451 71010550 chr19 35981451 35981550 chr16 876151 876250 chr8 1923001 1923100 chr7 1251251 1251350 chr1 38606051 38606150 chr10 131650501 131650600 chr4 140201651 140201750 chr14 105052601 105052700 chr10 3378851 3378950 chr14 106095451 106095550 chr12 6933201 6933300 chr8 54164801 54164900 chr13 106063101 106063200 chr10 94448551 94448650 chr8 54164701 54164800 chr17 79459401 79459500 chr7 158818151 158818250 chr6 25727351 25727450 chr5 1010951 1011050 chr1 2424651 2424750 chr3 128724951 128725050 chr12 322951 323050 chr10 3591201 3591300 chr10 3591251 3591350 chr1 2424701 2424800 chr7 1687001 1687100 chr17 27396901 27397000 chr4 7252451 7252550 chr10 134610401 134610500 chr7 1388151 1388250 chr5 2207001 2207100 chr6 37503051 37503150 chr10 131752851 131752950 chr8 143546801 143546900 chr15 102094651 102094750 chr14 101128351 101128450 chr3 64338501 64338600 chr3 64338551 64338650 chr2 209271151 209271250 chr1 15655901 15656000 chr16 29267801 29267900 chr12 107297101 107297200 chr22 43621801 43621900 chr10 5406551 5406650 chr17 79109751 79109850

[0602] In such embodiments, preferably the signature comprises a set of genomic locations defining 12 or more genomic regions, for example a set of genomic locations defining 15 or more genomic regions, a set of genomic locations defining 20 or more genomic regions, a set of genomic locations defining 25 or more genomic regions, a set of genomic locations defining 30 or more genomic regions, a set of genomic locations defining 50 or more genomic regions, a set of genomic locations defining 75 or more genomic regions, a set of genomic locations defining 100 or more genomic regions, a set of genomic locations defining 125 or more genomic regions, a set of genomic locations defining 150 genomic regions.

Methods for Determining a Solid Cancer cfDNA Methylome Signature

[0603] The present invention also provides methods for determining a solid cancer cfDNA methylome signature. Suitably, such signatures are used, for example, in detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of a solid cancer. They can also suitably be used with the methods and kits for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of a solid cancer and in methods for treatment of solid cancer.

[0604] In one embodiment, the invention provides a method for determining a solid cancer cfDNA methylome signature for use in detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of the solid cancer, the method comprising: [0605] (i) characterizing the methylome sequence of a plurality of cfDNA molecules in a first sample comprising cfDNA from a subject known to have the solid cancer, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0606] (ii) determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample by aligning the methylome sequences; [0607] (iii) determining the methylation ratio of each CpG locus and/or average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample; [0608] repeating steps (i) to (iii) for one or more further samples comprising cfDNA each from subjects known to have the solid cancer; [0609] performing a variance analysis of all or a selection of the methylation ratios of the CpG loci and/or all or a selection of average methylation ratios of the genomic regions of the samples; [0610] selecting a group of CpG loci and/or genomic regions associated with a feature of the samples; and [0611] selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature.

[0612] In certain embodiments the solid cancer is prostate cancer (for example acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer, and particularly acinar adenocarcinoma prostate cancer or ductal adenocarcinoma prostate cancer). In certain embodiments, the solid cancer is a metastatic cancer. In certain embodiments, the solid cancer is a relapsed and/or refractory solid cancer. In certain embodiments, the solid cancer is a subtype of a solid cancer, for example a subtype of a prostate cancer, for example a prostate cancer with specific molecular characteristics and/or genetic characteristics of the cancer cells.

[0613] The first sample is a sample that comprises cfDNA. The sample may suitably be a blood sample, a plasma sample, or a urine sample. In certain embodiments, the sample is a blood sample or a plasma sample. In certain embodiments, the sample is a urine sample.

[0614] Each further sample is a sample that comprises cfDNA. Each further sample may suitably be a blood sample, a plasma sample, or a urine sample. In certain embodiments, one or more further sample(s) is/are blood sample(s) or plasma sample(s). In certain embodiments, one or more further sample(s) is/are urine sample(s). In certain embodiments, all of the further samples are of the same type, for example each further sample is a blood sample; or each further sample is a plasma sample; or each further sample is a urine sample. In certain embodiments, each further sample is a blood sample; or each further sample is a plasma sample.

[0615] In one preferred embodiment, the first sample and each further sample are all samples of the same type. For example, the first sample and each further sample are all blood samples; or the first sample and each further sample are all plasma samples; or the first sample and each further sample are all urine samples. In certain embodiments, the first sample and each further sample are all blood samples; or the first sample and each further sample are all plasma samples.

[0616] In one embodiment, the first sample comprising cfDNA is from a subject known to have or suspected of having metastatic solid cancer. For example, the sample comprising cfDNA is from a subject known to have metastatic solid cancer.

[0617] In one embodiment, the one or more further samples comprising cfDNA are each from subjects known to have or suspected of having metastatic solid cancer. For example, the one or more further samples comprising cfDNA are each from subjects known to have metastatic solid cancer.

[0618] In one embodiment, the first and each further sample comprising cfDNA are each from subjects known to have or suspected of having metastatic solid cancer. For example, the first and each further sample comprising cfDNA are each from subjects known to have metastatic solid cancer.

[0619] In one embodiment, the first sample and one or more of the further samples are from the same subject, for example the same subject but at different time points, for example before treatment, during a treatment, after a treatment, before progression, after progression, after relapse, and/or after change of the disease to metastatic cancer.

[0620] In one embodiment, the first sample and each of the further samples are from the same subject, for example the same subject but at different time points, for example before treatment, during a treatment, after a treatment, before progression, after progression, after relapse, and/or after change of the disease to metastatic cancer.

[0621] In one embodiment, the first sample and one or more of the further samples are from different subjects. The different subjects may all have the same type of the solid cancer, or may all have a different type of the solid cancer, or some may have the same and some may have a different type of the solid cancer. A type of solid cancer may be metastatic, and a different type may be non-metastatic cancer. Another type of solid cancer may be a solid cancer that responds to a certain treatment (e.g. a hormonal agent), and a solid cancer that does not respond to that treatment (e.g. a hormonal agent). For prostate cancer, different types of that solid cancer include acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, and small cell prostate cancer. For prostate cancer, different types of that solid cancer also include castration sensitive prostate cancer and castration resistant prostate cancer.

[0622] In one embodiment, the first sample and one or more of the further samples are from different subjects. The different subjects may all have the same subtype of the solid cancer, or may all have a different subtype of the solid cancer, or some may have the same and some may have a different subtype of the solid cancer. A subtype of solid cancer may be subtype based on characteristics of the cancer cells, and in particular molecular and genetic characteristics of the cells. An example of prostate cancer subtypes include androgen sensitive prostate cancer, androgen insensitive prostate cancer, AR copy number gain, and prostate cancer with an aggressive clinical course.

[0623] In one embodiment, the first sample and one or more of the further samples have different levels of cancer fraction of cfDNA. In one embodiment, the first sample and one or more of the further samples have similar levels of cancer fraction of cfDNA. The level of cancer fraction in a cfDNA sample can be determined by, for example, using methods that estimate tumour fraction using genomic markers.

[0624] Each subject is preferably the same species, for example each subject (i.e. the first subject and each of the one or more further subjects) are human.

[0625] In certain embodiments, the method comprises the additional step of obtaining a biological sample from the first subject and/or obtaining a biological sample from one or more further subjects, for example from each of the one or more further subjects.

[0626] The method for determining a solid cancer cfDNA methylome signature may further comprise isolating the cfDNA from the first sample, and isolating the cfDNA from the one or more further samples. Methods for isolating the cfDNA from the sample described elsewhere herein may be used in the method for determining a solid cancer cfDNA methylome signature.

[0627] The method comprises characterizing the methylome sequence of a plurality of cfDNA molecules in a first sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule. The method also comprises characterizing the methylome sequence of a plurality of cfDNA molecules in each of one or more further samples, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule. Methods for characterizing the methylome sequence of a plurality of cfDNA molecules described elsewhere herein may be used in the method for determining a solid cancer cfDNA methylome signature.

[0628] A plurality of cfDNA molecules may be, for example, at least 100, at least 1000, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). Preferably, a plurality of cfDNA molecules may be, for example, at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). More preferably, a plurality of cfDNA molecules may be, for example, at least 100,000, at least 500,000, at least 1,000,000 (10.sup.6), at least 5,000,000 (5×10.sup.6), at least 10,000,000 (10.sup.7), at least 100,000,000 (10.sup.8), or at least 1,000,000,000 (10.sup.9). The plurality of cfDNA molecules that are characterised for the first sample and for each of the one or more further samples may be the same or may be different.

[0629] The method comprises determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) by aligning the methylome sequences in the first sample. The method also comprises determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) by aligning the methylome sequences in each of of the one or more further samples. Aligning the methylome sequences can, for example, be carried out using a variety of techniques known in the art. For example, a DNA sequence alignment tool, (e.g., BSMAP (PMID: 19635165), Bismark (PMID: 21493656), gemBS (PMID: 30137223), Arioc (PMID: 29554207), BS-Seeker2 (PMID: 24206606), MethylCoder (PMID: 21724594) or BatMeth2 (PMID: 30669962)) can be used to align the reads. The reads may be aligned to reference genome (for example hg38, hg19, hg18, hg17 or hg16).

[0630] In certain embodiments, the method comprises removing duplications of reads of the same DNA molecule (i.e. duplications of reads of the same cfDNA molecule). In this step, sequence reads having exactly the same sequence and start and end base pairs (i.e. the same unclipped alignment start and unclipped alignment end of the sequence) are removed, as they are likely to be duplicate sequence reads of the same sequence (i.e. duplicate of reads of the same cfDNA molecule). For example, PCR duplications can be removed as part of the aligning step, such as using Picard tools v2.1.0 (http://broadinstitute.github.io/picard).

[0631] Preferably, determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample comprises aligning the methylome sequences with a reference genome for the subject, for example for a human subject by aligning the methylome sequences with hg38, hg19, hg18, hg17 or hg16.

[0632] Preferably, determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the one or more further samples comprises aligning the methylome sequences for each of the one or more further samples with a reference genome for the subject, for example for a human subject by aligning the methylome sequences with hg38, hg19, hg18, hg17 or hg16.

[0633] Preferably, the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to the same reference genome, for example the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to hg38; or the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to hg19; or the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to hg18; or the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to hg17; or the methylome sequences in the first sample and the methylome sequences in each of the one or more further samples are aligned to hg16.

[0634] In certain preferred embodiments, the cfDNA molecules in the first sample and the one or more further samples may correspond to a CpG locus or a genomic region of 2 to 5000 bp. More preferably, cfDNA molecules correspond to a CpG locus or a genomic region of 2 to 5000 bp, 2 to 4000 bp, 2 to 3000 bp, 2 to 2000 bp, 2 to 1000 bp, 2 to 800 bp, 2 to 600 bp, 2 to 500 bp, 2 to 400 bp, 2 to 300 bp, or 2 to 200 bp. In one very preferred embodiment, the cfDNA molecules correspond to a CpG locus or a genomic region of 2 to 200 bp for example 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190 or 200 bp. In another preferred embodiment, the cfDNA molecules correspond to a CpG locus or a genomic region of 10 to 150 bp, 20 to 150 bp, 50 to 150 bp, 50 to 120 bp, 80 to 120 bp, 90 to 110 bp. In one preferred embodiment, the cfDNA molecules correspond to a genomic region of 100 bp.

[0635] The method comprises determining the methylation ratio of each CpG locus or the average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample, and determining the methylation ratio of each CpG locus or the average methylation ratio of each genomic region of (preferably 2 to 200 bp) in each of the one or more further samples.

[0636] The average methylation ratio is the average of the methylation ratios of all the CpG loci within a given genomic region, and can be calculated by determining the sum of the methylation ratios of all CpG within a given genomic region and dividing the sum by the number of CpG within the given genomic region. If a genomic region has only 1 CpG locus, the average methylation is the same as the methylation ratio for the single CpG locus in the genomic region.

[0637] The method comprises repeating steps (i) to (iii) for one or more further samples comprising cfDNA each from subjects known to have the solid cancer. As such, the method comprises:

characterizing the methylome sequence of a plurality of cfDNA molecules in each of one or more further samples comprising cfDNA each from a subject known to have the solid cancer, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in each of one or more further samples by aligning the methylome sequences;
determining the methylation ratio of each CpG locus or the average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in each of one or more further samples.

[0638] Thus, for the first sample and for each of the one or more further samples, the methylation ratio of each CpG locus or the average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the characterised cfDNA molecules are determined.

[0639] In certain embodiments, there is one further sample. In certain embodiments there is more than one further sample.

[0640] Preferably there are 2 or more further samples, 3 or more further samples, 4 or more further samples, 5 or more further samples, 6 or more further samples, 7 or more further samples, 8 or more further samples, 9 or more further samples, 10 or more further samples, 12 or more further samples, 15 or more further samples, 20 or more further samples, 25 or more further samples, 30 or more further samples, 40 or more further samples, 50 or more further samples, 60 or more further samples, 70 or more further samples, 80 or more further samples, 90 or more further samples, 100 or more further samples, 200 or more further samples, 300 or more further samples, 400 or more further samples, 500 or more further samples or 1000 or more further samples.

[0641] In one preferred embodiment there are 5 or more further samples, 10 or more further samples, 15 or more further samples, 20 or more further samples, 25 or more further samples, 50 or more further samples, 100 or more further samples, 200 or more further samples, 300 or more further samples, 400 or more further samples, 500 or more further samples or 1000 or more further samples.

[0642] In one preferred embodiment there are 10 or more further samples, 15 or more further samples, 20 or more further samples, 25 or more further samples, 50 or more further samples, 100 or more further samples, 200 or more further samples, 300 or more further samples, 400 or more further samples, 500 or more further samples or 1000 or more further samples.

[0643] The method comprises performing a variance analysis of all or a selection of the methylation ratios of the CpG loci and/or all or a selection of average methylation ratios of the genomic regions of the samples. A variance analysis results in groupings of CpG locus and/or genomic regions associated with features of the samples.

[0644] A cfDNA sample from a subject having a solid cancer is a heterogenous mixture of cfDNA from a primary source (for example, for a blood or plasma sample the primary source of cfDNA molecules are cfDNA from white blood cells, or in a urine sample the primary source of cfDNA molecules is a mixture of cfDNA from white blood cells, immune cell and urinary tract lining cells) and cfDNA from cancer cells. cfDNA in different samples (i.e. samples from different subjects and/or from the same subject at different time points) have differences in methylation levels. The inventors have surprisingly found that very useful methylome signatures can be found by performing a variance analysis of methylation ratios of CpG loci and/or average methylation ratios of genomic regions in multiple cfDNA samples from cancer patients. As not all DNA ends up as cfDNA, in view of the method of the invention determining variance in cfDNA samples, the signatures found using the method include CpG loci and/or genomic regions that are found in cfDNA samples. Additionally, the signatures found using this method can include both cancer-specific and tissue specific methylation. Thus, signatures found using the method of the invention will be especially useful and accurate when used in methods for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of a solid cancer in a cfDNA sample, and especially in a sample of the same type as was used to find the signature.

[0645] A selection of the methylation ratios and/or a selection of average methylation ratios may be, for example at least 95%, at least 90%, at least 80%, at least 70%, at least 60%, at least 50%, at least 40%, at least 30%, at least 20%, at least 10%, or at least 5% methylation ratios and/or average methylation ratios. A selection of the methylation ratios and/or a selection of average methylation ratios of the genomic regions of the samples may be, for example less than 95%, less than 90%, less than 80%, less than 70%, less than 60%, less than 50%, less than 40%, less than 30%, less than 20%, less than 10%, or less than 5% methylation ratios and/or average methylation ratios.

[0646] A selection of the methylation ratios and/or a selection of average methylation ratios of the genomic regions of the samples may be a selection of the methylation ratios of the CpG loci and/or a selection of average methylation ratios of the genomic regions for one or more chromosomes. For example, selection of the methylation ratios and/or a selection of average methylation ratios of the genomic regions for one or more of chromosome 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, X and/or Y.

[0647] A selection of the methylation ratios and/or a selection of average methylation ratios of the genomic regions of the samples may be a selection of the methylation ratios of the CpG loci and/or a selection of average methylation ratios of the genomic regions wherein all samples have at least 1 characterised cfDNA molecule covering of each of the CpG loci and/or genomic regions. For example, wherein each sample has at least 10 (for example at least 15, 20, 25, 50, 100 or 1000) characterised cfDNA molecules covering each of the CpG loci and/or genomic regions.

[0648] In one preferred embodiment, the variance analysis performed is dimensionality reduction. For example, the variance analysis performed is a principal component analysis, a logistic regression analysis, a nearest neighbour analysis, a support vector machine, a neural network model, a NMF (non-negative matrix factorisation), an ICA (independent component analysis), FA (factor analysis), surrogate variable analysis (SVA), and independent surrogate variable analysis (ISVA).

[0649] In one preferred embodiment, the variance analysis performed is a principal component analysis.

[0650] In embodiments wherein the variance analysis performed is a principal component analysis, the CpG locus and/or genomic regions associated with features of the samples are the groupings of the different principal components, such as principal component 1, principal component 2, principal component 3, principal component 4, principal component 5, principal component 6, principal component 7, principal component 8 or a higher principal component.

[0651] The variance analysis performed will group CpG loci and/or genomic regions associated with different feature of the samples.

[0652] The variance analysis (for example the dimensionality reduction) is optionally followed by feature selection methods. An optional feature selection method can be implemented using R, python languages or equivalent statistical application or software.

[0653] The method comprises selecting a group of CpG loci and/or genomic regions associated with a feature of the samples, i.e. selecting a group from all of the groups that the variance analysis results in. For example, in embodiments wherein the variance analysis performed is a principal component analysis, the selecting a group of CpG loci and/or genomic regions associated with a feature of the samples comprises selecting one of principal component 1, principal component 2, principal component 3, principal component 4, principal component 5, principal component 6, principal component 7, principal component 8 or a higher principal component.

[0654] A feature of the samples may be any feature of the samples, which are each from subjects known to have the solid cancer and which all comprise cfDNA. Examples of a feature of the samples that a group of CpG loci and/or genomic regions may be associated with include, but are not limited to, level of solid cancer fraction in the cfDNA, a type of solid cancer, a subtype of solid cancer, a prognosis, aggression of the solid cancer, and susceptibility of the solid cancer to a treatment.

[0655] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated with a level of solid cancer fraction in the cfDNA.

[0656] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated with a type of solid cancer, for example associated with metastatic cancer; associated with non-metastatic cancer; associated with a type of solid cancer that responds to a certain treatment (e.g. a hormonal agent); or associated with a solid cancer that does not respond to a certain treatment (e.g. a hormonal agent). For a solid cancer that is a prostate cancer, in certain embodiments the group selected is a group of CpG loci and/or genomic regions associated with a type of solid cancer, for example associated with castration resistant prostate cancer; associated with castration sensitive prostate cancer; associated with acinar adenocarcinoma prostate cancer; associated with ductal adenocarcinoma prostate cancer; associated with transitional cell cancer of the prostate; associated with squamous cell cancer of the prostate; or associated with small cell prostate cancer.

[0657] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated with a subtype of solid cancer, for example associated with molecular characteristics of the cancer cells; and/or associated with genetic characteristics of the cancer cells. For a solid cancer that is a prostate cancer, in certain embodiments the group selected is a group of CpG loci and/or genomic regions associated with a subtype of the solid cancer, for example associated with AR copy number gain; and/or associated with an aggressive clinical course.

[0658] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated with a prognosis, for example associated with a good prognosis (for example survival of the subject upon treatment is from at least 1 month to at least 90 years); or associated with a poor prognosis (for example survival of a subject that is expected to be from less than 5 years to less than 1 month).

[0659] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated aggression of the solid cancer.

[0660] In certain embodiments, the group selected is a group of CpG loci and/or genomic regions associated with susceptibility of the solid cancer to a treatment. For example associated with susceptibility of the solid cancer to a treatment with one or more of the following: a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, a chemotherapy agent and a radionuclide treatment.

[0661] The method further comprises selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature. This may include selecting all of the CpG loci and/or genomic regions in the group or selecting a plurality of the CpG loci and/or genomic regions in the group, for example selecting at least 95%, at least 90%, at least 80%, at least 70%, at least 60%, at least 50%, at least 40%, at least 30%, at least 20%, at least 10%, or at least 5%, or for example selecting less than 95%, less than 90%, less than 80%, less than 70%, less than 60%, less than 50%, less than 40%, less than 30%, less than 20%, less than 10%, or less than 5%.

[0662] Selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature may comprise selecting at least 10,000 CpG loci and/or genomic regions, at least 8000 CpG loci and/or genomic regions, at least 5000 CpG loci and/or genomic regions, at least 4000 CpG loci and/or genomic regions, at least 3000 CpG loci and/or genomic regions, at least 2000 CpG loci and/or genomic regions, at least 1000 CpG loci and/or genomic regions, at least 800 CpG loci and/or genomic regions, at least 700 CpG loci and/or genomic regions, at least 600 CpG loci and/or genomic regions, at least 500 CpG loci and/or genomic regions, at least 400 CpG loci and/or genomic regions, at least 300 CpG loci and/or genomic regions, at least 250 CpG loci and/or genomic regions, at least 200 CpG loci and/or genomic regions, at least 150 CpG loci and/or genomic regions, at least 100 CpG loci and/or genomic regions, at least 50 CpG loci and/or genomic regions or at least 10 CpG loci and/or genomic regions.

[0663] Selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature may comprise selecting 10,000 or fewer CpG loci and/or genomic regions, 8000 or fewer CpG loci and/or genomic regions, 5000 or fewer CpG loci and/or genomic regions, 4000 or fewer CpG loci and/or genomic regions, 3000 or fewer CpG loci and/or genomic regions, 2000 or fewer CpG loci and/or genomic regions, 1000 or fewer CpG loci and/or genomic regions, 800 or fewer CpG loci and/or genomic regions, 700 or fewer CpG loci and/or genomic regions, 600 or fewer CpG loci and/or genomic regions, 500 or fewer CpG loci and/or genomic regions, 400 or fewer CpG loci and/or genomic regions, 300 or fewer CpG loci and/or genomic regions, 250 or fewer CpG loci and/or genomic regions, 200 or fewer CpG loci and/or genomic regions, 150 or fewer CpG loci and/or genomic regions, 100 or fewer CpG loci and/or genomic regions, 50 or fewer CpG loci and/or genomic regions or 10 or fewer CpG loci and/or genomic regions.

[0664] Selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature may comprise selecting 10,000 CpG loci and/or genomic regions, 8000 CpG loci and/or genomic regions, 5000 CpG loci and/or genomic regions, 4000 CpG loci and/or genomic regions, 3000 CpG loci and/or genomic regions, 2000 CpG loci and/or genomic regions, 1000 CpG loci and/or genomic regions, 800 CpG loci and/or genomic regions, 700 CpG loci and/or genomic regions, 600 CpG loci and/or genomic regions, 500 CpG loci and/or genomic regions, 400 CpG loci and/or genomic regions, 300 CpG loci and/or genomic regions, 250 CpG loci and/or genomic regions, 200 CpG loci and/or genomic regions, 150 CpG loci and/or genomic regions, 100 CpG loci and/or genomic regions, 50 CpG loci and/or genomic regions or 10 CpG loci and/or genomic regions.

[0665] Preferably, the method comprises selecting at least 5 CpG loci (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) and/or at least 5 genomic regions (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) in the group to provide a cfDNA methylome signature.

[0666] In one embodiment, the method comprises selecting at least 5 CpG loci in the group to provide a cfDNA methylome signature, for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000 CpG loci. In one preferred embodiment, the method comprises selecting at least 10 CpG loci, at least 100 CpG loci, at least 250 CpG loci, or at least 500 CpG loci in the group to provide a cfDNA methylome signature. For example the method comprises selecting 10 CpG loci, 100 CpG loci, 250 CpG loci, 500 CpG loci in the group to provide a cfDNA methylome signature.

[0667] In another embodiment, the method comprises selecting at least 5 genomic regions in the group to provide a cfDNA methylome signature, for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000 genomic regions. In one preferred embodiment, the method comprises selecting at least 10 genomic regions, at least 100 genomic regions, at least 250 genomic regions, or at least 500 genomic regions in the group to provide a cfDNA methylome signature. For example the method comprises selecting 10 genomic regions, 100 genomic regions, 250 genomic regions, 500 genomic regions in the group to provide a cfDNA methylome signature.

[0668] In one preferred embodiment, selecting the CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting the CpG loci and/or genomic regions in the group that have strong (for example high) association with the feature to provide the cfDNA methylome signature. The CpG loci and/or genomic regions with strong (for example high) association with the feature may be CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions most correlated with the feature in the group. For example, CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 8000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 6000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 5000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 4000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 3000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 2000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions most correlated with the feature in the group; CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 800 CpG loci and/or genomic regions most correlated with the feature in the group; or CpG loci and/or genomic regions with strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions most correlated with the feature in the group.

[0669] In one preferred embodiment, CpG loci and/or genomic regions correlated with the feature in the group that have strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions most correlated with the feature in the group. More preferably, CpG loci and/or genomic regions correlated with the feature in the group that have strong (for example high) association with the feature are CpG loci and/or genomic regions that are within the top 800 CpG loci and/or genomic regions most correlated with the feature in the group; or even more preferably CpG loci and/or genomic regions most correlated with the feature in the group that have strong (for example high) association with the feature may be CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions most correlated with the feature in the group.

[0670] In one embodiment wherein the level of methylation variance is determined using a principal component analysis, selecting the CpG loci and/or genomic regions in the group comprises selecting a plurality of CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8, for example selecting a plurality of CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 that have strong (for example high) association with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 5000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 4000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 3000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 2000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8.

[0671] In one embodiment wherein the level of methylation variance is determined using a principal component analysis, selecting the CpG loci and/or genomic regions in the group that that have strong (for example high) association with the feature comprises selecting a plurality of CpG loci and/or genomic regions of principal component 1 correlated with the feature of principal component 1, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; or selecting CpG loci and/or genomic regions that are within the top 5000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; selecting CpG loci and/or genomic regions that are within the top 4000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; selecting CpG loci and/or genomic regions that are within the top 3000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; selecting CpG loci and/or genomic regions that are within the top 2000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; selecting CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1; or selecting CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions of principal component 1 most correlated with the feature of principal component 1.

[0672] The method for determining a solid cancer cfDNA methylome signature may further comprise comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region in one or more of the following: [0673] a sample of non-cancerous tissue of origin of the solid cancer; [0674] a sample of the solid cancer; [0675] a cell-line of the solid cancer; [0676] a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype); [0677] a sample of white blood cells; and/or [0678] a sample of cfDNA from a healthy subject (for example an age-matched healthy subject); and [0679] optionally determining if the selected CpG locus and/or genomic region are associated with methylation patterns in the tissue of origin of the solid cancer and/or the solid cancer.

[0680] A sample of non-cancerous tissue of origin of the solid cancer, sample of the solid cancer, cell-line of the solid cancer; and/or sample of white blood cells may come from the same subject as the first sample and/or the one or more further samples comprising cfDNA; and/or a sample of non-cancerous tissue of origin of the solid cancer, sample of the solid cancer, cell-line of the solid cancer, and/or sample of white blood cells may come from a different subject as the first sample and/or the one or more further samples comprising cfDNA; and/or a sample of non-cancerous tissue of origin of the solid cancer, sample of the solid cancer, cell-line of the solid cancer, and/or sample of white blood cells may come from a different subject as the first sample and each of the one or more further samples comprising cfDNA.

[0681] In embodiments where the sample is a sample of the solid cancer, a sample of non-cancerous tissue of origin of the solid cancer and/or a sample of white blood cell, preferably the sample is from the same subject as the subject of the first sample or a subject of the one or more further samples. Additionally, or alternatively, samples of the solid cancer, samples of non-cancerous tissue of origin of the solid cancerm and/or samples of white blood cell from one or more different subjects to the subject of the first sample and the subjects of the one or more further samples are compared.

[0682] If a sample is from a different subject to the subject of the first sample and/or the subjects of the one or more further samples, preferably the different sample is from a subject that is age-matched subject with the subject of the first sample and/or the subjects of the one or more further samples.

[0683] In one preferred embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region with one or more of the following:

a sample of white blood cells from the subject; and/or
a sample cfDNA from a healthy subject.

[0684] In one embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region with one or more of the following:

a sample of white blood cells from the subject;
a sample of the solid cancer from the subject; and/or
a sample of non-cancerous tissue of origin of the solid cancer from the subject.

[0685] In one embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region with one or more of the following:

a sample of cfDNA from a healthy subject (for example an age-matched healthy subject); and/or
a sample of non-cancerous tissue of origin of the solid cancer from the subject from a healthy subject (for example an age-matched healthy subject).

[0686] In one embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region with one or more of the following:

a sample of the solid cancer from multiple different subjects and optionally a sample of the solid cancer from the subject;
a cell-line of the solid cancer from multiple different subjects; and/or
a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0687] In one embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region with one or more of the following:

a sample of the solid cancer from multiple different subjects and optionally a sample of the solid cancer from the subject;
cell-lines of the solid cancer from multiple different subjects;
a sample of white blood cells from the subject;
samples of white blood cells multiple different subjects; and/or
samples of non-cancerous tissue of origin of the solid cancer from multiple different subjects;
a sample of non-cancerous tissue of origin of the solid cancer from the subject; and/or
a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0688] In one embodiment, the method for determining a solid cancer cfDNA methylome signature further comprises comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG loci and/or genomic regions in a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype), and preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200 or 500 samples) each from a different subject known to have the solid cancer (for example each from a different age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the each cfDNA sample from the different subjects is known and/or wherein each the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0689] The method for determining a solid cancer cfDNA methylome signature may further comprise determining a reference value for each of the selected CpG loci and/or genomic regions. In certain embodiments, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in a cfDNA sample from one or more healthy subjects. In certain embodiments, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in one or more white blood cell samples. In certain embodiments, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in a sample of tissue from one or more healthy subjects. In certain embodiments, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in one or more samples of solid cancer tumour and/or one or more solid cancer cell lines. In certain embodiments, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0690] In one embodiment, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in a cfDNA sample from one or more healthy subjects. In another embodiment, the reference value is based on the methylation level (e.g. the methylation ratio for a CpG locus or the average methylation ratio for a genomic region) of the same CpG locus and/or genomic region in one or more white blood cell samples.

[0691] In certain embodiments, a reference value for each of the selected CpG loci and/or genomic regions is the average methylation ratio of the same CpG locus and/or genomic region in or covered by:

a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject;
a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a sample of white blood cells from a subject, for example the subject or a healthy subject;
a characterized methylome sequence of a white blood cell;
a characterized methylome sequence of a prostate cancer cell line;
a characterized methylome sequence of a cancerous prostate cell;
a characterized methylome sequence of a non-cancerous prostate cell; or
a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0692] The method for determining a solid cancer cfDNA methylome signature may further comprise determining two or more (for example 3 or more, 4 or more, 5 or more, 10 or more, 15 or more, or 20 or more) reference values for each of the selected CpG loci and/or genomic regions (for example 2, 3, 4, 5, 6, 7, 8, 9 10, 15, 20, 30, 40, 50, 100, 200, 500 or 1000 reference values for each of the selected CpG loci and/or genomic regions). The two or more reference values may be selected from the average methylation ratio of the same CpG locus and/or genomic region in or covered by one or more of the following:

a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject;
a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a sample of white blood cells from a subject, for example the subject or a healthy subject;
a characterized methylome sequence of a white blood cell;
a characterized methylome sequence of a prostate cancer cell line;
a characterized methylome sequence of a cancerous prostate cell; or
a characterized methylome sequence of a non-cancerous prostate cell;
a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0693] The method for determining a solid cancer cfDNA methylome signature may further comprise establishing an algorithm for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of the solid cancer using the cfDNA methylome signature.

[0694] The algorithm may be established using, for example, a random forest classifier, a regression analysis algorithm, for example a least absolute shrinkage and selection operator (LASSO) algorithm, a Naïve Bayes classifier, a support vector machine, a perceptron learning algorithm, a decision tree, a gradient boosting tree, a neural network or k-nearest neighbour algorithm. The algorithm can be implemented using R, python languages or equivalent statistical application or software (such as STATA) by one of ordinary skill in the art.

[0695] In certain embodiments, the algorithm is for determining the presence of solid cancer in a further sample comprising DNA using the cfDNA methylome signature.

[0696] In certain embodiments, the algorithm is for determining the level of a solid cancer in a further sample comprising DNA using the cfDNA methylome signature, for example the level of solid cancer tumour fraction.

[0697] In certain embodiments, the algorithm is for determining a subtype of solid cancer in a further sample comprising DNA using the cfDNA methylome signature.

[0698] In preferred embodiments the algorithm comprises comparing the methylation status, the methylation ratio, or the average methylation ratio, for some or all of the selected CpG loci and/or genomic regions of the cfDNA methylome signature to the methylation status, the methylation ratio, or the average methylation ratio for some or all of the selected CpG loci and/or genomic regions in a further sample comprising DNA. Additionally, or alternatively, the algorithm comprises comparing the methylation status, the methylation ratio, or the average methylation ratio, for some or all of the selected CpG loci and/or genomic regions of the cfDNA methylome signature to a reference value for each CpG locus and/or genomic region.

[0699] The invention will now be illustrated in a non-limiting way by reference to the following Example.

EXAMPLES

Example 1: New Prostate Cancer Plasma Methylation Signatures

Materials and Methods

Study Design

[0700] Plasma samples were collected within 30 days of treatment initiation and at progression in two biomarker studies, separately approved by the Istituto Scientifico Romagnolo per lo Studio e la Cura dei Tumori (IRST), Meldola, Italy (REC 2192/2013) and Royal Marsden, London, UK (REC 04/Q0801/6) and in the PREMIERE trial (EudraCT: 2014-003192-28, NCT02288936) that was sponsored and conducted by the Spanish Genito-Urinary oncology Group (SOGUG) (FIGS. 1 and 2). All patients provided written informed consent for these analyses.

[0701] These cohorts were described in Romanel et al. (Romanel, A., et al. Sci Transl Med 7, 312re310 (2015)) and Conteduca et al (Conteduca, V. et al., Ann Oncol 28, 1508-1516, (2017)). Briefly, patients needed to have histologically or biochemically confirmed prostate adenocarcinoma and be starting abiraterone or enzalutamide treatment for progressive mCRPC. Patients were required to receive abiraterone or enzalutamide until disease progression as defined by at least two of the following: a rise in PSA, worsening symptoms, or radiological progression defined as progression in soft-tissue lesions measured by computed tomography (CT) imaging according to modified Response Evaluation Criteria in Solid Tumors or progression on bone scanning according to criteria adapted from the Prostate Cancer Clinical Trials Working Group 2 guidelines. Patients with sufficient vials to allow both genome and methylome assessment were prioritised. Metastases were obtained at rapid warm autopsy in the Peter MacCallum warm autopsy program CASCADE (Cancer tissue Collection After Death) described by Alsop et al. (Alsop, K. et al. A, Nat Biotechnol 34, 1010-1014 (2016). (HREC 15/98, FIG. 2).

Plasma DNA Sequencing

[0702] Circulating DNA (10-25 ng) was extracted from plasma using the QIAamp Circulating Nucleic Acid kit (Qiagen™) and quantified using the Quant-iT high-sensitivity Picogreen double-stranded DNA Assay Kit (Invitrogen by Thermo Fisher™). Germline DNA was extracted from white blood cells using the QIAamp DNA kit (Qiagen™). Genomic NGS was performed as described previously (Romanel, A. et al. Sci Transl Med 7, 404 312re310 (2015)). For methylation assessment, raw plasma DNA was bisulfite treated using the ZYMO™ Gold Kit as per the manufacturer's protocol. Swift Bioscience™ Methyl-Seq was used to generate libraries. CpGs were selected from prior data generated using Illumina Infinium HumanMethylation450k microarray (Roche Nimblegen™ targeted capture kit, Epi CpGiant). Probes were designed to hybridize to strands of fully methylated, partially methylated and fully unmethylated derivatives of the target as described below. Libraries were quantified by KAPA library quantification kit (Roche™) before pooling and sequencing on an Illumina™ HiSeq 2500 using paired-end 100-base pair reads. Sequencing matrices for targeted methylome and LP-WGBS are shown in FIGS. 2 and 3, and details on the pipelines for analysis of sequencing data are provided below.

Processing of Targeted Methylation NGS Data

[0703] Data were processed using fastqc to assess quality and read through adapters were trimmed using Trimmomatic v0.36. Since DNA was bisulfite treated, reads were aligned based on three nucleotides (thymine (T), adenosine (A), guanine (G)) to the human genome (hg)19 using the BSMAP v2.90 (Xi, Y. & Li, W., BMC Bioinformatics 10, 232 (2009); Bolger, A. M., et al, Bioinformatics 30, 2114-2120 (2014)). The duplicated reads were removed with Picard tools v2.1.0 (http://broadinstitute.github.io/picard), and unaligned reads were clipped (hard-clipped) using the bamUtil 1.0.13 (Jun, G et al, Genome Res 25, 918-925 (2015)).

[0704] The CpG methylation ratio of each loci was calculated using formula (I), which takes cytosine (C) and thymidine (T) counts from all reads covering each CpG loci.

[00002]$\begin{array}{cc}\mathrm{Methylation}\mathrm{Ratio}=\frac{C}{C+T}& \left(I\right)\end{array}$

[0705] From all sites included in the predesigned capture panel (Roche Nimblegen SeqCap EpiGiant), only sites with a minimum coverage of 10 reads were considered for further analysis of CpG (FIG. 5). The methylation ratio was computed using the methylKit R package v1.6.2 (Akalin, A. et al. Genome Biol 13, R87 (2012)).

Selection of Optimal Data Inputs for PCA

[0706] Adjacent CpG methylation levels are usually highly related, and previously studies have demonstrated high sensitivity of identifying tissue-specific methylation markers using sliding window approaches (Lehmann-Werman, R. et al. Proc Natl Acad Sci USA 113, E1826-1834 (2016); Guo, S. et al. Nat Genet 49, 635-642 (2017); Sun, K. et al. Proc Natl Acad Sci USA 112, E5503-5512 (2015)). Here adjacent CpG sites were combined into methylation segments of fixed length (the term “methylation segment” and the term “segment” as used in the examples section may also be referred to as a genomic region), and the average methylation ratio across all CpGs within the segment was calculated and used to represent the methylation ratio of the segment using methylKit R package v1.6.2 (Akalin, A. et al. Genome Biol 13, R87 (2012)). Initially 100 bp with sliding window of 50 bp were used and generated >1.47 million windows across all CpGs in the target panel. Principal component analysis (PCA) was applied using the FactoMineR v1.41 package.

[0707] To eliminate potential biases due to the selection of segmentation length, segmentation length parameters were optimised. To do so, segments of 10 bp, 100 bp, 1000 bp and 10,000 bp were tested with sliding windows of 5 bp, 50 bp, 500 bp and 5000 bp, respectively. It was found that the smaller the window size, the more data that had to be drop when combining plasma samples due to variable inputs and sequencing coverage (FIG. 6). It was also found that the average methylation ratio of 100 bp segments with 50 bp sliding windows showed high consistency with the methylation ratio estimated at single CpG level (FIG. 7). The correlation of PC1 with genomically-determined tumour fraction was >90% regardless of window sizes (FIG. 48).

[0708] Thus, to preserve more detailed methylation information, and to guarantee successful execution in a reasonable amount of time, the setting of 100 bp segments with 50 bp sliding window was applied for the rest of the analysis. However, other segment sizes and windows could have been used.

Principal Component Analysis of Targeted Plasma Methylome

[0709] The methylation segments for which methylation ratios available in all baseline samples (n=19) and for which the standard deviation values were in the upper two quartiles, were subjected to principal component analysis (FactorMineR R package v1.41, as described in Lê, S., Josse, J. & Husson, F. FactoMineR: An R Package for Multivariate Analysis. 2008 25, 18, doi:10.18637/jss.v025.i01 (2008).).

[0710] More specifically, unscaled PCA using FactoMineR (http://factominer.free.fr) (Lê, S., Josse, J. & Husson, F. FactoMineR: An R Package for Multivariate Analysis. 2008 25, 18 (2008)) was applied. The PCA model comes with the eigenvector, eigenvalues and correlation matrix comprised of correlation coefficient by each segment. The distribution of the top-K highly correlated segments was plotted based on the correlation matrix returned by PCA, and these segments were highly representative of each eigenvector (e.g., principal component 1, or PC1). To identify the optimal value K of highly correlated segments, multiple K values equal to 10, 100, 1,000, and 10,000 were tested and intra-sample variance calculated, and the correlation between the median of the average methylation ratios with genomically-determined tumour fraction was determined (FIGS. 8 and 9).

[0711] Significant principal components were determined using a permutation test as implemented in the jackstraw R package (v1.2) (https://CRAN.R-project.org/package=jackstraw). The projection of all the samples based on the PCA eigenvectors was based on the average methylation ratio of each segment (i.e. average methylation ratio of all the CpG loci within each region) used in the initial PCA for all the samples. Missing values were imputed based on the PCA method as implemented in the missMDA R package (v1.13), as described in Josse, J. & Husson, F. missMDA: A Package for Handling Missing Values in Multivariate Data Analysis. 2016 70, 31, doi:10.18637/jss.v070.i01 (2016).

Tumour Fraction Estimation

[0712] Genomically-determined tumour fraction was determined from targeted next-generation sequencing (NGS) using CLONET as described in Romanel et al. (2015) and Prandi et al. (Prandi, D. et al. Genome Biol 15, 439 (2014)). On high-coverage targeted methylation NGS, PC1 values were calculated as described above, and the median of PC1 values extracted from healthy volunteers were set as 0%, while the median of PC1 values derived from LNCaP samples were set as 100% tumour purity. The tumour fractions of all the plasma samples were obtained with interpolation using PC1 projected values. For tumour fraction estimation based on low-passage whole genome sequencing (LP-WGS) on bisulfite-treated or non-treated plasma DNA, ichorCNA (Adalsteinsson, V. A. et al. Nat Commun 8, 1324 (2017)) was used as described below. For LP-WGBS PC1 projected values were used.

Analysis of LP-WGS by ichorCNA

[0713] LP-WGS on both bisulfite-treated and untreated plasma DNA was performed with a target 1× coverage. For each sample, reads from LP-WGS on untreated plasma DNA were aligned to the hg19 using BWA-MEM version 0.7.12-r1039 and de-duplicated using Picard tools v2.1.0. The human genome was then divided into non-overlapping bins of 1 million base pairs, and, for each sample, the de-duplicated reads were counted per bin using HMM Copy (http://compbio.bccrc.ca/software/hmmcopy/) (Ha, G. et al. Genome Res 22, 1995-2007 (2012)). Next, ichorCNA (https://github.com/broadinstitute/ichorCNA) was applied to estimate the tumour content of each sample (Adalsteinsson, V. A. et al. Nat Commun 8, 1324 (2017)). The algorithm first removed bins in the centromere regions with a flanking region of 100,000 base pairs. For all the remaining bins read counts were corrected by GC content and mappability issues. The normalised read counts were then fed into the Hidden Markov model (HMM), which is a probabilistic model assigning each bin into one possible state (hemizygous deletions (HETD, 1 copy), copy neutral (NEUT, 2 copies), copy gain (GAIN, 3 copies), amplification (AMP, 4 copies), and high-level amplification (HLAMP, 5 or more copies). Based on the copy number profile, the model estimated a ploidy and tumour content for every sample. Finally, the algorithm was initiated with ploidy values 2 and 3, and normal fraction, which is 1 minus tumour fraction of 0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.95. The solution with maximum likelihood among all of these initial combinations was automatically assigned. The CNA status was estimated based on the log R values of each 1 Mbp region obtained by the ichorCNA analysis with fixed threshold of 0.5 (GAIN: log R≥0.5, LOSS: log R≤0.5).

Analysis of Low Passage Whole Genome Bisulfite Sequencing (LP-WGBS)

[0714] Reads from LP-WGBS were processed as high coverage NGS. To calculate PC1 values derived from LP-WGBS, the default segmentation length of 100 bp was used and the average methylation ratio of each segment (i.e. average methylation ratio of all the CpG loci within each region) was calculated based on formula (I) to determine the methylation ratio of each loci, and then then mean of all CpG loci in a segment was calculated to arrive at the average methylation ratio for a segment. To maximize the available information obtained from the data, methylation data from higher coverage bisulfite data based regularised iterative PCA algorithm (Josse, J. & Husson, F. missMDA: A Package for Handling Missing Values in Multivariate Data Analysis. 2016 70, 31) (missMDA R package (v1.13)) was inputted, and projected on the PCA model as described above. The regularisation process with random initialisation can also circumvent the over-fitting problem, which might reduce the generalization capabilities of the findings.

Analysis of Illumina HumanMethylation450 BeadChip Dataset

[0715] The microarray processed data were obtained from the Gene Expression Omnibus (Edgar, R., et al, Nucleic Acids Res 30, 207-210 (2002)) repository (GSE84043). From the dataset probes overlapping with PC1 segments were selected. The average methylation ratio of each segment was obtained considering the median of the 13 values of the overlapping probes. The tumour fraction estimates by different methods were obtained by the sample information published (Fraser, M. et al, Nature 541, 359-364 (2017)).

Statistical Analysis

Overview

[0716] Pearson correlation was used to measure the association between two parameters (principal component values versus genomically determined tumour fraction estimation, or different approaches of tumour fraction estimations). The association between copy number status of each region and principal components was estimated using the Kruskal-Wallis test. Mann-Whitney U test was used to test significance between two groups (AR gain versus AR non-gain—see FIG. 45). Hazard ratio in overall survival analysis was calculated using the Mantel-Haenszel method. For all tests, a significance threshold of 0.05 was required unless otherwise specified.

Correlation and Association Analysis

[0717] Correlation analyses of continuous measures were performed using the Pearson correlation method as implemented in the R v3.4.0 stats package. The association analysis between principal components and CNA of each region was performed by grouping the principal component values of each sample based on the CNA observed for the region (LOSS, NEUTRAL and GAIN). The differences in the principal component values distribution among groups was then assessed using the Kruskal-Wallis test (one-way ANOVA on ranks) as implemented in the R v3.4.0 stats package.

Methylation Ratio Difference with Kruskal-Wallis and Dunn's Test

[0718] The samples were grouped based on tissue of origin and clinical status (white blood cells, plasma healthy volunteer, plasma baseline and plasma progression). Samples were grouped by ct-MethSig and AR-MethSig, and the average methylation ratio of each 100 bp segment was estimated in each group of samples. To keep the analysis consistent, only segments present in all samples (340,467 segments) were considered. All the selected segments were split in two groups based on the overlap with the promoter region of known genes (263,262 non-promoter segments, 77,205 promoter segments). The promoter region was defined as 1k base-pair upstream and downstream of the transcription start site (TSS). The significance of the differences among each group was calculated using Kruskal-Wallis test (one-way ANOVA on ranks) as implemented in the R v3.4.0 (https://www.R-project.org (2018)) stats package. After defining the significance of the differences, the difference of the average methylation ratio across each group was assessed using the Dunn's test as implemented in FSA R package v0.8.22 (https://github.com/droglenc/FSA).

Functional Enrichment Analysis

[0719] Functional enrichment analysis (chemical and genetic perturbations, MSigDB) was executed using the enrich R package (v0.1) based on all the MSigDB main categories (MSigDB database v6.0) (Liberzon, A. et al. Cell Syst 1, 417-425 (2015)) with a significance threshold of 0.05 on Benjamini corrected p values.

Motif Enrichment Analysis

[0720] Motif enrichment analysis was used to identify potential transcriptomic regulators of methylation signatures (MethSig). MethSig top 1000 correlated segments were submitted to find the possible motif binding sequences over-represented as compared to the default background set (Zambelli, F., et al, Nucleic Acids Res 41, W535-543 (2013)). The pipeline (Pscan-Chip) (Zambelli, F., et al, Nucleic Acids Res 41, W535-543 (2013)) originally designed for the analysis of chromatin immunoprecipitation followed by next generation sequencing technologies was applied. The program automatically scanned 75 bp preceding and after the ‘peak’ regions that were submitted with controlled background, and know transcriptional factor binding motifs obtained from JASPAR version 2018. Local enrichment p-value was two-tailed and denoted whether the motif was over-represented in the 150-bp region compared to the genomic regions flanking them. Global enrichment denoted whether the motif binding sequence was over-represented in the region with respect to global background composed of pan-genome putative regulatory regions from various cell lines. The analysis on top 1000 highly correlated segments with PC1 (i.e. ct-MethSig) or PC3 (i.e. AR-MethSig) was performed and other randomly selected regions from the custom, targeted enrichment panel. The result of AR-MethSig was validated by an orthogonal pipeline (Heinz, S. et al. Mol Cell 38, 576-589 (2010)), and the finding was consistent to original approach as described above.

Gaussian Mixture Model (GMM)

[0721] Average methylation ratios of ct-MethSig segments derived from LNCaP cell lines, and healthy volunteer plasma were extracted. To estimate the probability density function (pdf), kernel density estimation (kde) was applied, assuming a mixture of two Gaussian distributions consistent with the input dataset of normal prostate epithelium (FIG. 28). The Gaussian mixture model (see formula (II) below) applies expectation-maximization (EM) to fit the mixtures of Gaussian distributions by an iterative process (Pedregosa, F. et al. J. Mach. Learn. Res. 12, 2825-2830 (2011)). In the experiment, the model was executed with maximum iterations of 100 times and ‘k-means’ method for initialization, and it was hypothesized that there were two Gaussian distributions, each of them with its own general covariance. The Gaussian mixture model was subject to cross-validation on random split set of regions over 100 times to prove the robustness of the approach (FIG. 38). The fitted GMM (number of class=2) was then used to predict the top 1000 segments of PC1, and thus arrive at the ct-MethSig segments of prostate epithelium (PrEC) (Pidsley, R. et al. Genome Biol 17, 208 (2016)).

Gaussian mixture model: g.sub.j(x)=ø.sub.θ.sub.j(x); where θ.sub.j=(μ.sub.i,σ.sub.j.sup.2)

Results

Results

Interrogating the Plasma DNA Methylome in Metastatic Prostate Cancer

[0722] The mCRPC plasma methylome and genome were concurrently characterized (FIG. 10). Plasma DNA was subjected to either high-coverage targeted or whole genome NGS in order to determine tumour fractions and copy number status. Tumour fractions were derived using genomic information at heterozygous single-nucleotide polymorphisms (SNPs) to computationally determine the abundance of deletions involving 8p21 or 21q22, designated as prostate cancer anchor lesions that were used previously as a proxy for tumour fraction (Prandi, D. et al, Genome Biol 15, 439 (2014); Carreira, S. et al. Sci Transl Med 6, 254ra125 (2014)). Plasma has been collected up to 30 days prior to abiraterone or enzalutamide (baseline) from 25 mCRPC patients (median age: 76 years; range: 42-90) participating in prospective biomarker protocols, with a wide range of genomically-determined tumour fractions and from across the disease spectrum (docetaxel-naïve or docetaxel-treated). From the 25 patients, plasma had been collected from 19 patients at radiographic progression and four control samples were collected from two healthy, male volunteers (aged 30 and 60, FIG. 11, FIGS. 1 and 2). The median and range of genomically-determined tumour fractions in the mCRPC cohort were 0.41 (0.04-0.89) and 0.42 (0.09-0.89) for baseline and progression plasma, respectively.

[0723] A separate aliquot of DNA was subjected to bisulfite treatment and target enrichment NGS for 5.5 million pan-genome CpG sites was performed (target coverage: ≥30×; key sequencing parameters in FIGS. 2 and 3). These CpGs were selected based on their known involvement in or proximity to regions that had been associated with disease (Roche Nimblegen™ targeted capture kit, Epi CpGiant). In total targeted capture was performed on 39 plasma samples (19 baseline, 16 progression, 4 healthy volunteer plasma samples from two individuals, FIG. 5 and FIG. 3). Low-pass whole genome bisulfite sequencing (LP-WGBS) was also performed on 46 plasma samples (24 baseline, 20 progression, two plasma samples from one healthy volunteer—FIGS. 4 and 5). Additionally, targeted bisulfite NGS on 15 white blood cell samples, including white blood cells collected prior to and 108 days after treatment with abiraterone from one patient was conducted (FIGS. 1 and 2).

[0724] Adjacent CpG methylation patterns are usually highly correlated (Guo, S. et al. Nat Genet 49, 635-642, (2017); Lehmann-Werman, R. et al. Proc Natl Acad Sci USA 113, E1826-1834 (2016)). A 100 base-pair sliding window was applied and the data divided into 1.47 million methylation segments as described above. In keeping with prior studies on tissues, the methylation ratio distribution across all methylation segments in plasma and white blood cell samples showed a density peak for hypermethylation and hypomethylation (FIG. 12). Regions with a minimum of 10× coverage were selected. When separated by annotation category (such as promoter, exon, intron), the distribution was consistent with the targeted regions (FIG. 13) (Yu, G., et al, Bioinformatics 31, 2382-2383 (2015)). It was observed that methylation segments in promoter regions were primarily hypomethylated whilst other categories were primarily hypermethylated (FIG. 14, top panel). The average methylation ratio distribution for segments in baseline, progression plasma and healthy volunteer plasma were then compared with white blood cell DNA, and significant differences between plasma and white blood cell samples were observed (P<10.sup.−15, Kruskal-Wallis test). The difference was more pronounced in cancer patients' plasma samples compared to healthy volunteers' ((respectively, Z scores for promoter regions were −20.3, −19.6 and −15.6 and non-promoter regions: −157.2, −170.1 and −5.9; all P<10.sup.−9, Dunn's test, FIG. 14, bottom panel). In keeping with previous studies that the cancer genome is characterized by more hypo-methylation events, the mCRPC plasma methylome that includes a mixture of cancer and normal DNA, is globally more hypomethylated than healthy volunteer plasma.

An Unbiased Approach Identifies Tumour Fraction as the Major Determinant of Global Plasma DNA Methylation Variance

[0725] The analytical framework was applied on baseline plasma methylome (n=19) to identify methylation features associated with genomically-determined tumour fraction. To use an unbiased approach to explore the complexity of pan-genome plasma methylation changes, principal component analysis (PCA) was performed. Different parameters were experimented on and confirmed the robustness of the finding on progression, healthy volunteer plasma methylome and LNCaP cell line methylome. To expand the applicability of the approach, segments highly correlated with principal components were extracted and tested on LP-WGBS plasma methylome, and external, well-defined tissue data sets using orthogonal approaches (FIG. 15).

[0726] The first principal component (PC1) contributed 42% of the variance (FIG. 16) and showed a high correlation with genomically-determined tumour fraction (r=−0.96, P=1.3×10.sup.−10, Pearson correlation, FIG. 17). To investigate whether treatment with AR targeting agents affected the association of PC1 with tumour fraction, PCA eigenvectors were used to project the progression samples, healthy volunteer controls (“0” tumour fraction) and the LNCaP prostate cancer cell line (100% tumour, three replicates, FIG. 18). After including the projected samples, the correlation of PC1 and genomically-determined tumour fraction remained high (r=−0.94, P=1.3×10.sup.−18, FIG. 19).

[0727] To evaluate the clinical applicability of the findings using LP-WGBS, scaled PC1 values were extracted from LP-WGBS. Applying Bland-Altman analysis, a good agreement was found between LP-WGBS derived tumour fraction estimation and estimates from high-coverage targeted NGS (95% limits of agreement: −0.25 to 0.15, bias: —0.05) introducing the opportunity for scalable and cost-efficient circulating tumour DNA detection and quantitation using LP-WGBS (FIG. 20).

Methylation Ratio can Serve as a Proxy for Tumour Fraction

[0728] To test features identified by NGS in datasets with fewer data-points, such as methylation arrays, it was hypothesized that the median of the average methylation ratios of the segments that most strongly correlated to the component features could serve as a proxy of tumour fraction. A high correlation (r≥0.93, Pearson correlation) of the average methylation ratio of the segments with genomically-determined tumour fraction was consistently observed in both negatively (i.e. hypermethylated) and positively (i.e. hypomethylated) correlated group when including 10 to 10,000 segments. Also, the intra-sample variance of average methylation ratios of segments in the top correlated segments gradually increased when more segments were included (FIGS. 8 and 9). The 1000 segments that showed the highest correlation with principle component 1 were selected (the selected 1000 segments are referred to herein as circulating tumour methylation signature or ct-MethSig, FIG. 22). These 1000 segments are shown in Tables 1 to 4 above, grouped by their origin (prostate tissue or cancer specific) and their correlation (negative (i.e. hyper-methylated) or positive (i.e. hypo-methylated)).

[0729] It was confirmed that the median of the average methylation ratios of the selected 1000 segments of the ctMethSig showed a high correlation with tumour fraction (520 segments in negatively (i.e. hypermethylated) correlated regions, hyper-methylated group: r=0.95, P=8.4×10.sup.−19; 480 segments in positively (i.e. hypomethylated) correlated regions, hypo-methylated group: r=−0.93, P=3×10.sup.−16, Pearson correlation, FIG. 22). It is noted that ct-MethSig did not include genes whose methylation status has been previously reported as diagnostic of prostate cancer such as, GSTP1, APC, and RASSF1 ((Massie, C. E., et al, J Steroid Biochem Mol Biol 166, 1-15) as the segments overlapping with these genes were not as strongly correlated with PC1 value as ct-MethSig (FIG. 23)).

[0730] Additionally, the finding that the median of the average methylation ratios of all 1000 segments of the ctMethSig can be used as a proxy for tumour fraction was tested in published tissue data sets and confirmed a high correlation with tumour fraction both in mCRPC (Beltran, H. et al. Nat Med 22, 298-305 (2016)) (hypermethylated group: r=0.92, P<1.5×10.sup.−6; hypomethylated group: r=−0.74, P<1.4 10.sup.−3, Pearson correlation, FIGS. 24A and 24B), and hormone-sensitive prostate cancer (HSPC) (Fraser, M. et al. Nature 541, 359-364 (2017)) (hypermethylated group: r=0.91, P<10.sup.−6°; hypomethylated group: r=−0.61, P<10.sup.−17, Pearson correlation) (FIG. 25).

Functional Enrichment Identifies Hypermethylation of Polycomb Repressor Complex 2 Targets in Circulating Prostate Cancer DNA

[0731] To study the biological processes underlying PC1, gene set enrichment analysis (GSEA) was performed on genes overlapping with ct-MethSig segments (i.e. the DNA segments of the genomic locations shown in Tables 1 to 4 above). Significant enrichment (adjusted P<10.sup.−4) was observed for targets of the polycomb repressor complex 2 (Lee, T. I. et al. Cell 125, 301-313 (2006)) (PRC2 related category in the Molecular Signature Database or MSigDB, FIG. 26). That was of particular interest as a previous mRNA profiling study showed that prostate cancer was distinguished from non-cancer prostate epithelium by down-regulation of genes that are repressed by PRC2 (Yu, J. et al. Cancer Res 67, 10657-10663 (2007)). It was noted that these PRC2 genes were only in the ct-MethSig hypermethylated group, representing an increase in methylation ratio with increasing fraction. Overall, the 520 negatively-correlated segments included 231 genes. Of these, 41 were collectively either components of PRC2-EED (Embryonic Ectoderm Development) (Cao, Q. et al. Nat Commun 5, 3127 (2014)) and SUZ12 (suppressor of zesta 12) (Hojfeldt, J. W. et al. Nat Struct Mol Biol 25, 225-232, (2018)) or H3K27ME3 (tri-methylation of lysine 27 on histone H3 protein subunit) (FIG. 26). A permutation test was performed and the result indicated that PRC2-regulated components were more enriched in ct-MethSig as compared to 1000 randomly selected genomic segments (FIG. 27). The inventors' discovery of hypermethylation in promoters upstream of these genes provides a biological explanation for their down-regulation and introduces a strategy for extending this biological difference to a liquid biopsy application (Beltran, H. et al, Nat Med 22, 298-305 (2016); Yu, J. et al, Cancer Res 67, 10657-10663 (2007)).

The Circulating Tumour Methylation Signature Comprises Segments Specific to Either Normal or Malignant Prostate Epithelium

[0732] It was postulated that ct-MethSig included components that were specific to either prostate malignant or non-malignant epithelium. The kernel density estimation of the ct-MethSig average methylation ratios in whole genome bisulfite sequencing data derived from the non-malignant prostate epithelium cell line (PrEC) (Pidsley, R. et al. Genome Res 28, 625-638, (2018)) was plotted and it was observed that there was a bimodal distribution (FIG. 28). A Gaussian mixture model was adapted on the average methylation ratios of ct-MethSig segments from the prostate cancer cell line LNCaP and the two healthy volunteer plasma samples and then the fitted Gaussian distribution was used on normal prostate epithelium (PrEC). PrEC segments were identified whose average methylation ratio distribution aligned with either LNCaP or healthy volunteer plasma. It was concluded that the former segments with average methylation ratios in normal prostate epithelium similar to LNCaP were prostate epithelium-specific, while the segments with average methylation ratios similar to healthy volunteer plasma were prostate cancer-specific (FIG. 28). These findings were confirmed by showing that CRPC metastases (bone, bladder, liver and lymph nodes, described further in FIG. 29) included segments attributed to both normal and cancerous prostate epithelium whilst normal prostate (54 year-old male donor, ENCODE donor ID: ENCDO451RUA) included only segments attributable to normal prostate epithelium. As a result, ct-MethSig could be split into two components, circulating cancer-specific and normal prostate-specific signatures. Circulating cancer-specific segments are shown in Tables 2 and 4, and normal prostate-specific segments are shown in Tables 1 and 3, above.

[0733] Finally, methylation microarray data from 553 prostate cancers from TCGA and 12 CRPC adenocarcinoma from Beltran et al. (Beltran, H. et al, Nat Med 22, 298-305 (2016)) was used to show that the distribution of ctMethSig segments in localized prostate cancer and CRPC tissue includes both cancer and normal components (FIG. 30).

Prostate Cancer Detection Using Plasma Methylome

[0734] To build a classifier for detection of prostate cancer to accurately categorise prostate cancer subjects and healthy subjects, metastatic prostate cancer plasma samples (N=44) were used as described before (FIGS. 1 and 2) plus fifteen leukocyte samples derived from patients and two healthy volunteer plasma and leukocyte samples. Patient plasma samples were labelled as class A, while the leukocyte and samples collected from healthy volunteer were labelled as class B. The steps to obtain the classifier are shown in FIG. 49.

[0735] The median of the average methylation ratios of all 1000 segments of ct-MethSig across all samples were used as input for random forest classifier (RFC), a classic machine learning classification method. A RFC model was built on and fitted a number of decision trees each of which categorized a subset of samples to improve the prediction accuracy and control for overfitting. The RFC was run with 1000 times cross-validation to ensure the stability of the model. Briefly, the samples were split into two groups—a training group (plasma DNA containing prostate tumour DNA) and a testing group plasma DNA not containing prostate tumour DNA. The classification model was initially built on the training group and the classifier was tested on the testing group. The model was initially built model selecting 10 trees in one forest, and the result showed 100% accuracy (STD=1%) on training and 95% on testing (STD=11%, FIG. 50A). When the number of trees in the forest were increased to 100, the model performance slightly improved to 100% accuracy (STD=1%) on training and 100% on testing (STD=6%, FIG. 50B).

[0736] To investigate whether the randomly selected 1, 10 or 100 segments, or all 1000 segments, of ct-MethSig could construct a reliable classifier, a fixed number of segments (1, 10, and 100) were randomly selected, and these segment(s) used to build RFC (n_estimators=100) with 1000-time iteration. The results indicated that using only 1 randomly selected the testing accuracy was 84% (STD %=20%). The testing accuracy gradually improved when more segments were included (FIGS. 51A to D).

[0737] In summary, the development of a methylation based classifier was achievable and able to identify plasma samples containing circulating tumour DNA with high accuracy.

Methylation Signatures Specific to an Individual's Cancer

[0738] Next plasma DNA methylation changes that could potentially identify distinct methylation subtypes were investigated. The second principal component (PC2) was driven by a single patient (02) and was not investigated further. In the third principal component (PC3) a weak correlation with tumour fraction was found (r=0.01, P=0.96, Pearson correlation) (FIG. 17) and this principal component was investigated in more detail. Similar to the methodology applied to ct-MethSig, the top 1000 segments that were most correlated with this component's values were identified. In contrast to ct-MethSig, these were predominantly positively correlated (i.e. hypomethylated) (FIG. 31). Using the median of the average methylation ratios of all 1000 segments of ctMethSig, it was possible to incorporate array-based methylation data from biopsies from intermediate-risk castration-sensitive prostate cancer (CSPC) (Fraser, M. et al, Nature 541, 359-364 (2017)) and mCRPC (Beltran, H. et al, Nat Med 22, 298-305 (2016)). It was found that the median of the average methylation ratios for the 1000 segments in CRPC plasma and tumour samples presented a greater variability in contrast to CSPC or white blood cells (FIGS. 32 and 35). It was noted that, in contrast to ct-MethSig, a change in tumour fraction before and after treatment did not change the median of the average methylation ratios of the top correlated segments with principal component 3 (FIG. 33). Similarly, inter-patient differences were greater than intra-patient variability in multiple metastases and plasma harvested from the same patient at autopsy (FIGS. 34 and 29).

[0739] Functional enrichment analysis on the top 1000 segments of PC3 (referred to herein as AR-MethSig and the segments shown in Table 8 above) showed enrichment in histone H3 tri-methylation markers (FIGS. 36A and 36B). It was hypothesized that this methylation signature was regulated by a common transcriptional pathway. Therefore known transcriptional factor binding sites (TFBSs) adjacent to within 75 base-pairs of the start of the top 1000 segments using a protocol described previously were investigated (Zambelli, F., et al, Nucleic Acids Res 41, W535-543 (2013)). Notably, the AR binding motif was the only significantly over-represented binding site (local enrichment P=6×10.sup.−4, global enrichment P=3×10.sup.−16; FIGS. 37 and 38). Thus profile was denoted as AR-MethSig.

[0740] AR-MethSig hypomethylation strongly associates with AR copy number gainNext, genome-wide copy number profiles were extracted from LP-WGS and confirmed high similarity between results from the same sample with and without bisulfite treatment (FIG. 39). Using LP-WGBS from plasma samples, copy number alterations were observed at a frequency consistent with previously described studies of mCRPC tissue or plasma (Annala, M. et al, Cancer Discov 8, 444-457(2018); Robinson, D. et al. Cell 162, 454 (2015)) (for example, most commonly: 8q21-24 gain: prevalence 70%; Xq12 gain: prevalence 60%; 8p21 loss: prevalence ≥50%, FIG. 40). More copy number changes were observed with increasing PC1 values, as an increasing tumour fraction improved copy number detection (FIG. 41). It was then confirmed that ct-MethSig or AR-MethSig were not located more frequently in regions of copy number alterations (FIG. 42). To integrate genomic copy number data with specific methylation signatures, the correlation of the copy number of every segment across the genome and PC1 values were evaluated (Kruskal-Wallis test, FIG. 43). Most notably, a significant difference in principal component 3 (PC3) value distributions was identified when comparing AR copy number gain and AR non-gain samples (P=0.018, Kruskal-Wallis test, FIG. 44).

The AR-Regulatory Methylation Signature May Identify Distinct Clinical Phenotypes

[0741] Given the association of PC3 values with AR copy number it was confirmed that patient plasma and tissue samples with AR gain had significantly lower average methylation ratios in the AR-MethSig segments (i.e. average methylation ratios in the AR-MethSig segments indicative of hypomethylation) than AR copy number normal samples (P<0.001 and P=0.023 respectively, Wilcoxon signed-rank test; FIG. 45). A high agreement was found for the median methylation ratio of AR-MethSig extracted from high-coverage targeted NGS and LP-WGBS (95% limits of agreement: −0.136 to 0.076; FIG. 46), again supporting the use of LP-WGBS that is amenable to clinical implementation for methylation-based patient stratification. No hormone-sensitive cancers harboring a low median of the average methylation ratios for the AR-MethSig (i.e. average methylation ratios in the AR-MethSig segments indicative of hypomethylation) were identified, nor did either of the two commonly studied AR-regulated prostate cancer cell lines have a low median of the average methylation ratios for the AR-MethSig (LNCaP and VCaP, FIG. 35). Evaluating the clinical relevance of AR-MethSig was of interest, and a change over time in AR-MethSig median methylation ratio was not observed, so fixed time-points over the disease independent of the time of sampling were chosen: namely time from start of androgen deprivation therapy (ADT) to death. It was observed that AR-MethSig low cancers (i.e. cancers having a median of the average methylation ratios in the AR-MethSig segments indicative of hypomethylation)) had poor clinical prognosis (HR=8.18, 95% CI=1.93-34.76, P=0.0044; Mantel-Cox log-rank test; FIG. 47).

Discussion

[0742] In Example 1, the present inventors performed next-generation sequencing (NGS) on plasma DNA with and without bisulfite treatment from mCRPC patients receiving either abiraterone or enzalutamide in the pre- or post-chemotherapy setting. Using principal component analysis on the mCRPC plasma methylome, the inventors surprisingly found that the main contributor to methylation variance (principal component one, or PC1) was strongly correlated with genomically-determined tumour fraction (r=−0.96; P<10.sup.−8). Further the 1000 top correlated segments of the PC1, “ct-MethSig”, which are presented in Tables 1 to 4 above, revealed that these segments comprised of methylation patterns specific to either prostate cancer or prostate normal epithelium.

[0743] The inventors used a custom target-capture approach to define the methylation status of pan-genome CpG islands. By using 100 bp sliding window strategy, the inventors obtained close to 0.5 million methylation segments with 10× coverage in all of the 19 “baseline” plasma DNA samples and used them to construct a principal component analysis. Novel to the inventors' approach was the construction of their model using solely mCRPC plasma DNA that has a variable ratio of normal DNA, primarily arising from white blood cells (Moss, J. et al. Nat Commun 9, 5068, (2018)), and validating the model using tumour DNA that harbors methylation changes that are either prostate epithelium-specific or cancer-specific. The method resulted in the ct-MethSig signature, the segments of which are shown in Tables 1 to 4. These segments can be used as described herein to very accurately determine the level of prostate cancer fraction in a cfDNA sample as shown, for example, in FIG. 50. The inventors have also found that they were able to implement the signature of Tables 1 to 4 in methylation data with variable CpG coverage, including methylation microarrays or reduced representation bisulfite sequencing.

[0744] The inventors found that the ct-MethSig did not include genes whose methylation status has been previously reported as diagnostic of prostate cancer such as, GSTP1, APC, and RASSF1 (Massie, C. E, et al, J Steroid Biochem Mol Biol 166, 1-15 (2017)). Although not wishing to be bound by theory, the present inventors being that this finding could be explained by highly variable methylation levels at the genomic segments of the signature in non-cancer plasma DNA compared to cancer plasma DNA.

[0745] As well as the signature of Tables 1 to 4 derived from the PC1 found by the present inventors that can be used to determine prostate cancer fraction from a sample, the inventors also surprisingly found a signature that can be used to extract information specific to an individual's cancer. That signature was derived from an orthogonal methylation signature (principal component three (PC3)), and the segments of this signature are defined in Table 8. The inventors surprisingly found that this signature can be used to identify a sub-group of cancers characterized by a more aggressive clinical course and that is enriched for AR copy number gain. In particular, this signature showed enrichment for androgen receptor binding sequences and hypomethylation at putative AR binding sites associated with AR copy number gain. Previous studies have reported worse outcome for patients with AR gain in plasma (Romanel, A. et al. Sci Transl Med 7, 312re310, (2015); Conteduca, V. et al., Ann Oncol 28, 1508-1516, (2017)) and given the high overlap between this genomic lesion and this signature, the inventors believe that this methylation signature identifies the same phenotype. Thus the inventors surprisingly found that a methylation signature can be used to detect a gene abnormality.

[0746] Thus, in summary, the present inventors' plasma methylome investigation using their innovative workflow has led to two novel signatures that can be used in methods, kits and uses as defined herein, to very accurately quantitate tumour fraction or identify distinct biologically-relevant subtypes of mCRPC with distinct biological mechanisms and differential clinical outcomes. As such, the signatures can be used for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises cfDNA.

Further Aspects of the Invention are Defined in the Following Numbered Clauses

[0747] § 1. A method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: [0748] characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0749] determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0750] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0751] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, [0752] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0753] calculating a methylation score using the average methylation ratio for each of the genomic regions; [0754] analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample.

[0755] § 2. The method of clause 1, wherein each of the genomic regions is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 characterized methylome sequences.

[0756] § 3. The method of clause 1 or 2, wherein each of the genomic regions is covered by at least 10 sequence reads, for example at least 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 sequence reads, and preferably wherein each sequence read or the majority of the sequence reads (for example at least 50%, 60%, 70%, 80% or 90% of the sequence reads) are from different characterized methylome sequences.

[0757] § 4. The method of any one of clauses 1 to 3, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises: [0758] determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or [0759] determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or [0760] comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0761] § 5. The method of clause 4, wherein the first group of genomic regions are all of the hypermethylated genomic regions for which the average methylation ratio has been determined, and the second group of genomic regions are all of the hypomethylated genomic regions for which the average methylation ratio has been determined.

[0762] § 6. The method of any one of clauses 1 to 5, wherein analyzing the methylation score to determine the level of prostate cancer fraction in the cfDNA sample comprises comparing the methylation score to one or more reference methylation scores, wherein a reference methylation score is a methylation score calculated for the same genomic regions (for example, calculated using the average methylation ratio for the same genomic regions) in one or more of the following a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0763] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0764] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0765] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0766] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0767] a cfDNA sample from a different subject having prostate cancer, preferably wherein the level of prostate cancer fraction in the cfDNA sample from the different subject is known (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and more preferably wherein each cfDNA sample has a different level of prostate cancer fraction); [0768] a characterized methylome sequence of a white blood cell; [0769] a characterized methylome sequence of a prostate cancer cell line; [0770] a characterized methylome sequence of a cancerous prostate cell; and/or [0771] a characterized methylome sequence of a non-cancerous prostate cell.

[0772] § 7. The method of any one of clauses 1 to 6, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises: [0773] determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined, and
wherein calculating a reference methylation score using the average methylation ratio for each genomic region comprises: [0774] determining the median (or the mean) of the average methylation ratios for all genomic regions; or
wherein calculating a methylation score using the average methylation ratio for each genomic region comprises: [0775] determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions for which the average methylation ratio has been determined, and the second group of genomic regions are all of the hypomethylated genomic regions for which the average methylation ratio has been determined), and
wherein calculating a reference methylation score using the average methylation ratio for each genomic region comprises: [0776] determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions for which the average methylation ratio has been determined, and the second group of genomic regions are all of the hypomethylated genomic regions for which the average methylation ratio has been determined).

[0777] § 8. The method of any one of clauses 1 to 6, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region,

and wherein the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by: [0778] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0779] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0780] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0781] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0782] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0783] a cfDNA sample from a different subject having prostate cancer, wherein preferably the level of prostate cancer fraction in the cfDNA sample from the different subject is known (more preferably multiple cfDNA samples (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably the level of prostate cancer fraction in each cfDNA sample from the different subjects is known, and more preferably wherein each cfDNA sample has a different level of prostate cancer fraction); [0784] a characterized methylome sequence of a white blood cell; [0785] a characterized methylome sequence of a prostate cancer cell line; [0786] a characterized methylome sequence of a cancerous prostate cell; and/or [0787] a characterized methylome sequence of a non-cancerous prostate cell.

[0788] § 9. The method of clause 8, wherein analyzing the methylation score to determine the level of prostate cancer DNA comprises determining the number of methylation ratio scores that are indicative of prostate cancer DNA.

[0789] § 10. The method of any one of clauses 1 to 9, wherein the methylome sequence of a cfDNA molecule is determined by using methylation aware sequencing (for example with bisulfite sequencing), methylation-sensitive restriction enzyme digestion, methylation-specific PCR, methylation-dependent DNA precipitation, methylated DNA binding proteins/peptides, or single molecule sequences without sodium bisulfite treatment.

[0790] § 11. The method of any one of clauses 1 to 10, wherein the methylome sequence of a cfDNA molecule is determined by performing methylation aware sequencing, for example wherein the methylation aware sequencing comprises treating the DNA molecule with sodium bisulfite and performing sequencing of the treated DNA molecule.

[0791] § 12. The method of any one of clauses 1 to 11, comprising determining the average methylation ratio at 25 or more, 50 or more, 100 or more, 150 or more, 200 or more, 300 or more, 400 or more, 500 or more, 600 or more, 700 or more, 800 or more, or 900 or more genomic regions (for example comprising determining the average methylation ratio at 25, 50, 100, 150, 200, 300, 400, 500, 600, 700, 800, 900 or 1000 genomic regions).

[0792] § 13. The method of any one of clauses 1 to 12, wherein the genomic regions are selected from: [0793] a 100 to 150 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0794] a 10 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus.

[0795] § 14. The method of any one of clauses 1 to 13, wherein the genomic regions are selected from: [0796] a 100 to 120 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0797] a 50 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus; or [0798] a 100 to 120 bp region comprising or having a genomic location defined in Table 5, and [0799] a 50 to 99 bp region within a genomic location defined in Table 5 and comprising at least one CpG locus; or [0800] a 100 to 120 bp region comprising or having a genomic location defined in Table 6, and [0801] a 50 to 99 bp region within a genomic location defined in Table 6 and comprising at least one CpG locus; or [0802] a 100 to 120 bp region comprising or having a genomic location defined in Table 7, and [0803] a 50 to 99 bp region within a genomic location defined in Table 7 and comprising at least one CpG locus.

[0804] § 15. The method of any one of clauses 1 to 14, wherein the genomic regions have a 100 bp genomic location defined in any one of Tables 1 to 4, Table 5, Table 6 or Table 7.

[0805] § 16. The method of any one of clauses 1 to 15, comprising characterising the average methylation ratio at 50 or more (for example 50), 100 or more (for example 100), 200 or more (for example 200), 500 or more (for example 500), or 800 or more (for example 800 or 1000) genomic regions, wherein the genomic regions each have a genomic location defined in Tables 1 to 4; or

characterising the average methylation ratio at 10 or more (for example 10), 50 or more (for example 50) or 100 or more (for example 100), wherein each of the genomic regions have a genomic location defined in Table 5; or
characterising the average methylation ratio at 10 or more (for example 10), 50 or more (for example 50) or at 100 or more (for example 100), wherein each of the genomic regions have a genomic location defined in Table 6; or
characterising the average methylation ratio at 10 or more (for example 10), 50 or more (for example 50) or at 100 or more (for example 100), wherein each of the genomic regions have a genomic location defined in Table 7.

[0806] § 17. The method of any one of clauses 1 to 16, wherein at least 25% of the genomic regions are prostate cancer specific genomic regions; or wherein at least 25% of the genomic regions are prostate tissue specific genomic regions.

[0807] § 18. The method of any one of clauses 1 to 17, wherein at least 40% of the genomic regions are prostate cancer specific genomic regions, for example at least 50, 60, 70, 80, 90 or 95% (for example 95, 96, 97, 98, 99 and 100%) of the genomic regions are prostate cancer specific genomic regions; or wherein at least 40% of the genomic regions are prostate tissue specific genomic regions, for example at least 50, 60, 70, 80, 90 or 95% (for example 95, 96, 97, 98, 99 and 100%) of the genomic regions are prostate tissue specific genomic regions.

[0808] § 19. The method of any one of clauses 1 to 18, wherein at least 40% of the genomic regions comprise, have or are within genomic locations defined in Tables 1 and/or 2, or Table 5 or Table 6 or Table 7, for example at least 50, 60, 70, 80, 90 or 95% (for example 95, 96, 97, 98, 99 and 100%) of the genomic regions comprise, have or are within a genomic location defined in Tables 1 and/or 2 or Table 5 or Table 6 or Table 7.

[0809] § 20. The method of any one of clauses 1 to 19, wherein a plurality of cfDNA molecules is at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000, at least 5,000,000, at least 10,000,000, or at least 100,000,000 cfDNA molecules.

[0810] § 21. The method of any one of clauses 1 to 20, wherein the prostate cancer is acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer (for example wherein the prostate cancer is acinar adenocarcinoma prostate cancer or ductal adenocarcinoma prostate cancer).

[0811] § 22 The method of any one of clauses 1 to 21 wherein the prostate cancer is castration resistant prostate cancer and/or is metastatic prostate cancer.

[0812] § 23. The method of any one of clauses 1 to 22, wherein the sample comprising cfDNA is a blood or plasma sample.

[0813] § 24. The method of any one of clauses 1 to 23, further comprising measuring the level of prostate-specific antigen (PSA) in a sample of blood from the subject, and determining if the subject has an abnormal level of PSA in the blood (for example a level of PSA in the blood of at least 4.0 ng/mL or, if the subject has had a previous PSA test, an increased level of PSA compared to the previous test).

[0814] § 25. The method of clause 24, wherein the subject has an abnormal level of PSA in the blood (for example a level of PSA in the blood of at least 4.0 ng/mL or, if the subject has had a previous PSA test, an increased level of PSA compared to the previous test); or wherein the subject has a normal level of PSA in the blood (for example a level of PSA in the blood of 4.0 ng/mL or less).

[0815] § 26. The method of any one of clauses 1 to 25, further comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the level of prostate cancer fraction in the two samples.

[0816] § 27. The method of any one of clauses 1 to 26 for screening and/or prognostication of prostate cancer, wherein prostate cancer is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%.

[0817] § 28. The method of any one of clauses 1 to 27, for detecting, screening and/or prognostication of metastatic prostate cancer, wherein metastatic prostate cancer is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%.

[0818] § 29. The method of any one of clauses 1 to 28, for detecting, screening and/or prognostication of prostate cancer, wherein metastatic prostate cancer with a poor prognosis is predicted when a level of prostate cancer is determined, for example a detectable level of prostate cancer, for example a percentage level of prostate cancer fraction of at least 0.01%.

[0819] § 30. An in-vitro diagnostic kit for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer, comprising one or more reagents for detecting the presence or absence of at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4, or comprising at least one CpG locus defined in Table 5, or comprising at least one CpG locus defined in Table 6, or comprising at least one CpG locus defined in Table 7.

[0820] § 31. The kit as defined in clause 30, wherein the kit comprises one or more reagents for detecting the presence or absence of at least 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, or 900 DNA molecules (for example 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules) having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Tables 1 to 4.

[0821] § 32. The kit as defined in clause 30 or 31, wherein the kit comprises oligonucleotides for specifically hybridizing to at least a section of the at least 10 DNA molecules (for example, at least 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, or 900 DNA molecules) having a DNA sequence corresponding to all or part of a genomic location defined in Tables 1 to 4.

[0822] § 33. The kit of any one of clauses 30 to 32, wherein at least one of the oligonucleotides for specifically hybridizing to at least a section of a DNA molecule is an amplification primer, for example each of the oligonucleotides for specifically hybridizing to at least a section of a DNA molecule is an amplification primer.

[0823] § 34. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform the method of any one of clauses 1 to 29.

[0824] § 35. A computer-executable software for performing the method of any one of clauses 1 to 29.

[0825] § 36. The kit of any one of clauses 30 to 33, wherein the kit comprises instructions for use which define how to determine the level of prostate cancer fraction in a sample comprising cfDNA from a subject, and/or comprises a computer product as defined in clause 34, and/or a computer-executable software as defined in clause 35.

[0826] § 37. A computer-implemented method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: [0827] receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in the sample; [0828] and wherein the computer readable medium comprises instructions that, when executed by the processor, causes the computer to perform a method of any one of clauses 1 to 29 (for example causes the computer to perform a method comprising the following steps: [0829] characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0830] determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0831] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0832] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, [0833] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0834] calculate a methylation score using the average methylation ratio for each of the genomic regions; [0835] analyze the methylation score to determine the level of prostate cancer fraction in the cfDNA sample).

[0836] § 38. A computer-implemented method for classifying a prostate cancer patient into one or more of a plurality of treatment categories, the method comprising determining the level of prostate cancer DNA in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: [0837] receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample obtained from a subject, wherein the sample comprises cfDNA; [0838] and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform a method of any one of clauses 1 to 29 (for example causes the computer to perform a method comprising the following steps: [0839] characterize the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0840] determine the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0841] a 100 to 200 bp region comprising or having a genomic location defined in Tables 1 to 4, and [0842] a 2 to 99 bp region within a genomic location defined in Tables 1 to 4 and comprising at least one CpG locus, [0843] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0844] calculate a methylation score using the average methylation ratio for each genomic region; [0845] analyse the methylation score to determine the level of prostate cancer fraction in the cfDNA sample).

[0846] § 39. The method of any one of clauses 1 to 29, 37 or 38 further comprising treating the subject for prostate cancer using a therapeutic agent for the treatment of prostate cancer;

or ceasing or altering treatment with a therapeutic agent for the treatment of prostate cancer; or initiating a non-therapeutic agent treatment for prostate cancer (for example initiation of treatment by surgery or radiation).

[0847] § 40. A method for treating prostate cancer in a subject comprising the method of one of clauses 1 to 29, 37 or 38 and further comprising treating the subject using a therapeutic agent for the treatment of prostate cancer, surgery, and/or radiotherapy; or a method for treating prostate cancer in a subject, comprising administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer after the subject has been determined to have prostate cancer based on a method as defined in one of clauses 1 to 29, 37 or 38.

[0848] § 41. The method of clause 40, wherein the method of clause 1 to 29, 37 or 38 is performed before and/or after treating the subject.

[0849] § 42. A method of any one of clauses 39 to 41, comprising performing the method of clause 1 to 29, 37 or 38 before treating the subject, and subsequently repeating the method of clause 1 to 29, 37 or 38 after the treatment, for example at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months or at least 36 months after treating the subject.

[0850] § 43. The method of clause 42, wherein the method comprises continuing to treat the subject with the therapeutic agent for the treatment of prostate cancer if the level of prostate cancer fraction is substantially the same in the initial and subsequent method or lower in the subsequent method than in the initial method.

[0851] § 44. The method of clause 42 or 43, wherein the method comprises [0852] ceasing or altering treatment with the therapeutic agent for the treatment of prostate cancer; and/or [0853] initiating treatment with a second therapeutic agent for the treatment of prostate cancer; and/or [0854] initiating a non-therapeutic agent treatment (e.g., surgery or radiation),
if the level of prostate cancer fraction is substantially the same in the initial and subsequent method or higher in the subsequent method than in the initial method.

[0855] § 45. A method of treating a subject in need of treatment with a therapeutic agent for the treatment of prostate cancer, comprising [0856] i) performing the method of any one of clauses 1 to 29, 37 or 38 to determine the level of prostate cancer fraction in the subject; [0857] ii) administering a therapeutic agent for the treatment of prostate cancer if the subject has a level of prostate cancer fraction (for example 0.01% or more prostate cancer fraction).

[0858] § 46. A therapeutic agent for the treatment of prostate cancer for use in the treatment of prostate cancer, whereby [0859] i) the method of any one of clauses 1 to 29, 37 or 38 is performed to determine the level of prostate cancer prostate cancer DNA in a subject; [0860] ii) the therapeutic agent is administered if the subject has a level of prostate cancer.

[0861] § 47. A method as defined in clause 40 to 45, or a therapeutic agent for the treatment of prostate cancer for use as defined in clause 46, wherein a second therapeutic agent for the treatment of prostate cancer is administered if the subject has a level of prostate cancer DNA (for example a detectable level of prostate cancer DNA, for example 0.01% or more prostate cancer DNA).

[0862] § 48. The method of clause 45, or a therapeutic agent for the treatment of prostate cancer for use as defined in clause 46, wherein [0863] (iii) at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months, or at least 36 months, after the administration of the therapeutic agent, a further sample comprising cfDNA is obtained from the subject, and the method of any one of clauses 1 to 29, 37 or 38 is performed to determine the level of prostate cancer DNA in the further sample.

[0864] § 49. A method of determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer comprising [0865] performing the method of any one of clauses 1 to 29, 37 or 38; [0866] determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to the level of prostate cancer, whereby one therapeutic agent is suitable for a subject with no level of prostate cancer fraction (for example an undetectable level of prostate cancer fraction) or a level of prostate cancer fraction of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of prostate cancer DNA (for example a percentage level of prostate cancer fraction of at least 0.01%); [0867] or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with no level of prostate cancer DNA (for example an undetectable level of prostate cancer DNA) or a level of prostate cancer DNA of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of prostate cancer DNA (for example a percentage level of prostate cancer fraction of at least 0.01%).

[0868] § 50. A method of determining a suitable treatment regimen for a subject having prostate cancer comprising [0869] performing the method of any one of clauses 1 to 29, 37 or 38; [0870] determining the treatment regimen by reference to the level of prostate cancer fraction, whereby a standard treatment is suitable for a subject having no level of prostate cancer fraction (for example an undetectable level of prostate cancer fraction) or a percentage level of prostate cancer fraction of less than 0.01%, and a non-standard treatment is suitable for a subject with a level of prostate cancer fraction (for example a detectable level of prostate cancer fraction) or a percentage level of prostate cancer fraction of at least 0.01%.

[0871] § 51. The method as defined in clause 50, wherein the standard treatment is a treatment with a therapeutic agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with two or more therapeutic agents for the treatment of prostate cancer;

or wherein the standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a chemotherapeutic agent for the treatment of prostate cancer and/or a immunotherapy treatment of prostate cancer and/or a targeted treatment of prostate cancer and/or a biologic agent treatment of prostate cancer.

[0872] § 52. A computerized method and/or computer-assisted method for determining one or more suitable therapeutic agents for the treatment of prostate cancer in a subject having prostate cancer, the method comprising performing the steps of clause 49; or a computerized method and/or computer-assisted method for determining a suitable treatment regimen for a subject having prostate cancer, the method comprising performing the steps of clause 50 or clause 51.

[0873] § 53. A method or therapeutic agent as defined in any one of clauses 39 to 52, wherein the therapeutic agent for the treatment of prostate cancer is selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, a chemotherapy agent;

for example:
a hormonal agent selected from LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), estrogens, and steroids (for example prednisone or dexamethasone);
a targeted agent selected from poly(ADP-ribose) polymerase (PARP) inhibitor (for example olaparib, rucaparib, niraparib or talazoparib), a epidermal growth factor receptor (EGFR) inhibitor (for example gefitinib, erlotinib, afatinib, brigatinib, icotinib, cetuximab, or osimertinib, adavosertib, lapatinib), and a tyrosine kinase inhibitor (for example imatinib, gefitinib, erlotinib, sunitinib);
a biologic agent selected from monoclonal antibodies (for example pertuzumab, trastuzumab and Solitomab), hormones (for example a hormonal agent selected from LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), and estrogens), interferons (for example interferons-α, -β, -γ), and interleukin-based products (for example interleukin-2);
an immunotherapy agent selected from a cancer vaccine (for example sipuleucel-T), T-cell therapy, monoclonal antibody therapy, immune checkpoint therapy (for example a PD-1 inhibitor (e.g pembrolizumab, nivolumab, cemiplimab spartalizumab), a PD-L1 inhibitor (e.g. atezolizumab, avelumab or durvalumab), or a CTLA-4 (e.g. ipilimumab)), and non-specific immunotherapies (for example interferons and inerleukins); or
a chemotherapy agent selected from docetaxel, cabazitaxel, and c-Met inhibitors (for example cabozantinib).

[0874] § 54. A method or therapeutic agent as defined in any one of clauses 39 to 52, wherein the therapeutic agent for the treatment of prostate cancer is a hormonal agent and optionally a chemotherapy agent and/or optionally a further hormonal agent and/or optionally a targeted agent and/or optionally a radionuclide agent and/or an immunotherapy agent (for example a LHRH agonist (for example leuprolide, goserelin, triptorelin, or histrelin) or a LHRH antagonist (for example degarelix), and optionally a chemotherapy agent (for example docetaxel, cabazitaxel, carboplatin) and/or optionally a further hormonal treatment (for example enzalutamide, abiraterone, darolutamide) and/or optionally a radionuclide agent (Radium223, PSMA-labelled radionuclide) and/or optionally a PARP inhibitor (for example olaparib, rucaparib, niraparib or talazoparib) and/or an immunotherapy agent (for example nivolumab, pembroluzimab, ipilumimab, durvalumab)).

[0875] § 55. A method for determining a solid cancer circulating free DNA (cfDNA) methylome signature for use in detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of the solid cancer, the method comprising: [0876] (i) characterizing the methylome sequence of a plurality of cfDNA molecules in a first sample comprising cfDNA from a subject known to have the solid cancer, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0877] (ii) determining the respective number of characterised cfDNA molecules corresponding to a CpG locus or a genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample by aligning the methylome sequences; [0878] (iii) determining the methylation ratio of each CpG locus and/or average methylation ratio of each genomic region of 2 to 10,000 bp (preferably 2 to 200 bp) in the first sample; [0879] repeating steps (i) to (iii) for one or more further samples comprising cfDNA each from subjects known to have the solid cancer; [0880] performing a variance analysis of all or a selection of the methylation ratios of the CpG loci and/or all or a selection of average methylation ratios of the genomic regions of the samples; [0881] selecting a group of CpG loci and/or genomic regions associated with a feature of the samples; and [0882] selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature.

[0883] § 56. The method of clause 55, wherein the method further comprises aligning the methylome sequences for the first sample with a reference genome for the subject; and aligning the methylome sequences for each of the one or more further samples with the same reference genome.

[0884] § 57. The method of clause 55 or 56, wherein the reference genome is selected from hg38, hg19, hg18, hg17 and hg16.

[0885] § 58. The method of any one of clauses 55 to 57, comprising selecting at least 25 CpG loci (for example at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) and/or at least 25 genomic regions (for example at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature.

[0886] § 59. The method of any one of clauses 55 to 58, wherein the variance analysis performed is a dimensionality reduction.

[0887] § 60. The method as defined in clause 59, wherein the dimensionality reduction is a principal component analysis, a logistic regression analysis, a nearest neighbor analysis, a support vector machine, a neural network model, a NMF (non-negative matrix factorisation), an ICA (independent component analysis) or a FA (factor analysis) is used to determine the level of methylation variance in the samples.

[0888] § 61. The method as defined in clause 60, wherein the variance analysis performed is a principal component analysis.

[0889] § 62. The method as defined in clause 61, wherein selecting a group of CpG loci and/or genomic regions associated with a feature of the samples comprises selecting one of principal component 1, principal component 2, principal component 3, principal component 4, principal component 5, principal component 6, principal component 7, principal component 8 or a higher principal component.

[0890] § 63. The method of any one of clauses 55 to 62, wherein selecting the CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting the CpG loci and/or genomic regions in the group that have strong association with the feature, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions most correlated with the feature in the group (for example selecting CpG loci and/or genomic regions that are within the top 8000, 5000, 3000, 2000, 1000, 800, 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions most correlated with the feature in the group).

[0891] § 64. The method of any one of clauses 55 to 63, wherein selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting at least 5 CpG loci (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) and/or at least 5 genomic regions (for example at least 8, at least 10, at least 12, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 75, at least 100, at least 200, at least 300, at least 400, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000 or at least 10,000) in the group to provide a cfDNA methylome signature.

[0892] § 65. The method of clause 61 or 62, or clauses 63 and 64 when dependent on clauses 61 or 62, wherein selecting CpG loci and/or genomic regions in the group to provide the cfDNA methylome signature comprises selecting a plurality of CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8, for example selecting CpG loci and/or genomic regions that are within the top 10,000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 5000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 4000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 3000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 2000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; selecting CpG loci and/or genomic regions that are within the top 1000 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8; or selecting CpG loci and/or genomic regions that are within the top 500, 400, 300, 250, 200, 150, 100, 50 or 10 CpG loci and/or genomic regions of principal component 1, 2, 3, 4, 5, 6, 7 or 8 most correlated with the feature of principal component 1, 2, 3, 4, 5, 6, 7 or 8.

[0893] § 66. The method of any one of clauses 55 to 65, wherein the first sample comprising cfDNA and each of the one or more further samples is a blood sample; or wherein the first sample comprising cfDNA and each of the one or more further samples is a plasma sample.

[0894] § 67. The method of any one of clauses 55 to 66, wherein the cancer is prostate cancer.

[0895] § 68. The method of any one of clauses 55 to 67 comprising repeating steps (i) to (iii) for 2 or more further samples, 3 or more further samples, 4 or more further samples, 5 or more further samples, 6 or more further samples, 7 or more further samples, 8 or more further samples, 9 or more further samples, 10 or more further samples, 12 or more further samples, 15 or more further samples, 20 or more further samples, 25 or more further samples, 30 or more further samples, 40 or more further samples, 50 or more further samples, 60 or more further samples, 70 or more further samples, 80 or more further samples, 90 or more further samples, 100 or more further samples, 200 or more further samples, 300 or more further samples, 400 or more further samples, 500 or more further samples or 1000 or more further samples comprising cfDNA each from subjects known to have the solid cancer.

[0896] § 69. The method of any one of clauses 55 to 68, wherein the first sample and one or more of the further samples are from different subjects (for example wherein the first sample and each of the one or more of the further samples are from different subjects) and/or wherein the first sample and one or more of the further samples are from the same subject, for example the same subject but at different time points, for example before treatment, during a treatment, after a treatment, before progression, after progression, and/or after change of the disease to metastatic cancer.

[0897] § 70. The method of any one of clauses 55 to 69, further comprising comparing the methylation state of each of the selected CpG loci and/or genomic regions in the first sample and in the one or more further samples with the methylation state of the same CpG locus and/or genomic region in one or more of the following: [0898] a sample of non-cancerous tissue of origin of the solid cancer; [0899] a sample of the solid cancer; [0900] a cell-line of the solid cancer; [0901] a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype); [0902] a sample of white blood cells; and/or [0903] a sample of cfDNA from a healthy subject (for example an age-matched healthy subject); and [0904] optionally determining if the selected CpG locus and/or genomic region are associated with methylation patterns in the tissue of origin of the solid cancer and/or the solid cancer.

[0905] § 71. The method of any one of clauses 55 to 70, further comprising determining a reference value (for example one more reference value, e.g. 3 or more, 4 or more, 5 or more, 10 or more, 15 or more, or 20 or more reference values) for each of the selected CpG loci and/or genomic regions, for example wherein a reference value for each of the selected CpG loci and/or genomic regions is the average methylation ratio of the same CpG locus and/or genomic region in or covered by: [0906] a cfDNA sample from a healthy subject, for example a healthy age-matched subject; [0907] a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject; [0908] a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient; [0909] a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line; [0910] a sample of white blood cells from a subject, for example the subject or a healthy subject; [0911] a characterized methylome sequence of a white blood cell; [0912] a characterized methylome sequence of a prostate cancer cell line; [0913] a characterized methylome sequence of a cancerous prostate cell; [0914] a characterized methylome sequence of a non-cancerous prostate cell; or [0915] a sample of cfDNA from a subject known to have the solid cancer (for example an age-matched subject known to have the solid cancer, and for example wherein the level of cancer fraction in the cfDNA sample from the different subject is known and/or wherein the sample is known to comprise cfDNA derived from a prostate cancer subtype).

[0916] § 72. The method of any one of clauses 55 to 71, further comprising establishing an algorithm for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, prognostication and/or treatment of the solid cancer using the cfDNA methylome signature, for example wherein [0917] the algorithm is for determining the presence of solid cancer in a further sample comprising DNA using the cfDNA methylome signature; and/or [0918] the algorithm is for determining the level of a solid cancer in a further sample comprising DNA using the cfDNA methylome signature, for example the level of solid cancer tumour fraction; and/or [0919] the algorithm is for determining a subtype of solid cancer in a further sample comprising DNA using the cfDNA methylome signature.

[0920] § 73. The method of clause 72, where the algorithm comprises comparing the methylation status, the methylation ratio, or the average methylation ratio, for some or all of the selected CpG loci and/or genomic regions of the cfDNA methylome signature to the methylation status, the methylation ratio, or the average methylation ratio for some or all of the selected CpG loci and/or genomic regions in a further sample comprising DNA; and/or wherein the algorithm comprises comparing the methylation status, the methylation ratio, or the average methylation ratio, for some or all of the selected CpG loci and/or genomic regions of the cfDNA methylome signature to a reference value for each CpG locus and/or genomic region.

[0921] § 74. A computer implemented method for determining a solid cancer cfDNA methylome signature for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of the solid cancer, the method comprising performing the method of any one of clauses 55 to 73.

[0922] § 75. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform the method of any one of clauses 55 to 73.

[0923] § 76. A computer-executable software for performing the method of any one of clauses 55 to 73.

[0924] § 77. A computer-implemented software for determining a solid cancer cfDNA methylome signature for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of the solid cancer, the method comprising: [0925] receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample from a subject known to have the solid cancer; [0926] and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform a method of any one of clauses 55 to 73.

[0927] Further aspects of the invention are defined in the following numbered clauses:

[0928] § 1. A method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising: [0929] characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule; [0930] determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of: [0931] a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and [0932] a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus, [0933] and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence; [0934] calculating a methylation score using the average methylation ratio for each of the genomic regions; [0935] analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0936] § 2. The method of § 1, wherein the method comprises determining the level of cfDNA in the sample that is derived from a prostate cancer subtype.

[0937] § 3. The method of § 1 or § 2, wherein each of the genomic regions is covered by at least one sequence read of at least two characterized methylome sequences, for example at least one sequence read of at least 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 characterized methylome sequences.

[0938] § 4. The method of any one of § 1 to § 3, wherein each of the genomic regions is covered by at least 10 sequence reads, for example at least 10, 12, 15, 20, 25, 50, 100, 200, 300, 400, 500, or 1000 sequence reads, and preferably wherein each sequence read or the majority of the sequence reads (for example at least 50%, 60%, 70%, 80% or 90% of the sequence reads) are from different characterized methylome sequences.

[0939] § 5. The method of any one of § 1 to § 4, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises:

determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or
determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score; or
comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region.

[0940] § 6. The method of § 5, wherein the first group of genomic regions are all of the hypermethylated genomic regions for which the average methylation ratio has been determined, and the second group of genomic regions are all of the hypomethylated genomic regions for which the average methylation ratio has been determined.

[0941] § 7. The method of any one of § 1 to § 6, wherein analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype comprises comparing the methylation score to one or more reference methylation scores, wherein a reference methylation score is a methylation score calculated for the same genomic regions (for example, calculated using the average methylation ratio for the same genomic regions) in one or more of the following

a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject;
a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a sample of white blood cells from a subject, for example the subject or a healthy subject;
a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (more preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably the each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype);
a characterized methylome sequence of a white blood cell;
a characterized methylome sequence of a prostate cancer cell line;
a characterized methylome sequence of a cancerous prostate cell; and/or
a characterized methylome sequence of a non-cancerous prostate cell.

[0942] § 8. The method of any one of § 1 to § 7, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises:

determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined, and
wherein calculating a reference methylation score using the average methylation ratio for each genomic region comprises:
determining the median (or the mean) of the average methylation ratios for all genomic regions for which the average methylation ratio has been determined; or
wherein calculating a methylation score using the average methylation ratio for each genomic region comprises
determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions, and the second group of genomic regions are all of the hypomethylated genomic regions), and
calculating a reference methylation score using the average methylation ratio for each genomic region comprises:
determining the median (or the mean) of the average methylation ratios for a first group of genomic regions to obtain a first methylation score and/or determining the median (or the mean) of the average methylation ratios for second group of genomic regions to obtain a second methylation score (for example wherein the first group of genomic regions are all of the hypermethylated genomic regions, and the second group of genomic regions are all of the hypomethylated genomic regions).

[0943] § 9. The method of any one of § 1 to § 8, wherein calculating a methylation score using the average methylation ratio for each genomic region comprises comparing the average methylation ratio at each genomic region to a reference methylation ratio for each genomic region to determine a methylation ratio score for each genomic region,

and wherein the reference methylation ratio is the average methylation ratio for the same genomic region in or covered by:
a cfDNA sample from a healthy subject, for example a healthy age-matched subject;
a tissue sample from a healthy subject, for example a prostate tissue sample from a healthy subject;
a cancer biopsy sample from a cancer patient, for example a prostate cancer biopsy sample from a prostate cancer patient;
a cancer cell line sample, for example a prostate cancer cell line sample from a prostate cancer cell line;
a sample of white blood cells from a subject, for example the subject or a healthy subject;
a cfDNA sample from a different subject having prostate cancer, wherein preferably the sample is known to comprise cfDNA derived from the prostate cancer subtype (preferably multiple cfDNA samples (for example at least 2, 3, 4, 5, 10, 20, 40, 50, 100, 200, 300 or 500 samples) each from a different subject having prostate cancer, wherein preferably each sample is known to comprise cfDNA derived from the prostate cancer subtype, and more preferably wherein each cfDNA sample has a different level of cfDNA derived from the prostate cancer subtype);
a characterized methylome sequence of a white blood cell;
a characterized methylome sequence of a prostate cancer cell line;
a characterized methylome sequence of a cancerous prostate cell; and/or
a characterized methylome sequence of a non-cancerous prostate cell.

[0944] § 10. The method of § 9, wherein analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype comprises determining the number of methylation ratio scores that are indicative of the prostate cancer subtype.

[0945] § 11. The method of any one of § 1 to § 10, wherein the methylome sequence of a cfDNA molecule is determined by using methylation aware sequencing (for example with bisulfite sequencing), methylation-sensitive restriction enzyme digestion, methylation-specific PCR, methylation-dependent DNA precipitation, methylated DNA binding proteins/peptides, or single molecule sequences without sodium bisulfite treatment.

[0946] § 12. The method of any one of § 1 to § 11, wherein the methylome sequence of a cfDNA molecule is determined by performing methylation aware sequencing, for example wherein the methylation aware sequencing comprises treating the DNA molecule with sodium bisulfite and performing sequencing of the treated DNA molecule.

[0947] § 13. The method of any one of § 1 to § 12, wherein the genomic regions are selected from:

a 100 to 150 bp region comprising or having a genomic location defined in Table 8, and
a 10 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus; or
a 100 to 150 bp region comprising or having a genomic location defined in Table 9, and
a 10 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus.

[0948] § 14. The method of any one of § 1 to § 13, wherein the genomic regions are selected from:

a 100 to 120 bp region comprising or having a genomic location defined in Table 8, and
a 50 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus; or
a 100 to 120 bp region comprising or having a genomic location defined in Table 9, and
a 50 to 99 bp region within a genomic location defined in Table 9 and comprising at least one CpG locus.

[0949] § 15. The method of any one of § 1 to § 14, wherein the genomic regions have a 100 bp genomic location defined in Table 8, or wherein the genomic regions have a 100 bp genomic location defined in Table 9.

[0950] § 16. The method of any one of § 1 to § 15, comprising characterising the average methylation ratio at 25 or more, 50 or more, 100 or more, 150 or more, 200 or more, 300 or more, 400 or more, or 500 or more genomic regions (for example comprising determining the average methylation ratio at 25, 50, 100, 150, 200, 300, 400 or 500 genomic regions), wherein the genomic regions have a genomic location defined in Table 8.

[0951] § 17. The method of any one of § 1 to § 15, comprising characterising the average methylation ratio at 25 or more, 50 or more, 100 or more, 150 or more, 200 or more, 300 or more, 400 or more, or 500 or more genomic regions (for example comprising determining the average methylation ratio at 25, 50, 100, 150, 200, 300, 400 or 500 genomic regions), wherein the genomic regions have a genomic location defined in Table 8.

[0952] § 16. The method of any one of § 1 to § 16, comprising characterising the average methylation ratio at 25 or more, 50 or more, 100 or more, 125 or more, or 150 genomic regions (for example comprising determining the average methylation ratio at 25, 50, 100, 125, or 150 genomic regions), wherein the genomic regions have a genomic location defined in Table 9.

[0953] § 18. The method of any one of § 1 to § 17, wherein at least 25% of the genomic regions are prostate tissue specific genomic regions; or wherein at least 25% of the regions are prostate cancer specific genomic regions.

[0954] § 19. The method of any one of § 1 to § 18, wherein at least 40% of the genomic regions are prostate cancer specific genomic regions, for example at least 50, 60, 70, 80, 90 or 95% (for example 95, 96, 97, 98, 99 and 100%) of the genomic regions are prostate cancer specific genomic regions; or wherein at least 40% of the genomic regions are prostate tissue specific genomic regions, for example at least 50, 60, 70, 80, 90 or 95% (for example 95, 96, 97, 98, 99 and 100%) of the genomic regions are prostate tissue specific genomic regions.

[0955] § 20. The method of any one of § 1 to § 19, wherein a plurality of cfDNA molecules is at least 10,000, at least 50,000, at least 100,000, at least 500,000, at least 1,000,000, at least 5,000,000, at least 10,000,000, or at least 100,000,000 cfDNA molecules.

[0956] § 21. The method of any one of § 1 to § 20, wherein the prostate cancer is acinar adenocarcinoma prostate cancer, ductal adenocarcinoma prostate cancer, transitional cell cancer of the prostate, squamous cell cancer of the prostate, or small cell prostate cancer.

[0957] § 22. The method of any one of § 1 to § 21 wherein the prostate cancer is castration resistant prostate cancer and/or is metastatic prostate cancer.

[0958] § 23. The method of § 1 to § 22, wherein the prostate cancer subtype is one that has an aggressive clinical course and/or androgen receptor (AR) copy number gain, for example an androgen-insensitive prostate cancer subtype.

[0959] § 24. The method of any one of § 1 to § 23, wherein the sample comprising cfDNA is a blood or plasma sample.

[0960] § 25. The method of any one of § 1 to § 24, further comprising measuring the level of prostate-specific antigen (PSA) in a sample of blood from the subject, and determining if the subject has an abnormal level of PSA in the blood (for example a level of PSA in the blood of at least 4.0 ng/mL), or, if the subject has had a previous PSA test, an increased level of PSA compared to the previous test.

[0961] § 26. The method of any one of § 1 to § 25, further comprising repeating the method on a second sample obtained from the subject after the subject has undergone a treatment for prostate cancer, wherein the second sample comprises cfDNA, and comparing the detectable level of cfDNA derived from a prostate cancer subtype in each sample.

[0962] § 27. The method of any one of § 1 to § 26, for screening and/or prognostication of prostate cancer, wherein prostate cancer with a poor prognosis is predicted when cfDNA derived from the prostate cancer subtype is identified in the sample, for example a detectable level of cfDNA derived from the prostate cancer subtype, for example a percentage level of cfDNA derived from the prostate cancer subtype of at least 0.01%.

[0963] § 28. An in-vitro diagnostic kit for use in the detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer, comprising one or more reagents for detecting the presence or absence of at least 10 DNA molecules having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Table 8 or Table 9.

[0964] § 29. The kit as described in § 28, wherein the kit comprises one or more reagents for detecting the presence or absence of at least 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, or 900 DNA molecules (for example 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, 900 or 1000 DNA molecules) having a DNA sequence corresponding to all or part of a genomic location comprising at least one CpG locus defined in Table 8 or Table 9.

[0965] § 30. The kit as described in § 28 or § 29, wherein the kit comprises oligonucleotides for specifically hybridizing to at least a section of the at least 10 DNA molecules (for example, at least 15, 20, 30, 40, 50, 75, 100, 150, 200, 250, 300, 400, 500, 600, 700, 800, or 900 DNA molecules) having a DNA sequence corresponding to all or part of a genomic location defined in Table 8 or Table 9.

[0966] § 31. The kit of any one of § 28 to § 30, wherein at least one of the oligonucleotides for specifically hybridizing to at least a section of a DNA molecule is an amplification primer, for example each of the oligonucleotides for specifically hybridizing to at least a section of a DNA molecule is an amplification primer.

[0967] § 32. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform the method of any one of § 1 to § 27.

[0968] § 33. A computer-executable software for performing the method of any one of § 1 to § 27.

[0969] § 34. The kit of any one of § 28 to § 31, wherein the kit comprises instructions for use which define how to determine whether a sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, and/or comprises a computer product as defined in § 32, and/or a computer-executable software as defined in § 33.

[0970] § 35. A computer-implemented method for detecting, screening, monitoring, staging, classification, selecting treatment for, ascertaining whether treatment is working in, and/or prognostication of prostate cancer in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in the sample;
and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform a method of any one of § 1 to § 27 (for example causes the computer to perform a method comprising the following steps:
characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determining the average methylation ratio at 10 or more genomic regions, each genomic region being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and
a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic regions is covered by at least one sequence read of at least one characterized methylome sequence;
calculating a methylation score using the average methylation ratio for each of the genomic regions;
analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0971] § 36. A computer-implemented method for classifying a prostate cancer patient into one or more of a plurality of treatment categories, the method comprising determining the level of prostate cancer DNA in a sample obtained from a subject, wherein the sample comprises circulating free DNA (cfDNA), the method comprising:

receiving a data set in a computer comprising a processor and a computer readable medium, wherein the data set comprises the methylome sequence of a plurality of cfDNA molecules in a sample obtained from a subject, wherein the sample comprises cfDNA;
and wherein the computer readable medium comprises instructions that, when executed by the processors, causes the computer to perform a method of any one of § 1 to § 27, for example causes the computer to perform a method comprising the following steps:
characterizing the methylome sequence of a plurality of cfDNA molecules in the sample, wherein the methylome sequence of a cfDNA molecule is the DNA sequence and the methylation profile of the molecule;
determining the average methylation ratio at 10 or more genomic regions, each of the genomic regions being selected from the group consisting of:
a 100 to 200 bp region comprising or having a genomic location defined in Table 8, and
a 2 to 99 bp region within a genomic location defined in Table 8 and comprising at least one CpG locus,
and wherein each of the genomic region is covered by at least one sequence read of at least one characterized methylome sequence;
calculating a methylation score using the average methylation ratio for each of the genomic regions;
analyzing the methylation score to determine whether the sample comprises cfDNA derived from a prostate cancer subtype.

[0972] § 37. The method of any one of § 1 to § 27, § 35 or § 36 further comprising treating the subject for prostate cancer using a therapeutic agent for the treatment of prostate cancer; or ceasing or altering treatment with a therapeutic agent for the treatment of prostate cancer; or initiating a non-therapeutic agent treatment for prostate cancer (for example initiation of treatment by surgery or radiation).

[0973] § 38. A method for treating prostate cancer in a subject comprising the method of § 1 to § 27, § 35 or § 36 and further comprising treating the subject using a therapeutic agent for the treatment of prostate cancer, surgery, and/or radiotherapy; or a method for treating prostate cancer in a subject, comprising administering to the subject an effective amount of a therapeutic agent for the treatment of prostate cancer after the subject has been determined to have prostate cancer subtype based on a method as defined in § 1 to § 27, § 35, or § 36.

[0974] § 39. The method of § 38, wherein the method of § 1 to § 27, § 35, or § 36 is performed before and/or after treating the subject.

[0975] § 40. A method of any one of § 37 to § 39, comprising performing the method of § 1 to § 27, § 35, or § 36 before treating the subject, and subsequently repeating the method of use § 1 to § 27, § 35, or § 36 after the treatment, for example at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months or at least 36 months after treating the subject.

[0976] § 41. The method of § 40, wherein the method comprises continuing to treat the subject with the therapeutic agent for the treatment of prostate cancer if the cfDNA derived from a prostate cancer subtype is detected in the sample and/or the sample comprises a level of cfDNA derived from the prostate cancer subtype that is substantially the same in the initial and subsequent method or lower in the subsequent method than in the initial method.

[0977] § 42. The method of § 40 or § 41, wherein the method comprises

ceasing or altering treatment with the therapeutic agent for the treatment of prostate cancer; and/or
initiating treatment with a second therapeutic agent for the treatment of prostate cancer; and/or
initiating a non-therapeutic agent treatment (e.g., surgery or radiation),
if the sample comprises cfDNA derived from a prostate cancer subtype and/or the sample comprises a level of cfDNA derived from a prostate cancer subtype that is substantially the same in the initial and subsequent method or higher in the subsequent method than in the initial method.

[0978] § 43. The method of § 42, wherein the second therapeutic agent is a chemotherapeutic agent or a PARP inhibitor.

[0979] § 44. A method of treating a subject in need of treatment with a therapeutic agent for the treatment of prostate cancer, comprising

i) performing the method of any one of § 1 to § 27, § 35, or § 36 to determine if the sample comprises cfDNA derived from a prostate cancer subtype and/or determine the level of cfDNA in the sample derived from a prostate cancer subtype;
ii) administering a therapeutic agent for the treatment of prostate cancer if the sample comprises cfDNA derived from a prostate cancer subtype and/or if the sample comprises a level of cfDNA derived from a prostate cancer subtype (for example 0.01% or more cfDNA derived from a prostate cancer subtype).

[0980] § 45. A therapeutic agent for the treatment of prostate cancer for use in the treatment of prostate cancer, wherein

i) the method of any one of § 1 to § 27, § 35 or § 36 is performed to determine if a sample comprises cfDNA derived from a prostate cancer subtype in a subject and/or determine the level of cfDNA in the sample derived from a prostate cancer subtype in a subject;
ii) the therapeutic agent is administered if the sample comprises cfDNA derived from a prostate cancer subtype in the subject and/or if the sample comprises a level of cfDNA derived from a prostate cancer subtype (for example 0.01% or more cfDNA derived from a prostate cancer subtype).

[0981] § 46. A method as described in § 39 to § 44, or a therapeutic agent for the treatment of prostate cancer for use as described in § 45, wherein a second therapeutic agent for the treatment of prostate cancer is administered if a sample from the subject has cfDNA derived from a prostate cancer subtype and/or has a level of cfDNA derived from a prostate cancer subtype (for example a detectable level of prostate cancer DNA, for example 0.01% or more cfDNA derived from a prostate cancer subtype).

[0982] § 47. The method as described in § 44, or a therapeutic agent for the treatment of prostate cancer for use as described in § 45, wherein

(iii) at least 1 week, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 1 month, at least 2 months, at least 3 months, at least 6 months, at least 9 months, at least 12 months, at least 24 months, or at least 36 months, after the administration of the therapeutic agent, a further sample comprising cfDNA is obtained from the subject, and the method of any one of § 1 to § 27, § 35, or § 36 is performed to determine if the further sample comprises cfDNA derived from a prostate cancer subtype in a subject and/or determine the level of cfDNA that is derived from a prostate cancer subtype.

[0983] § 48. A method of determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer comprising

performing the method of any one of § 1 to § 27, § 35 or § 36;
determining the one or more suitable therapeutic agents for the treatment of prostate cancer by reference to whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby one therapeutic agent is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a level of cfDNA derived from a prostate cancer subtype of less than 0.01%, and two or more therapeutic agents are suitable for a subject with a level of cfDNA derived from a prostate cancer subtype (for example a percentage level of cfDNA derived from a prostate cancer subtype of at least 0.01%);
or whereby a therapeutic agent selected from a first list of therapeutic agents is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a level of cfDNA derived from a prostate cancer subtype of less than 0.01%, and a therapeutic agent from a second list of therapeutic agents, or two or more therapeutic agents from the first list, is suitable for a subject with a level of cfDNA derived from a prostate cancer subtype (for example a percentage level of cfDNA derived from a prostate cancer subtype of at least 0.01%).

[0984] § 49. A method of determining a suitable treatment regimen for a subject having prostate cancer comprising

performing the method of any one of claims § 1 to § 27, § 35 or § 36;
determining the treatment regimen by reference whether the sample comprises cfDNA derived from a prostate cancer subtype and/or the level of cfDNA in the sample that is derived from a prostate cancer subtype, whereby a standard treatment is suitable for a subject with a sample having no cfDNA derived from a prostate cancer subtype (for example an undetectable level of cfDNA derived from a prostate cancer subtype) or a percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA sample of less than 0.01%, and a non-standard treatment is suitable for a subject when a level cfDNA derived from a prostate cancer subtype (for example a detectable level of cfDNA derived from a prostate cancer subtype in the cfDNA sample) or a percentage level of cfDNA derived from a prostate cancer subtype in the cfDNA sample is determined of at least 0.01%.

[0985] § 50. The method as claimed in § 49, wherein the standard treatment is a treatment with a therapeutic agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with two or more therapeutic agents for the treatment of prostate cancer;

or wherein the standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a non-standard treatment is a treatment with a hormonal agent for the treatment of prostate cancer, and a chemotherapeutic agent for the treatment of prostate cancer and/or a immunotherapy treatment of prostate cancer and/or a targeted treatment of prostate cancer and/or a biologic agent treatment of prostate cancer.

[0986] § 51. A computerized method and/or computer-assisted method for determining one or more suitable therapeutic agents for the treatment of prostate cancer for a subject having prostate cancer, the method comprising performing the steps of § 48; or for selecting a treatment regimen for a subject having prostate cancer, the method comprising the steps of § 49 or § 50.

[0987] § 52. A method or therapeutic agent as described in any one of § 37 to § 51, wherein the therapeutic agent for the treatment of prostate cancer is selected from the group consisting of a hormonal agent, a targeted agent, a biologic agent, an immunotherapy agent, a chemotherapy agent;

for example: a hormonal agent selected from LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), estrogens, and steroids (for example prednisone or dexamethasone);
a targeted agent selected from poly(ADP-ribose) polymerase (PARP) inhibitor (for example olaparib, rucaparib, niraparib or talazoparib), a epidermal growth factor receptor (EGFR) inhibitor (for example gefitinib, erlotinib, afatinib, brigatinib, icotinib, cetuximab, or osimertinib, adavosertib, lapatinib), and a tyrosine kinase inhibitor (for example imatinib, gefitinib, erlotinib, sunitinib);
a biologic agent selected from monoclonal antibodies (for example pertuzumab, trastuzumab and Solitomab), hormones (for example a hormonal agent selected from LHRH agonists (for example leuprolide, goserelin, triptorelin, or histrelin), LHRH antagonists (for example degarelix), androgen blockers (for example abiraterone or ketoconazole), anti-androgens (for example flutamide, bicalutamide, nilutamide, enzalutamide, apalutamide or darolutamide), and estrogens), interferons (for example interferons-α, -β, -γ), and interleukin-based products (for example interleukin-2);
an immunotherapy agent selected from a cancer vaccine (for example sipuleucel-T), T-cell therapy, monoclonal antibody therapy, immune checkpoint therapy (for example a PD-1 inhibitor (e.g pembrolizumab, nivolumab, cemiplimab spartalizumab), a PD-L1 inhibitor (e.g. atezolizumab, avelumab or durvalumab), or a CTLA-4 (e.g. ipilimumab)), and non-specific immunotherapies (for example interferons and inerleukins);
a chemotherapy agent selected from docetaxel, cabazitaxel, and c-Met inhibitors (for example cabozantinib).