Clonal haematopoiesis
11613786 · 2023-03-28
Assignee
Inventors
Cpc classification
C12Q1/6883
CHEMISTRY; METALLURGY
International classification
Abstract
The present invention relates to clonal expansion of somatic cells in subjects, and acquired selective advantage of cell clones during the lifetime of a subject. In particular, the invention relates to methods for predicting the development of cancer based on the observation of specific genetic mutations in somatic cell clones, as well as to methods for treating or preventing cancer in a subject, in which clonal expansion of cells comprising specific modifications is observed.
Claims
1. A method of detecting DNMT3A mis-sense mutations in hematopoietic stem cells (HSCs), the method comprising the steps of: (a) obtaining a blood sample from a human subject (b) isolating HSCs from the blood sample; (c) sequencing DNMT3A nucleic acids from one or more of the isolated HSCs; (d) detecting the presence of a mis-sense mutation in the sequenced DNMT3A nucleic acids, wherein the mis-sense mutation is G543C, F732C, Y735C, R749C, F751C, W753C, or L889C; and (e) reducing the incidence of HSCs comprising said mutations in the subject and administering to the subject HSCs in which the mutations are absent.
2. The method of claim 1, wherein the subject is at least 50 years of age.
3. The method of claim 1, wherein the subject is or has been exposed to a human carcinogen in sufficient amount and/or frequency for such carcinogen to be a potential cause of hematological malignancy.
4. The method of claim 3, wherein the carcinogen comprises a tobacco product, an organic solvent, a virus, a compound found in grilled red meat, ionizing radiation, a heavy metal or compound thereof, or any combination thereof.
5. The method of claim 1, wherein the HSCs in which the mutations are absent are administered in a bone marrow transplantation.
6. The method of claim 1, wherein the HSCs in which the mutations are absent are administered in a blood transfusion.
7. A method of determining whether a subject is predisposed to a hematological malignancy, the method comprising the steps of: (a) obtaining a blood sample from a human subject (b) isolating hematopoietic stem cells (HSCs) from the blood sample; (c) sequencing DNMT3A nucleic acids from one or more of the isolated HSCs; (d) detecting the presence of a mis-sense mutation in the sequenced DNMT3A nucleic acids, wherein the mis-sense mutation is G543C, F732C, Y735C, R749C, F751C, W753C, or L889C; (e) determining that the subject is predisposed to a hematological malignancy if the mis-sense mutation is detected; and (f) reducing the incidence of HSCs comprising said mutations in the subject and administering to the subject HSCs in which the mutations are absent.
8. The method of claim 7, wherein the hematological malignancy is a myeloproliferative neoplasm, a myelodysplastic syndrome, acute myeloid leukemia or chronic lymphocytic leukemia.
9. The method of claim 7, wherein the human subject is at least 50 years of age.
10. The method of claim 7, wherein the human subject has been exposed to a human carcinogen in sufficient amount and/or frequency for such carcinogen to be a potential cause of hematological malignancy.
11. The method of claim 10, wherein the carcinogen comprises a tobacco product, an organic solvent, a virus, a compound found in grilled red meat, ionizing radiation, a heavy metal or compound thereof, or any combination thereof.
12. The method of claim 7, wherein the HSCs in which the mutations are absent are administered in a bone marrow transplantation.
13. The method of claim 7, wherein the HSCs in which the mutations are absent are administered in a blood transfusion.
14. A method of selecting hematopoietic cells for transplantation, the method comprising the steps of: (a) obtaining a sample comprising hematopoietic cells from a human subject; (b) sequencing DNMT3A nucleic acids from one or more of the hematopoietic cells; (c) detecting the absence of G543C, F732C, Y735C, R749C, F751C, W753C, and L889C mis-sense mutations in the sequenced DNMT3A nucleic acids; and (d) collecting the remaining hematopoietic cells in the sample for transplantation.
15. The method of claim 14, wherein the hematopoietic cells are collected for autologous transplantation.
16. The method of claim 14, wherein the hematopoietic cells are collected for allogenic transplantation.
17. The method of claim 14, wherein the sample is obtained from the human subject prior to diagnosis of a hematological malignancy.
18. The method of 14, wherein the sample is derived from cord blood.
Description
BRIEF DESCRIPTION OF THE FIGURES
(1) The following detailed description, given by way of example, but not intended to limit the invention solely to the specific embodiments described, may best be understood in conjunction with the accompanying drawings.
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
DETAILED DESCRIPTION OF THE INVENTION
(22) The development of disease often involves dynamic processes that begin years or decades before disease onset. However, the process of pathogenesis often goes undetected until after the patient develops symptoms and presents with advanced disease.
(23) Cancer arises due to the combined effects of multiple somatic mutations, which are likely to be acquired at different times (Nowell, P. C. The clonal evolution of tumor cell populations. Science 194, 23-28 (1976)). Early mutations may be present in an individual's body many years before disease develops. In some models of cancer development, early mutations lead to clonal expansions by stem cells or other progenitor cells (Reya, T., Morrison, S. J., Clarke, M. F. & Weissman, I. L. Stem cells, cancer, and cancer stem cells. Nature 414, 105-111 (2001)). Such clonal expansions might create a favourable context for the selection of later, cooperating mutations while simultaneously increasing the likelihood that later mutations will affect cells that already contain the earlier, initiating mutations. To understand the pathogenesis of proliferative diseases, it is important to know the extent to which clonal expansions occur and precede malignancies.
(24) Several lines of evidence suggests that haematopoietic stem cell (HSC) population dynamics may precede many haematological malignancies including myeloproliferative neoplasms (Jamieson, C. H. M. et al. The JAK2 V617F mutation occurs in hematopoietic stem cells in polycythemia vera and predisposes toward erythroid differentiation. Proc. Natl. Acad. Sci. U.S.A 103, 6224-6229 (2006)), myelodysplastic syndromes (Jaiswal, S. & Ebert, B. L. MDS Is a Stem Cell Disorder After All. Cancer Cell 25, 713-714 (2014)), acute myeloid leukaemia (AML) (Potter, N. E. & Greaves, M. Cancer: Persistence of leukaemic ancestors. Nature 506, 300-301 (2014); Vasanthakumar, A. & Godley, L. A. On the origin of leukemic species. Cell Stem Cell 14, 421-422 (2014)), and chronic lymphocytic leukemia (Damm, F. et al. Acquired initiating mutations in early hematopoietic cells of CLL patients. Cancer Discov. (2014). doi:10.1158/2159-8290.CD-14-0104). For example, in some patients, stem cells carrying a subset of the mutations present in the cancer cells are able to survive chemotherapy; subsequently, these cells acquire novel mutations, triggering relapse (Ding, L. et al. Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing. Nature 481, 506-510 (2012); Shlush, L. I. et al. Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia. Nature 506, 328-333 (2014); Corces-Zimmerman, M. R., Hong, W.-J., Weissman, I. L., Medeiros, B. C. & Majeti, R. Preleukemic mutations in human acute myeloid leukemia affect epigenetic regulators and persist in remission. Proc. Natl. Acad. Sci. U.S.A 111, 2548-2553 (2014)).
(25) Clonal mosaicism for large chromosomal abnormalities, reflecting expansion of a specific cellular clone, appears to arise in about 2% of healthy aging individuals and is a risk factor for later haematopoietic cancers (Laurie, C. C. et al. Detectable clonal mosaicism from birth to old age and its relationship to cancer. Nat. Genet. 44, 642-650 (2012); Jacobs, K. B. et al. Detectable clonal mosaicism and its relationship to aging and cancer. Nat. Genet. 44, 651-658 (2012); Schick, U. M. et al. Confirmation of the reported association of clonal chromosomal mosaicism with an increased risk of incident hematologic cancer. PloS One 8, e59823 (2013)). In principle, clonal expansion among HSCs—a phenomenon termed clonal haematopoiesis—could be much more common, if only a minority of cases are accompanied by large chromosomal abnormalities (similarly to AML14).
(26) Many studies today sequence blood-derived DNA from thousands of individuals to identify inherited risk factors for common diseases. The inventors reasoned that such data offered the opportunity to test the hypothesis that clonal haematopoiesis may be common and associate with subsequent cancer and mortality in its common form, and to identify the genes in which mutations drive clonal expansions.
(27) The inventors therefore analysed the exome sequences from 12,380 individuals and identified 3,111 putative somatic mutations based on their presence at unusual allelic fractions, corresponding to an average of approximately one putative somatic mutation for every four subjects. For 65 of 65 mutations tested, molecular validation confirmed that the mutant allele was present at a low allelic fraction (significantly less than 50%) and thus could not have been inherited.
(28) The inventors have found that clonal haematopoiesis with somatic mutations affects at least 10% of the elderly and increases in frequency with advancing age (
(29) The method of the invention involves analysis of at least part of the genome of a sample from a subject. The sample can contain one more cells, which for example can be haematopoietic stem cells (HSCs), committed myeloid progenitor cells having long term self-renewal capacity or mature lymphoid cells having long term self-renewal capacity.
(30) In some embodiments the part of the genome that is sequenced may be limited to specific genes, the whole exome or parts of an exome. For example, the sequencing may be whole exome sequencing (WES).
(31) In an advantageous embodiment, the subject is a human. In another advantageous embodiment, the human may be at least 50 years of age. In other embodiments, the human may exhibit one or more risk factors of being a smoker, undergoing therapy for cancer, or having been exposed to a solvent as defined herein.
(32) Most clonal haematopoiesis appears to be driven by mutations in a specific subset of the genes recognized as drivers of blood malignancies (Shih, A. H., Abdel-Wahab, O., Patel, J. P. & Levine, R. L. The role of mutations in epigenetic regulators in myeloid malignancies. Nat. Rev. Cancer 12, 599-612 (2012)), such as DNMT3A, ASXL1, and TET2 (
(33) Many if not most haematological malignancies appear to be preceded by an extended period during which a haematopoietic clone with somatic mutations could be detected simply by sequencing the DNA in peripheral blood. Such clones were detected in 42% of the subjects who were diagnosed with malignancies 6-36 months later (
(34) Appropriate perspective should be exercised when cancer-associated mutations are observed as an incidental finding in other studies or diagnostic tests: our results suggest that such findings may be common and do not justify a diagnosis of haematological malignancy. However, the present data show that defined somatic mutations are associated with particularly elevated risk; studies of large numbers of elderly individuals identify those somatic mutations with greatest likelihood of subsequent malignancy.
(35) As used herein, an increased likelihood of progression means that the subject is more likely, in embodiments is statistically more likely, to develop cancer than a subject in which the mutations referred to herein have not been detected. For example, the subject has a higher likelihood of developing cancer when expressed as a percentage of subjects who develop cancer, as opposed to those who do not, within a defined time period. A defined time period can be from as little as six months or less, to 1, 1.5, 2, 2.5, 3, 4, 5, 6, 7, 8, 9, 10 years or more.
(36) Statistical significance can mean that the associated p-value is 0.05 or less.
(37) In embodiments, the increase in likelihood can be expressed as the increase in likelihood over one year. For example, the increase in likelihood can be 0.25%, 0.5%, 0.75%, 1%, 1.25%, 1.5%, 1.75%, 2% or more over one year.
(38) Progression to a cancerous state denotes the development of a novel cancer, or malignancy, in the subject. A cancer is, in embodiments, a haematological malignancy. Examples of such cancers include myeloproliferative neoplasms (MPN), myelodysplastic syndromes (MDS) and chronic myelomonocytic leukemia (CMML), as well as acute stages, i.e. acute myeloid leukemia (AML). MPNs can comprise a variety of disorders such as chronic myeloid leukemia (CML) and non-CML MPNs such as polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF).
(39) The invention requires sequencing of at least part of the genome of a subject. Sequencing can be carried out according to any suitable technique, many of which are generally known in the art. Many proprietary sequencing systems are available commercially and can be used in the context of the present invention, such as for example from Illumina, USA. Single-cell sequencing methods are known in the art, as noted for example by Eberwine et al., Nature Methods 11, 25-27 (2014) doi:10.1038/nmeth.2769 Published online 30 Dec. 2013; and especially single-cell sequencing in microfluidic droplets (Nature 510, 363-369 (2014) doi:10.1038/nature13437).
(40) Sequencing can be of specific genes only, specific parts of the genome, or the whole genome. Where specific genes are sequenced, the gene(s) sequences are preferably selected from the group consisting of DNMT3A, ASXL1, TET2, PPM1D and JAK2. In embodiments, specific parts of genes can be sequenced; for example in DNMT3A, exons 7 to 23 can be sequenced. In embodiments, specific mutations can be interrogated, such as the JAK2 mutation V617F. Additionally, or alternatively, specific mutations can be avoided, such as ASXL1 p.G646fsX12 and p.G645fsX58.
(41) Where a part of a genome is sequenced, that part can be the exome. The exome is the part of the genome formed by exons, and thus an exon sequencing method sequences the expressed sequences in the genome. There are 180,000 exons in the human genome, which constitute about 1% of the genome, or approximately 30 million base pairs. Exome sequencing requires enrichment of sequencing targets for exome sequences; several techniques can be used, including PCR, molecular inversion probes, hybrid capture of targets, and solution capture of targets. Sequencing of targets can be conducted by any suitable technique.
(42) Mutations in genes can be disruptive, in that they have an observed or predicted effect on protein function, or non-disruptive. A non-disruptive mutation is typically a missense mutation, in which a codon is altered such that it codes for a different amino acid, but the encoded protein is still expressed.
(43) DNMT3A is DNA (cytosine-5-)-methyltransferase 3 alpha and is encoded on chromosome 2. See Human Genome Nomenclature Committee reference HGMC 2978.
(44) ASXL1 is additional sex combs like transcriptional regulator 1 and is encoded on chromosome 20. See Human Genome Nomenclature Committee reference HGMC 18318.
(45) TET2 is tet methylcytosine dioxygenase 2 and is encoded on chromosome 4. See Human Genome Nomenclature Committee reference HGMC 25941.
(46) PPM1D is protein phosphatase, Mg2+/Mn2+ dependent, 1D and is encoded on chromosome 17. See Human Genome Nomenclature Committee reference HGMC 9277.
(47) JAK2 is janus kinase 2 and is encoded on chromosome 9. See Human Genome Nomenclature Committee reference HGMC 6192.
(48) In the context of the present invention, a “treatment” is a procedure which alleviates or reduces the negative consequences of cancer on a patient. Many cancer treatments are known in the art, and some are set forth herein. Any treatments or potential treatments can be used in the context of the present invention.
(49) A treatment is not necessarily curative, and may reduce the effect of a cancer by a certain percentage over an untreated cancer. For example, the number of cancerous cells in a subject may be diminished by the treatment, or the overall mass of cancer tissue may be diminished.
(50) The percentage reduction or diminution can be from 10% up to 20, 30, 40, 50, 60, 70, 80, 90, 95, 99 or 100%.
(51) Methods of treatment may be personalised medicine procedures, in which the DNA of an individual is analysed to provide guidance on the appropriate therapy for that specific individual. The methods of the invention may provide guidance as to whether treatment is necessary, as well as revealing progress of the treatment and guiding the requirement for further treatment of the individual.
(52) Sequencing of DNA can be performed on tissues or cells. Sequencing of specific cell types (for example, haematopoietic cells obtained by flow sorting) can identify mutations in specific cell types that provide specific predictive value. Some cell types may provide a greater predictive value than other cell types.
(53) Sequencing can also be conducted in single cells, using appropriate single-cell sequencing strategies. Single-cell analyses can be used to identify high-risk combinations of mutations co-occurring in the same cells. Co-occurrence signifies that the mutations are occurring in the same cell clone and carry a greater risk, and therefore have a greater predictive value, that occurrence of the same mutations in different individual cells.
(54) In certain embodiments, the mutations identified in the subject can be checked against databases of mutations which are associated with cancer. One such database is the Catalogue of Somatic Mutations in Cancer (COSMIC); Forbes, S. A. et al. COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer; Nucleic Acids Res. 39, D945-950 (2011). In particular, version 69 of the COSMIC database is referred to.
(55) If analysis of the sample from the subject does not reveal the presence of any of the specific mutations identified herein as indicative of increase risk of development of cancer, or of the presence of clonal haematopoiesis, the sample can be further analysed for the presence of somatic mutations. The presence of a plurality of somatic mutations, at a level above that normally expected for a random mutation, is deemed to suggest the presence of clones. The threshold level for indicating the presence of clones is 0.1 mutations per megabase of sequenced DNA.
(56) Typical analytical pipelines for identifying somatic mutations in cancer seek to identify mutations that are present in tumor tissue but absent from paired normal tissue from the same individual (Cibulskis, K. et al. Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples. Nat. Biotechnol. 31, 213-219 (2013)). Because the analysis provided herein uses a single DNA sample from each subject, a novel strategy is provided for identifying somatic mutations based on allelic fractions.
(57) Assuming that a somatic mutation will be present in only a subset of the cells contributing DNA to analysis, the mutant allele will be present in fewer than 50% of the sequence reads arising from that genomic site (
(58) Certain sequences, such as those with high GC content, repetitive elements and/or low sequence complexity are prone to sequencing errors and false positive creation due to artifacts caused by enzyme slippage and other reading errors. Hence, care must be taken to ensure that any sequence changes observed in these regions are real and not artifact. Due to the higher likelihood of misalignment and PCR artefacts, somatic mutations in the following regions are optionally excluded from analysis: 1) Low complexity regions and sites harboring markers failing Hardy Weinberg equilibrium tests in the 1000 Genomes Project phase 1 (Li, H. Toward better understanding of artifacts in variant calling from high-coverage samples. Bioinforma. Oxf Engl. (2014). doi:10.1093/bioinformatics/btu356); (see Github files: lh3/varcmp/blob/master/scripts/LCR-hs37d5.bed.gz and lh3/varcmp/blob/master/scripts/1000g.hwe-bad.bed); 2) Sites with excess coverage within the 1000 Genomes Project phase 1 (Genovese, G., Handsaker, R. E., Li, H., Kenny, E. E. & McCarroll, S. A. Mapping the Human Reference Genome's Missing Sequence by Three-Way Admixture in Latino Genomes. Am. J. Hum. Genet. 93, 411-421 (2013)); 3) Segmental duplications of the human genome (Bailey, J. A., Yavor, A. M., Massa, H. F., Trask, B. J. & Eichler, E. E. Segmental duplications: organization and impact within the current human genome project assembly. Genome Res. 11, 1005-1017 (2001); Bailey, J. A. et al. Recent segmental duplications in the human genome. Science 297, 1003-1007 (2002)) (see UCSC Genome Browser hg19 database: /goldenPath/hg19/database/genomicSuperDups.txt.gz); 4) Regions excluded from the strict mask of the 1000 Genomes Project phase 1(1000 Genomes Project Consortium et al. An integrated map of genetic variation from 1,092 human genomes. Nature 491, 56-65 (2012)) (see 1000 Genomes project data hosted by the European Bioinformatics Institute FTP site: vol1/ftp/phase1/analysis_results/supporting/accessible_genome_masks/20120824_strict_mas k.bed).
(59) These filters defined regions covering ˜60% of the GRCh37 human genome reference and ˜70% of the coding regions and they excluded 161,158 out of the 1,812,331 variants called in the cohort described in the Examples.
(60) Due to enrichment bias in exome libraries, allelic fractions for inherited heterozygous mutations are not expected to be centered around 50%. The average expected allelic fraction for the alternate allele of a heterozygous single nucleotide polymorphisms (SNPs) is actually 47%±4% (
(61) Putative somatic mutations include but are not limited to those alleles satisfying the following criteria: 1) non-silent/disruptive nucleotide changes, indels, missense mutations, frameshifts, stop mutations (addition or deletion), read-through mutations, splice mutations; 2) confirmed change not due to a sequencing error or artifact of the testing system.
(62) In embodiments, the mutation is a putative somatic mutation if: a) the mutation is a SNV; b) the mutation results in a disruptive change in the encoded polypeptide or regulation of the gene; c) the mutation has an allelic fraction above 10%; and d) the mutation includes changes in regions other than those identified as being prone to errors and artifacts, including but not limited to, e.g., low sequence complexity, high GC content, repetitive elements, and the like.
(63) Inclusive somatic mutations are defined as those alleles satisfying the following criteria: 1) SNVs or indels of length one or two base pairs or more; 2) disruptive mutation; 3) allelic fraction above 5%; and 4) not a false positive.
(64) In the context of the cohort of patients analyzed herein, samples are classified as indicative of likelihood of clonal haematopoiesis and/or progress towards a cancerous state if the sample comprises at least one of: (i) 3 or more exomic putative somatic mutations; (ii) 0.1 putative somatic mutation per megabase of sequenced DNA; or (iii) 50 putative somatic mutations per genome.
(65) Subjects which are positive as assessed by somatic mutation analysis considered at increased risk of developing cancer and/or having a higher proportion of haematopoietic clones, as for the foregoing subjects which are judged positive on the basis of specific gene mutations.
(66) In embodiments of the present invention, the analysis of the genomes of single cells by single cell sequencing can be used to provide information about the relationship between mutations and cell types. For example, the presence of a mutation in multiple cells of a defined cell type can further strengthen the conclusion that the mutation is clonal. Moreover, the presence of more than one mutation in a single cell can be evidence of clonal expansion, if the mutations are repeated ly found together,
(67) The presence of multiple somatic mutations, as set forth above, can be an indicator of clonal hematopoiesis even in the absence of the presence of driver mutations, for instance the driver mutations identified herein. Accordingly, sequencing in accordance with the present invention can comprise sequencing of genome, exome or specific genes in pooled cells from a sample, such as a blood sample, to identify the presence of driver mutations and/or putative somatic mutations. Alternatively, or in addition, sequencing can comprise the sequencing of the genome, exome or specific genes of one or more single cells, in order to identify the presence of mutations in genes in specific cell types. Initial screens can comprise sequencing to identify driver mutations in a sample, or the presence of putative somatic mutations in a sample. Samples testing positive can be followed dup by single cell sequencing to identify the cell types which harbor the specific mutations, and the identity of mutations which occur together in a single cell.
(68) Subjects can accordingly be subjected to treatment for cancer conditions, including wherein for example (a) the treatment comprises treating said subject by reducing the incidence of haematopoietic clones comprising said mutation in the subject's blood or (b) the treatment or monitoring includes repeating the method as to the subject monthly, bi-monthly or quarterly and treating said subject by reducing the incidence of haematopoietic clones comprising said mutation in the subject's blood or (c) the treatment or monitoring comprises including the subject as a candidate to receive a bone marrow transplant, or (d) the treatment or monitoring includes administering to the subject a bone marrow transplant, or (e) transfusing the subject with blood in which said mutations are absent.
(69) Blood in which mutations are absent can be autologous blood, derived from blood samples taken from the same patient at an earlier point in time; including for example cord blood. Alternatively, or in addition, the blood in which mutations are absent can be allogenic blood, derived from an individual in which the mutations are absent.
(70) In embodiments, a bone marrow transplant can be effected.
(71) Initial detection of clonal haematopoiesis can justify more frequent screening to detect the presence of cooperating mutations at low allele frequencies that presage cancer.
(72) In addition, the use of DNA sequencing to ascertain at-risk cohorts and monitor clonal expansions, as reported here, will facilitate clinical trials of prevention strategies to reduce progression to malignancy.
(73) Clonal haematopoiesis is also a marker for declining health of HSC populations, potentially reflecting aging, attrition, and a declining ability to contain novel neoplasms.
(74) The subject, as referred to herein, is preferably a mammal and advantageously a human. It has been observed that clonal expansion can be determined with rapidly increasing frequency in human subjects of 50 years of age or more. This is in contrast with methods of the prior art, in which clonal expansion is only apparently significant in subjects of greater age, such as 60 or 70 years of age.
(75) Accordingly, there is defined a population of subjects which are tested using any of the methods set forth in accordance with the present invention, wherein that population of subjects comprises humans of at least 50 years of age.
(76) Further populations of subjects which are tested using any of the methods of the present invention comprise subjects undergoing cancer therapy, such as chemotherapy or radiotherapy; these therapeutic approaches increase the risk of developing haematopoietic malignancies and the promotion of haematopoietic clones.
(77) Other populations include subjects which have been exposed to a carcinogen, such as for example tobacco products and/or organic solvents such as textile dyes, paints or inks, and/or red meat, for example grilled, fried or roasted red meat, a virus, ionising radiation or a heavy metal compound such as a lead compound.
(78) There are many methods known in the art for determining the genotype of a patient and for identifying or analyzing whether a given DNA sample contains a particular somatic mutation. Any method for determining genotype can be used for determining genotypes in the present invention. Such methods include, but are not limited to, amplimer sequencing, DNA sequencing, fluorescence spectroscopy, fluorescence resonance energy transfer (or “FRET”)-based hybridization analysis, high throughput screening, mass spectroscopy, nucleic acid hybridization, polymerase chain reaction (PCR), RFLP analysis and size chromatography (e.g., capillary or gel chromatography), all of which are well known to one of skill in the art.
(79) The methods of the present invention, such as whole exome sequencing and targeted amplicon sequencing, have commercial applications in diagnostic kits for the detection of the somatic mutations in patients. A test kit according to the invention may comprise any of the materials necessary for whole exome sequencing and targeted amplicon sequencing, for example, according to the invention. In a particular advantageous embodiment, a diagnostic for the present invention may comprise testing for any of the genes in disclosed herein. The kit further comprises additional means, such as reagents, for detecting or measuring the sequences of the present invention, and also ideally a positive and negative control.
(80) The present invention further encompasses probes that are immobilized on a solid or flexible support, such as paper, nylon or other type of membrane, filter, chip, glass slide, microchips, microbeads, or any other such matrix, all of which are within the scope of this invention. The probe of this form is now called a “DNA chip”. These DNA chips can be used for analyzing the somatic mutations of the present invention. The present invention further encompasses arrays or microarrays of nucleic acid molecules that are based on one or more of the sequences described herein. As used herein “arrays” or “microarrays” refers to an array of distinct polynucleotides or oligonucleotides synthesized on a solid or flexible support, such as paper, nylon or other type of membrane, filter, chip, glass slide, or any other suitable solid support. In one embodiment, the microarray is prepared and used according to the methods and devices described in U.S. Pat. Nos. 5,446,603; 5,545,531; 5,807,522; 5,837,832; 5,874,219; 6,114,122; 6,238,910; 6,365,418; 6,410,229; 6,420,114; 6,432,696; 6,475,808 and 6,489,159 and PCT Publication No. WO 01/45843 A2, the disclosures of which are incorporated by reference in their entireties.
(81) For the purposes of the present invention, sequence identity or homology is determined by comparing the sequences when aligned so as to maximize overlap and identity while minimizing sequence gaps. In particular, sequence identity may be determined using any of a number of mathematical algorithms. A nonlimiting example of a mathematical algorithm used for comparison of two sequences is the algorithm of Karlin & Altschul, Proc. Natl. Acad. Sci. USA 1990; 87: 2264-2268, modified as in Karlin & Altschul, Proc. Natl. Acad. Sci. USA 1993; 90: 5873-5877.
(82) Another example of a mathematical algorithm used for comparison of sequences is the algorithm of Myers & Miller, CABIOS 1988; 4: 11-17. Such an algorithm is incorporated into the ALIGN program (version 2.0) which is part of the GCG sequence alignment software package. When utilizing the ALIGN program for comparing amino acid sequences, a PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4 can be used. Yet another useful algorithm for identifying regions of local sequence similarity and alignment is the FASTA algorithm as described in Pearson & Lipman, Proc. Natl. Acad. Sci. USA 1988; 85: 2444-2448.
(83) Advantageous for use according to the present invention is the WU-BLAST (Washington University BLAST) version 2.0 software. WU-BLAST version 2.0 executable programs for several UNIX platforms can be downloaded from the FTP site for Blast at the Washington University in St. Louis website. This program is based on WU-BLAST version 1.4, which in turn is based on the public domain NCBI-BLAST version 1.4 (Altschul & Gish, 1996, Local alignment statistics, Doolittle ed., Methods in Enzymology 266: 460-480; Altschul et al., Journal of Molecular Biology 1990; 215: 403-410; Gish & States, 1993; Nature Genetics 3: 266-272; Karlin & Altschul, 1993; Proc. Natl. Acad. Sci. USA 90: 5873-5877; all of which are incorporated by reference herein).
(84) In all search programs in the suite the gapped alignment routines are integral to the database search itself. Gapping can be turned off if desired. The default penalty (Q) for a gap of length one is Q=9 for proteins and BLASTP, and Q=10 for BLASTN, but may be changed to any integer. The default per-residue penalty for extending a gap (R) is R=2 for proteins and BLASTP, and R=10 for BLASTN, but may be changed to any integer. Any combination of values for Q and R can be used in order to align sequences so as to maximize overlap and identity while minimizing sequence gaps. The default amino acid comparison matrix is BLOSUM62, but other amino acid comparison matrices such as PAM can be utilized.
(85) Alternatively or additionally, the term “homology” or “identity”, for instance, with respect to a nucleotide or amino acid sequence, can indicate a quantitative measure of homology between two sequences. The percent sequence homology can be calculated as (Nref−Ndif)*100/−Nref, wherein Ndif is the total number of non-identical residues in the two sequences when aligned and wherein Nref is the number of residues in one of the sequences. Hence, the DNA sequence AGTCAGTC will have a sequence identity of 75% with the sequence AATCAATC (N Nref=8; N Ndif=2). “Homology” or “identity” can refer to the number of positions with identical nucleotides or amino acids divided by the number of nucleotides or amino acids in the shorter of the two sequences wherein alignment of the two sequences can be determined in accordance with the Wilbur and Lipman algorithm (Wilbur & Lipman, Proc Natl Acad Sci USA 1983; 80:726, incorporated herein by reference), for instance, using a window size of 20 nucleotides, a word length of 4 nucleotides, and a gap penalty of 4, and computer-assisted analysis and interpretation of the sequence data including alignment can be conveniently performed using commercially available programs (e.g., Intelligenetics™ Suite, Intelligenetics Inc. CA). When RNA sequences are said to be similar, or have a degree of sequence identity or homology with DNA sequences, thymidine (T) in the DNA sequence is considered equal to uracil (U) in the RNA sequence. Thus, RNA sequences are within the scope of the invention and can be derived from DNA sequences, by thymidine (T) in the DNA sequence being considered equal to uracil (U) in RNA sequences. Without undue experimentation, the skilled artisan can consult with many other programs or references for determining percent homology.
(86) The invention further encompasses kits useful for screening nucleic acids isolated from one or more patients for any of the somatic mutations described herein and instructions for using the oligonucleotide to detect variation in the nucleotide corresponding to one or more of the somatic mutations, such as but not limited to, one or more genes selected from the group consisting of DNMT3A, TET2, ASXL1, PPM1D and JAK2 of the isolated nucleic acid.
(87) In other embodiments of this invention, the step of assaying is selected from the group consisting of: restriction fragment length polymorphism (RFLP) analysis, minisequencing, MALDI-TOF, SINE, heteroduplex analysis, single strand conformational polymorphism (SSCP), denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE).
(88) The present invention also encompasses a transgenic mouse which may express one or more of the herein disclosed somatic mutations. Methods for making a transgenic mouse are well known to one of skill in the art, see e.g., U.S. Pat. Nos. 7,709,695; 7,667,090; 7,655,700; 7,626,076; 7,566,812; 7,544,855; 7,538,258; 7,495,147; 7,479,579; 7,449,615; 7,432,414; 7,393,994; 7,371,920; 7,358,416; 7,276,644; 7,265,259; 7,220,892; 7,214,850; 7,186,882; 7,119,249; 7,112,715; 7,098,376; 7,045,678; 7,038,105; 6,750,375; 6,717,031; 6,710,226; 6,689,937; 6,657,104; 6,649,811; 6,613,958; 6,610,905; 6,593,512; 6,576,812; 6,531,645; 6,515,197; 6,452,065; 6,372,958; 6,372,957; 6,369,295; 6,323,391; 6,323,390; 6,316,693; 6,313,373; 6,300,540; 6,255,555; 6,245,963; 6,215,040; 6,211,428; 6,201,166; 6,187,992; 6,184,435; 6,175,057; 6,156,727; 6,137,029; 6,127,598; 6,037,521; 6,025,539; 6,002,067; 5,981,829; 5,936,138; 5,917,124; 5,907,078; 5,894,078; 5,850,004; 5,850,001; 5,847,257; 5,837,875; 5,824,840; 5,824,838; 5,814,716; 5,811,633; 5,723,719; 5,720,936; 5,688,692; 5,631,407; 5,620,881; 5,574,206 and 5,569,827. The transgenic mouse may be utilized to mimic haematopoietic disease conditions and may be useful to test novel treatments for blood cancer diseases disease in a mouse model.
(89) Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined in the appended claims.
(90) The present invention will be further illustrated in the following Examples which are given for illustration purposes only and are not intended to limit the invention in any way.
EXAMPLES
Example 1: Identification of Characteristic Mutations and Candidate Drivers in Clonal Haematopoiesis
(91) The exome sequences from 12,380 individuals were analysed and 3,111 putative somatic mutations identified based on their presence at unusual allelic fractions corresponding to an average of approximately one putative somatic mutation for every four subjects.
(92) In detail, a total of 12,380 Swedish research participants with psychiatric diagnoses (Table S1) were ascertained from the Swedish National Hospital Discharge Register, which captures all inpatient hospitalizations. Controls were randomly selected from population registers. We treated cases and controls as a single cohort for all analyses presented below, as none of the mutational variables analyzed below showed any relationship to psychiatric diagnosis after controlling for other factors such as age and smoking.
(93) Excluding bipolar subjects, medical histories (from 1965 to 2011) of 11,164 of the subjects enrolled in the study were extracted from the Swedish national in- and outpatient register (median follow-up was 32 months). Information about vital status (from 2006 to 2012) was extracted from the population register and the Cause of Death register (median follow-up was 42 months). To identify individuals with haematologic malignancies, we included diagnoses within ICD10 code groups C81-C96 (malignant neoplasms of lymphoid, haematopoietic and related tissue), D45 (polycythemia vera), D46(myelodysplastic syndromes), D47 (other neoplasms of uncertain behavior of lymphoid, haematopoietic and related tissue), and D7581 (myelofibrosis) and the same diagnoses within the corresponding ICD9 and ICD8 groups.
(94) Sequencing data were aligned against the GRCh37 human genome reference using BWA ALN version 0.5.9. (Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754-1760 (2009)) On average across samples each base pair of the target intervals was observed 95 times.
(95) Genotypes and allelic counts were computed across the genome using the Haplotype Caller from the Genome Analysis Toolkit version 3.1-1 (McKenna, A. et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 20, 1297 (2010)), which generated genotypes for 1,812,331 variant sites across 12,380 subjects. Due to the specific default parameters used by the Haplotype Caller and aimed at genotyping inherited mutations, we recognized that several mutations present in sequencing reads in the 5-10% allele fraction range, and that could have been called, were not reported. To mitigate this issue, we used the Unified Genotyper from the Genome Analysis Toolkit to genotype 208 variants reported as seen seven or more times in haematopoietic or lymphoid cancers in the Catalogue Of Somatic Mutations In Cancer (COSMIC) database3 v69 (released Jun. 2, 2014), with the exception of a few that we deemed inherited mutations or PCR sequencing artifacts rather than somatic events (Table S2). We kept all mutations for which the alternate allele was observed on at least three sequencing reads in an individual's sequencing data. These thresholds yielded 26 additional mutations that were not called by the Haplotype Caller. We did not use these mutations for our unbiased analysis of enrichment of disruptive mutations.
(96) Definition of Putative, Inclusive, and Candidate Driver Somatic Mutations
(97) Due to the higher likelihood of misalignment and PCR artifacts, we excluded from analysis somatic mutations in the following regions: 1) Low complexity regions and sites harboring markers failing Hardy Weinberg equilibrium tests in the 1000 Genomes Project phase 1: see Github files: lh3/varcmp/blob/master/scripts/LCR-hs37d5. bed.gz and https://github.com/lh3/varcmp/blob/master/scripts/1000g.hwe-bad.bed) 2) Sites with excess coverage within the 1000 Genomes Project phase 1 3) Segmental duplications of the human genome (see UCSC Genome Browser hg19 database: goldenPath/hg19/database/genomicSuperDups.txt.gz) 4) Regions excluded from the strict mask of the 1000 Genomes Project phase 1 (see 1000 Genomes Project data hosted by the European Bioinformatics Institute FTP site: vol1/ftp/phase1/analysis_results/supporting/accessible_genome_masks/20120 824_strict_mask.bed)
(98) These filters defined regions covering ˜60% of the GRCh37 human genome reference and ˜70% of the coding regions and they excluded 161,158 out of the 1,812,331 variants called in the cohort.
(99) Due to enrichment bias in exome libraries, allelic fractions for inherited heterozygous mutations are not expected to be centered around 50%. The average expected allelic fraction for the alternate allele of a heterozygous single nucleotide polymorphisms (SNPs) is actually 47%±4% (
(100) For this cohort, we define as putative somatic mutations those alleles satisfying the following criteria: 1) SNVs 2) Observed once or twice (minor allele frequency less than 0.01%) in this cohort 3) Allelic fraction above 10% 4) Failed the hypothesis that the alternate allelic count was distributed as a binomial process with mean 45% with a designed false positive rate of 10.sup.−5
(101) We define as inclusive somatic mutations those alleles satisfying the following criteria: 1) SNVs or indels of length one or two base pairs 2) Observed at most six times (minor allele frequency less than 0.025%) in the cohort 3) Allelic fraction above 5% 4) Failed the hypothesis that the alternate allelic count was distributed as a binomial process with mean 47% for SNVs and 40% for indels with a designed false positive rate of 0.01.
(102) These definitions yielded 4,275 putative somatic mutations and 53,474 inclusive somatic mutations across 12,380 subjects. Upon further analysis, a large fraction of these mutations originated from the first two sequencing waves (
(103) We excluded the 534 subjects from the first two waves and the outlier subject from any subsequent analyses in which putative or inclusive somatic mutations were used. This resulted in a refined set of 3,111 putative somatic mutations and 42,282 inclusive somatic mutations from 11,845 subjects.
(104) Mutational profiles for inherited mutations (
(105) Finally, we define as candidate driver somatic mutations those alleles satisfying the following criteria: 1) Disruptive and missense mutations in gene DNMT3A localized in exons 7 to 23 2) Disruptive mutations in gene ASXL1 with the exclusion of ASXL1 p.G646fsX12 and p.G645fsX58 3) Disruptive mutations in gene TET2 4) Disruptive mutations in gene PPM1D 5) Missense mutation JAK2 p.V617F 6) Mutations reported at least seven times in haematopoietic and lymphoid malignancies using the Catalogue of Somatic Mutations in Cancer3 with the exclusions of inherited mutations and potential PCR artifacts (Table S2)
(106) This definition does not take allelic fractions into account.
(107) Due to low coverage in one small region of ASXL1 (
(108) We performed a validation experiment for 65 mutations selected among putative somatic mutations and candidate driver somatic mutations from 12 subjects. A library preparation method utilizing a two round tailed amplicon PCR strategy was used to create targeted sequencing libraries for sequencing at high coverage on an Illumina MiSeq instrument. Alignment of sequencing reads against the GRCh37 human genome reference was performed using BWA MEM version 0.7.711 and allelic fractions were computed using the Unified Genotyper from the Genome Analysis Toolkit version 3.2-2.2
(109) For 65 of 65 mutations tested, molecular validation confirmed that the mutant allele was present at a low allelic fraction (significantly less than 50%) and thus could not have been inherited (
Example 2: DNMT3A and Other Driver Mutations
(110) We further performed validation for 30 candidate driver somatic mutations from two well-known recurrently mutated sites, DNMT3A p.R882H and JAK2 p.V617F. These were genotyped using TaqMan fluorescent assays in a droplet-based digital PCR system.12 Relative concentrations of each allele were quantitated through multiplexed fluorophores counted across approximately 15,000 nanoliter-sized droplets. Each somatic mutation that we attempted to validate was confirmed as somatic, including five JAK2 p.V617F mutations mutations showing at allelic fractions close to or above 50% (
(111) A total of 190 mutations across 185 subjects were identified in the DNMT3A gene (Table S4). Studies of mutations in haematologic malignancies have found DNMT3A mutations to be more common in cancers from females than in cancers from males (Markova, J. et al. Prognostic impact of DNMT3A mutations in patients with intermediate cytogenetic risk profile acute myeloid leukemia. Eur. J. Haematol. 88, 128-135 (2012); Roller, A. et al. Landmark analysis of DNMT3A mutations in hematological malignancies. Leukemia 27, 1573-1578 (2013)). We found that DNMT3A somatic mutations were also more common in females than in males (104/5780 vs. 81/6600; P=0.016 after adjusting for age using a linear regression model). We observed 48 disruptive mutations, and 142 in-frame indels or missense mutations including 23 mutations affecting the R882 amino acid of which 15 are R882H mutations known to dominantly inhibit wild-type DNMT3A (Russler-Germain, D. A. et al. The R882H DNMT3A Mutation Associated with AML Dominantly Inhibits Wild-Type DNMT3A by Blocking Its Ability to Form Active Tetramers. Cancer Cell 25, 442-454 (2014)). We also observed an enrichment within the DNMT3A FF interface region bounded by amino acid F732 and amino acid F772 (Jurkowska, R. Z. et al. Oligomerization and binding of the Dnmt3a DNA methyltransferase to parallel DNA molecules: heterochromatic localization and role of Dnmt3L. J. Biol. Chem. 286, 24200-24207 (2011)), similarly to what seen in DNMT3A mutations in acute myeloid leukemia (see Lawrence et al., Nature 505, 495-501 (23 Jan. 2014) doi:10.1038/nature12912).
(112) Of the 20 missense mutations within the FF interface region, 10 generated new cysteine residues (
(113) Based on the three-dimensional structure of DNMT3A, most of the predicted de novo disulfide bonds in mutant proteins would lead to severe structural change in the protein by disrupting the catalytic domain or influencing the oligomerization process (
(114) Our analysis identifies previously unknown cysteine forming mutations in DNMT3A in a cohort of patients, which we predict would lead to loss of enzymatic function.
(115) The vast majority of the mutations were dispersed across the genome. However, four genes (DNMT3A, TET2, ASXL1, and PPM1D) exhibited disproportionately high numbers of somatic mutations. Whereas the 95% of the mutations observed across the genome were missense and synonymous changes, the somatic mutations observed in DNMT3A, TET2, ASXL1, and PPM1D showed a different pattern: they strongly tended to disrupt gene protein-coding sequence by introducing a frameshift, nonsense, or splice-site disruption (commonly called disruptive mutations, though such mutations can also create proteins with altered or disregulated function) (
(116) The fourth implicated gene, PPM1D, which functions as a regulator of p53, has been described more frequently as mutated in malignancies of other cell types (Chuman, Y. et al. PPM1D430, a novel alternative splicing variant of the human PPM1D, can dephosphorylate p53 and exhibits specific tissue expression. J. Biochem. (Tokyo) 145, 1-12 (2009)). Of the 15 protein-truncating mutations observed in PPM1D, 12 occurred in the last exon, which is also the site of protein-truncating mutations described in cancer patients (Ruark, E. et al. Mosaic PPM1D mutations are associated with predisposition to breast and ovarian cancer. Nature 493, 406-410 (2013); Kleiblova, P. et al. Gain-of-function mutations of PPM1D/Wip1 impair the p53-dependent G1 checkpoint. J. Cell Biol. 201, 511-521 (2013); Akbari, M. R. et al. PPM1D mutations in circulating white blood cells and the risk for ovarian cancer. J. Natl. Cancer Inst. 106, djt323 (2014); Zhang, L. et al. Exome sequencing identifies somatic gain-of-function PPM1D mutations in brainstem gliomas. Nat. Genet. 46, 726-730 (2014)). Loss of the C-terminal localization domain of PPM1D is reported to activate PPM1D, repress p53, and thereby impair the p53-dependent G1 checkpoint, promoting proliferation.
(117) In addition to these disruptive mutations, DNMT3A also exhibited a strong (P<0.001) excess of missense mutations (
(118) Because DNMT3A, TET2, and ASXL1 are frequently mutated in haematological malignancies, we hypothesized that other recurring cancer mutations might also promote clonal haematopoiesis. We therefore considered 208 specific variants that have been reported in the Catalogue of Somatic Mutations in Cancer (Forbes, S. A. et al. COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer. Nucleic Acids Res. 39, D945-950 (2011)) to be mutated frequently (found in at least seven patients) in haematopoietic and lymphoid malignancies. We found 98 of these recurring mutations in our cohort, with 56 occurring in genes other than DNMT3A, TET2, ASXL1, and PPM1D. These recurrent mutations included the gain-of-function mutation JAK2 p.V617F found in 24 subjects; the DNMT3A mutation p.R882H (a proposed dominant negative; Russler-Germain, D. A. et al. The R882H DNMT3A Mutation Associated with AML Dominantly Inhibits Wild-Type DNMT3A by Blocking Its Ability to Form Active Tetramers. Cancer Cell 25, 442-454 (2014)) found in 15 subjects and the SF3B1 mutation p.K700E found in 9 subjects (Papaemmanuil, E. et al. Somatic SF3B1 mutation in myelodysplasia with ring sideroblasts. N. Engl. J. Med. 365, 1384-1395 (2011)). These mutations—including both the recurring, cancer-associated mutations and the disruptive mutations in DNMT3A, TET2, ASXL1, and PPM1D described above—comprised a set of 327 candidate driver somatic mutations for clonal haematopoiesis across 14 genes in 308 subjects (
Example 3: Clonal Haematopoiesis with Unknown Drivers
(119) Somatic mutations may either be “drivers” that contribute to clonal expansion or simply passive “passengers”. We tested whether subjects with clonal haematopoiesis with candidate drivers (CH-CD) also tended to carry additional putative somatic mutations. Subjects with CH-CD did indeed tend to carry more putative somatic mutations overall (mean 1.5, in addition to the candidate driver mutations themselves) than subjects without candidate drivers did (mean 0.23,
(120) Some 459 subjects had multiple putative somatic mutations without any of the candidate drivers described above. When multiple mutations were observed in the same individual, such mutations tended to have more-similar allelic fraction estimates than pairs of somatic mutations ascertained in different individuals (P<0.001, Mann-Whitney test for allelic fraction differences within and between subjects), consistent with the possibility that they were present in the same clone.
(121) Based on these results, we hypothesized that the presence of multiple somatic mutations might itself be an informative marker for clonal haematopoiesis, even when the exome sequencing analysis had not identified a candidate driver mutation. To consider cases of clonal haematopoiesis without obvious driver mutations, we sought to define a highly specific criterion for clonal haematopoiesis that depended only on the number (rather than identity) of the mutations. We identified 3,111 putative somatic mutations present at unusual allelic fractions. Whereas most individuals (9,927) had no putative somatic mutations, 1,333 had one; 313 had two; and 272 had from three up to eighteen (with 545 having sequence data of insufficient quality for detection). This distribution suggested that even if a random (“Poisson”) process generated many of the mutations observed in individuals with one or two mutations, a Poisson process (with a constant mean) could not explain the surprisingly high numbers of individuals with three to eighteen detectable mutations. In our analyses below, we classified subjects carrying three or more putative somatic mutations as having clonal haematopoiesis with unknown drivers (CH-UD); there were 195 such subjects.
(122) In some cases of CH-UD, additional analysis suggested potential candidate drivers. Somatic loss of chromosome Y (LOY) is known to be common in elderly men and a potential driver or a marker for clonal haematopoiesis (Forsberg, L. A. et al. Mosaic loss of chromosome Y in peripheral blood is associated with shorter survival and higher risk of cancer. Nat. Genet. 46, 624-628 (2014)). Sequence-coverage measurements across chromosome Y were used to estimate its copy number. Aligned sequencing reads are assigned mapping quality equal to 0 by BWA ALN1 when an alternative equally good alignment was identified by the aligner. Such reads on the sex chromosomes paralogous regions (PAR) have less predictive value to estimate LOY as they might come from the X chromosome even when aligned to the Y chromosome. We therefore measured for each subject: 1) number of sequencing reads over the Y chromosome with mapping quality greater than 0; 2) number of sequencing reads over regions X:1-2699520 (GRCh37 PAR1), X:154931044-155270560 GRCh37 PAR2), and over regions X:88456802-92375509 and Y:2917959-6616600 (GRCh37 PAR3) with mapping quality equal to 0.
(123) We then computed the relative amount of sequencing reads for each subject by dividing those numbers by the total number of aligned reads over the GRCh37 human genome reference for each subject (
(124) We found that LOY was more common in male subjects with CH-UD than in male subjects without clonal haematopoiesis (P<0.001, after adjusting for age using a linear regression model) and male subjects with CH-CD (P=0.002, after adjusting for age using a linear regression model). Approximately one fourth of male subjects with CH-UD showed some evidence for somatic LOY (
Example 4: Clonal Haematopoiesis and Advancing Age
(125) Detectable clonal haematopoiesis with candidate driver mutations (CH-CD) was rare among young individuals (0.74% before the age of 50) but much more common in the older population (5.7% after the age of 65) (
(126) Given that 459 subjects had multiple somatic mutations in the absence of candidate driver mutations, we sought to understand the extent to which this state arises dynamically over the lifespan (as opposed to being a lifelong property—for example, due to somatic mutations that occurred in embryonic development; Campbell, I. M. et al. Parental Somatic Mosaicism Is Underrecognized and Influences Recurrence Risk of Genomic Disorders. Am. J. Hum. Genet. 95, 1-10 (2014)). We therefore analyzed the age-dependent frequency of somatic genome states defined by the number of putative somatic mutations detected in the exome, excluding all subjects with candidate driver mutations. In contrast to the strongly age-dependent acquisition of CH-CD, the observation of a single somatic mutation in the exome was common at all ages (
Example 5: Clonal Haematopoiesis and Subsequent Cancer and Mortality
(127) We sought to understand how clonal haematopoiesis relates to subsequent cancer and mortality. Of the 503 individuals with evidence for clonal haematopoiesis (CH-CD or CH-UD), we were able to monitor subsequent medical history (median 33 months; range 2-7 years) in 455. Of these subjects, 15 developed haematological malignancies within three years from DNA sampling, with 8 developing myeloid malignancies and 6 developing lymphoid malignancies (Table S6). The myeloid malignancies arose in three subjects with SRSF2 p.P95H mutations, two subjects with JAK2 p.V617F, one subject with DNMT3A p.P904L, one with TP53 p.R248Q, and one subjects with CH-UD. The lymphoid malignancies arose in one subject with DNMT3A p.H613D, one subject with SF3B1 p.K700E and in four subjects with CH-UD.
(128) There were 55 subjects with a previous diagnosis of haematological malignancy. Of these, 14 showed clonal haematopoiesis (Table S7). Previous history of haematological malignancy was a strong risk factor for clonal haematopoiesis (odds ratio [OR]=6.0; 95% confidence interval [CI] 3.1 to 12; P<0.001, adjusting for age and sex using a linear regression model). There were also 31 subjects (42%) who developed haematological malignancies more than six months after DNA sampling. Of these, 13 showed clonal haematopoiesis (
(129) Subjects with clonal haematopoiesis (CH-CD or CH-UD) exhibited reduced overall survival (
Example 6: Malignant Clones in DNA Samples
(130) Two of the subjects with clonal haematopoiesis were diagnosed with myeloid malignancies just two months after DNA sampling (in both cases, this was their first diagnosis of any malignancy.) We hypothesized that the clone inferred from the exome sequence analysis might have been the malignant clone, at a pre-clinical stage. To evaluate this hypothesis, we performed whole-genome sequencing on both DNA samples to an average coverage of 108 times for each base pair of the genome (see Supplementary Appendix for details).
(131) High coverage whole-genome sequencing data were generated for Subject #1 and Subject #2 who were diagnosed with a myeloid malignancy two months after DNA sampling. Sequencing data were generated using four lanes from an Illumina HiSeq X Ten instrument for each subject with pair ended sequencing reads of 151 base pairs each and aligned against the GRCh37 human genome reference using BWA MEM version 0.7.7.11 Base pairs across the genome were sequenced on average 108 times per subject. Genotypes and allelic counts were computed across the genome using the Haplotype Caller from the Genome Analysis Toolkit version 3.2-2. Mutations of interest were further filtered out if: 1) already in the 1000 Genomes Project phase 1 dataset (see 1000 Genomes Project data hosted by the European Bioinformatics Institute FTP site: vol1/ftp/phase1/analysis_results/integrated_call_sets/ALL.wgs.integrated_phase1_v3.20101123.snps_indels_sv.sites.vcf.gz) 2) excluded from high confidence regions for the Genome in a Bottle genotype calls for NA1287828 (see Genome in a Bottle data hosted by the National Center for Biotechnology Information FTP site: giab/ftp/data/NA12878/variant_calls/NIST/union13callableMQonlymerged_addcert_nouncert_excludesimplerep_e xcludesegdups_excludedecoy_excludeRepSeqSTRs_noCNVs_v2.18_2mindatasets_5minYesNoRatio.bed.gz) 3) excluded from the strict mask of the 1000 Genomes Project phase 1 (see 1000 Genomes Project data hosted by the European Bioinformatics Institute FTP site: vol1/ftp/phase1/analysis_results/supporting/accessible_genome_masks/20120 824_strict_mask.bed) 4) within low complexity regions see Github files: lh3/varcmp/blob/master/scripts/LCRhs37d5.bed.gz) 5) present in more than two percent of the reads from each subject. These filters defined a dataset of 69,104 mutations across ˜50% of the GRCh37 human genome reference and ˜60% of the coding regions. When looking at mutations that failed the hypothesis that the alternate allelic count was distributed as a binomial process with mean 0.5 with a designed false positive rate of 0.01 or mutations at loci sequenced on average more than 200 times per subject, we observed that several of these mutations were clustering in hotspots. Upon further inspection, most of these calls were due to misalignment due to a paralogous region that was partially deleted in the human genome reference. We therefore further filtered out these mutations whenever they were found to be less than 1,000 bp from each other, further defining a refined dataset of 67,919 mutations across the two subjects. All putative somatic mutations were confirmed in whole-genome sequencing data.
(132) Whole-genome sequence analysis of the pre-clinical DNA sample revealed 1,153 putative somatic mutations in Subject #1 and 660 putative somatic mutations in Subject #2 (
Example 7: Genetic Relationship of Malignancies to Earlier Clones
(133) For two research subjects in the study, we were able to obtain and analyze bone marrow biopsies from their subsequent malignancies at the time of the first diagnosis. The first was Subject #2 (the subject whose earlier DNA sample we also analyzed by whole-genome-sequencing above, revealing the two CEBPA mutations), who was diagnosed with AML two months after DNA sampling. The other subject (Subject #3) was diagnosed with AML 34 months after DNA sampling. For both biopsy samples we generated (i) whole-exome sequence data to identify and measure the allelic fractions of protein-altering mutations, and (ii) low coverage whole-genome sequence data to identify large-scale gains or losses of chromosomal segments.
(134) Whole-exome sequencing data and low coverage whole-genome sequencing data of bone marrow biopsies were generated for Subject #2 and Subject #3. DNA was obtained from the diagnostic specimen available at the Clinical Genetics Department at Uppsala University (biobank application Bba-827-2014-064). 85 ng/μl and 88 ng/μl were obtained for, respectively, Subject #2 and Subject #3 in 10 μl water. The ThruPLEX-FD kit (Rubicon Genonics) was used to prepare three separate sequencing libraries from each subject starting from 2 μl of DNA. The three libraries were then pooled and subjected to exome capture using the SeqCap EZ Human Exome Library v3.0 kit according to standard protocols. Additionally, a fourth library was prepared with a separate index to perform lowpass whole-genome sequencing to assess the karyotypic profile of each subject.
(135) The pool of the three exome captured sequencing libraries for each individual was sequenced on one third of an Illumina Rapid Run flowcell (Hiseq 2500) at the Science for Life Laboratory in Sweden. The low-pass whole genome libraries were spiked in at a concentration of 1% each yielding 2.9 million read-pairs for Subject #2 and 3.2 million read-pairs for Subject #3. Sequencing reads of 101 base pairs each were aligned against the GRCh37 human genome reference using BWA MEM version 0.7.7 (Li, H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. ArXiv Prepr. ArXiv13033997 (2013)).
(136) Genotypes and allelic counts were computed across the genome using the Haplotype Caller from the Genome Analysis Toolkit version 3.2-2. Mutations of interest were further filtered out if: 1) excluded from high confidence regions for the Genome in a Bottle genotype calls for NA1287828 (see Genome in a Bottle data hosted by the National Center for Biotechnology Information FTP site: giab/ftp/data/NA12878/variant_calls/NIST/union13callableMQonlymerged_addcert_nouncert_excludesimplerep_excludesegdups_excludedecoy_excludeRepSeqSTRs_noCNVs_v2.18_2mindatasets_5minYesNoRatio.bed.gz) 2) excluded from the strict mask of the 1000 Genomes Project phase 1 (see 1000 Genomes Project data hosted by the European Bioinformatics Institute FTP site: vol1/ftp/phase1/analysis_results/supporting/accessible_genome_masks/20120824_strict_mask.bed) 3) within low complexity regions (see Github files: 1h3/varcmp/blob/master/scripts/LCRhs37d5.bed.gz)
(137) Analysis of the malignancy from Subject #2 (two months after initial DNA sampling) confirmed the presence of the mutations detected in the earlier DNA sample (including the two CEBPA mutations and two passenger mutations), now at higher allelic fractions (20.5% vs. the earlier 15.5%) and roughly consistent with the 50% blast count in the biopsy. We did not detect any novel clone other than the clone inferred two months before. Malignancies defined by pairs of CEBPA mutations tend to have a favorable prognosis, and indeed this patient enjoyed complete remission following chemotherapy and did not relapse.
(138) The initial DNA sample from Subject #3 contained a TP53 p.R248Q mutation at an allelic fraction of 24%. At diagnosis, the TP53 mutation had expanded to 86%, consistent with loss of heterozygosity and with the 86% blast cell count in the bone marrow biopsy. Low coverage whole-genome sequence data from the biopsy indicated losses of chromosome 17 and 5q (consisted with karyotype findings), and a complex karyotype pattern of gains and losses on chromosomes 12, 13, 16, and 19 (
Example 8: Subjects #1-#3
(139) Subject #1
(140) 85-years old male, diagnosed with myelodysplastic syndrome 2 months after DNA sampling. Died of unspecified leukemia 15 months after first diagnosis.
(141) Searching for mutations in genes previously observed as significantly mutated in acute myeloid leukemia in high coverage whole-genome sequencing data at the time of DNA sampling revealed recurrent somatic mutations ASXL1 p.G646fsX12 and RUNX1 p.L98fsX24, as well as somatic mutations TET2 p.Y1148fsX5, TET2 p.N1266S, and STAG2 p.E472_splice and further confirmed previously identified somatic mutation SRSF2 p.P95H (Table S9). Mutations in ASXL1 and TET2 were localized in regions of low coverage or no coverage and could not be detected in whole-exome sequencing data. Mutations in RUNX1 and STAG2 were not called in whole-exome sequencing data because observed in, respectively, only three and two sequencing reads. The somatic mutation ASXL1 p.G646fsX12 was at higher allelic fraction than the other candidate drivers, suggesting that this might have been the initiating lesion.
(142) Interestingly, mutations in ASXL1 have been shown to often co-occur in myelodysplastic syndromes with mutations in genes RUNX1 and SRSF2. Copy number analysis of whole-genome sequencing data revealed a normal karyotype.
(143) Subject #2
(144) 64-year-old male, diagnosed with acute leukemia 2 months after DNA sampling. Previous history unremarkable, was referred to the haematology unit due to fatigue and pancytopenia. Bone marrow examination showed a hypercellular marrow with 50% blast cells expressing CD34, CD117, CD13 and cytoplasmic MPO, i.e. AML FAB MO. Cytogenetics showed a normal karyotype. Following intense remission induction and consolidation chemotherapy, the patient obtained sustained complete remission. Four years later, he successfully underwent cystectomy due to a low differentiated urothelial cancer in the urinary bladder.
(145) High coverage whole-genome sequencing data at the time of DNA sampling revealed a 33 base pairs somatic insertion CEBPA p.K313_V314ins11 in the basic leucine zipper domain of the protein and previously observed in a different subject. The mutation in CEBPA was not called from whole-exome sequencing data due to the shorter 76 base pairs reads used. Upon further inspection of the data through the Integrative Genomics Viewer37 we also observed a 1 base pair frameshift deletion CEBPA p.P70fsX90 at lower allelic fraction of ˜7%, in agreement with the observation that in-frame C-terminal mutations, usually occurring in the basic-leucine zipper (bZIP) domain, are associated with frameshift N-terminal mutations in CEBPA (Barjesteh van Waalwijk van Doorn-Khosrovani, S. et al. Biallelic mutations in the CEBPA gene and low CEBPA expression levels as prognostic markers in intermediate-risk AML. Hematol. J. Off J. Eur. Haematol. Assoc. EHA 4, 31-40 (2003)). This mutation was not automatically called by the Haplotype Caller from the Genome Analysis Toolkit due to low allelic counts. Copy number analysis of whole-genome sequencing data both at the time of DNA sampling and at the time of diagnosis confirmed a normal karyotype (
(146) Whole-exome sequencing data of the bone marrow biopsy further confirmed the presence of the two CEBPA mutations and of three previously identified putative somatic mutations (Table S10). Estimated collective allelic fractions for these three putative somatic mutations increased in frequency between DNA sampling and first diagnosis (15.5% vs. 20.5%; P=0.037, left-tailed Fisher exact test).
(147) Subject #3
(148) 75-year-old female, diagnosed with AML 34 months after DNA sampling. SLE with mainly cutaneous manifestations since 15 years which had been treated with steroids but not chemotherapy. Referred to the haematology unit due to pancytopenia, fatigue and pulmonary infection. Bone marrow examination showed a hypercellular marrow with 86% blast cells with no maturation and expressing CD34, CD117, CD13 and cytoplasmic MPO, i.e. AML FAB MO. Cytogenetics showed a complex karyotype including monosomy 17 and 5q-. The patient received palliative treatment with hydroxyurea and died one month later due to the leukemia.
(149) Whole-exome sequencing data at the time of DNA sampling revealed somatic mutation TP53 p.R248Q at an estimated allelic fraction of 24%. Whole-exome sequencing data of the bone marrow biopsy confirmed this somatic mutation at a much higher estimated allelic fraction of 86%. Copy number analysis from low coverage whole-genome sequencing data confirmed that the malignancy was monosomy for chromosome 17,39 had a partial loss of chromosome arm 5q, 40 and a complex karyotype pattern involving chromosomes 12, 13, 16, and 19 (
(150) To test if these events were already present at the time of DNA sampling, we analyzed allelic fractions for the following six regions deleted in the malignancy:
(151) 1) chromosome 17 (
(152) 2) chromosome arm 5q from Mbp 72 to Mbp 155 (
(153) 3) chromosome arm 12p up to Mbp 26 (
(154) 4) chromosome arm 13q from Mbp 91 (
(155) 5) chromosome arm 16q (
(156) 6) chromosome arm 19q up to Mbp 35 (
(157) For each region we tested if allelic fractions for alleles retained in the malignancy and alleles lost in the malignancy were significantly different at the time of DNA sampling using a Mann-Whitney test. This test resulted significant for chromosome 17 (45.5% vs. 48.3%; P<0.001,
(158) Statistics and Figures
(159) Cox proportional hazards analyses and Kaplan-Meier plots were performed and generated using the R survival package. Forest plots were generated using the R metafor package.
(160) All remaining figures were generated using the R ggplot2 package and Google Drawings.
(161) Tables
(162) TABLE-US-00001 TABLE S1 Mean age and standard deviation of different groups ascertained in the cohort. Group Count Age Total 12,380 55 ± 12 Male 6,600 52 ± 11 Male control 3,187 56 ± 11 Male schizophrenia 2,964 53 ± 11 Male bipolar 454 NA Female 5,780 56 ± 12 Female control 3,063 57 ± 12 Female schizophrenia 2,006 55 ± 12 Female bipolar 711 NA
(163) TABLE-US-00002 TABLE S2 Mutations observed at least seven times in hematologic and lymphoid cancers in the Catalogue Of Somatic Mutations In Cancer (COSMIC) database v69 (released Jun. 2, 2014) and excluded from analysis in this study. Mutation ASXL1 p.G646fsX12 is a genuine recurrent somatic mutation but due to low coverage at the site of the mutation it was impossible to distinguish true positives from PCR artifacts. Number of observations in hematopoietic and Reason for Variant Amino acid change COSMIC ID lymphoid cancer exclusion rs10521 NOTCH1 COSM33747 11 Inherited mutation p.D1698D COSM1461158 rs3822214 KIT COSM28026 16 Inherited mutation p.M541L rs10663835 CNDP1 COSM307404 8 Inherited mutation p.L20_E21insL COSM1683699 rs55980345 PKD1L2 COSM314177 7 Inherited mutation p.N236fsX26 COSM314178 COSM1684461 COSM1684462 rs139115934 ASXL1 COSM36205 15 Inherited mutation p.E1102D rs146317894 OR52D1 COSM1683657 7 Inherited mutation p.T204fsX33 rs147836249 TET2 COSM87107 7 Inherited mutation p.F868L NA ASXL1 COSM34210 319 Potential PCR p.G646fsX12 COSM1411076 slippage error due COSM1658769 to G homopolymer run NA ASXL1 CO5M85923 0 Potential PCR p.G645fsX58 COSM1180918 slippage error due to G homopolymer run NA NOTCH1 COSM13047 15 Potential PCR p.V1578delV slippage error due to CAC tandem repeat
(164) TABLE-US-00003 TABLE S3 List of candidate driver somatic mutations detected in the cohort. Position dbSNP 138 Reference Altenate Reference Alternate COSMIC COSMIC Chromosome (GRCh37) ID Allele Allele Count Count ID Count Gene Annotation 2 25,457,164 NA T C 31 28 NA 0 DNMT3A NM_022552:exon23:c.A2723G:p.Y908C 2 25,457,164 NA T C 81 19 NA 0 DNMT3A NM_022552:exon23:c.A2723G:p.Y908C 2 25,457,164 NA T C 94 29 NA 0 DNMT3A NM_022552:exon23:c.A2723G:p.Y908C 2 25,457,168 NA C T 65 41 NA 0 DNMT3A NM_022552:exon23:c.G2719A:p.E907K 2 25,457,173 NA A C 121 35 NA 0 DNMT3A NM_022552:exon23:c.G2714G:p.L905R 2 25,457,173 NA A T 97 30 NA 0 DNMT3A NM_022552:exon23:c.G2714A:p.L905Q 2 25,457,176 rs149095705 G A 55 6 87007 6 DNMT3A NM_022552:exon23:c.C2711T:p.P904L 2 25,457,176 rs149095705 G A 65 21 87007 6 DNMT3A NM_022552:exon23:c.C2711T:p.P904L 2 25,457,176 rs149095705 G A 81 13 87007 6 DNMT3A NM_022552:exon23:c.C2711T:p.P904L 2 23,457,176 rs149095705 G A 88 11 87007 6 DNMT3A NM_022552:exon23:c.C2711T:p.P904L 2 25,457,192 NA G A 67 40 NA 0 DSMT3A NM_022552:exon23:c.C2695T:p.R899C 2 25,457,204 NA C T 82 23 335620 0 DSMT3A NM_022552:exon23:c.G2683A:p.V895M 335621 2 25,457,209 NA C T 72 25 NA 0 DSMT3A NM_022552:exon23:c.G2678A:p.W893X 2 25,457,215 NA CG C 51 12 NA 0 DSMT3A NM_022552:exon23:c.2671_2672G 2 25,457,218 NA C T 59 13 1482984 1 DSMT3A NM_022552:exon23:c.2669A:p.G890D 256042 2 25,457,242 rs147001633 C G 50 5 3356083 14 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 99740 2 25,457,242 rs147001633 C G 75 9 3356083 14 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 99740 2 25,457,242 rs147001633 C T 27 3 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 30 12 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 44 5 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 32944 2 25,457,242 rs147001633 C T 45 5 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 47 7 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 48 10 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 48 15 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 48 7 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 50 7 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 51 16 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 52 6 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 53 8 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 56 17 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 60 10 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,242 rs147001633 C T 63 8 442676 392 DSMT3A NM_022552:exon23:c.G2645C:p.R882P 52944 2 25,457,243 rs377577594 G A 29 3 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,243 rs377577594 G A 29 8 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,243 rs377577594 G A 31 4 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,243 rs377577594 G A 59 10 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,243 rs377577594 G A 69 8 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,243 rs377577594 G A 77 24 1165704 164 DSMT3A NM_022552:exon23:c.C2644T:p.R882C 53042 2 25,457,249 NA T C 58 19 120499 3 DSMT3A NM_022552:exon23:c.A2638G:p.M880V 2 25,458,595 rs373014701 A G 38 11 231568 2 DSMT3A NM_022552:exon22:c.T2578C:p.W860R 2 25,458,595 rs373014701 A G 43 12 231568 2 DSMT3A NM_022552:exon22:c.T2578C:p.W860R 2 25,459,595 rs373014701 A G 50 14 231568 2 DSMT3A NM_022552:exon22:c.T2578C:p.W860R 2 25,459,595 rs373014701 A G 86 17 231568 2 DSMT3A NM_022552:exon22:c.T2578C:p.W860R 2 25,459,595 rs373014701 A G 87 11 231568 2 DSMT3A NM_022552:exon22:c.T2578C:p.W860R 2 75,458,619 NA T C 49 23 NA 0 DSMT3A NM_022552:exon22:c.A2554G:p.M852V 2 25,458,646 NA C T 93 20 NA 0 DSMT3A NM_022552:exon22:c.G2527A:p.G843S 2 25,458,696 NA T C 40 16 NA 0 DSMT3A NM_022552:exon23:c.2479−2A>G 2 25,459,804 NA C A 28 6 NA 0 DSMT3A NM_022552:exon22:c.2478+1G>T 2 25,459,837 NA G A 28 7 99739 1 DSMT3A NM_022552:exon21:c.C2446T:pC2446T:p.Q816X 2 25,461,998 NA C T 23 5 NA 0 DSMT3A NM_022552:exon21:c.2408+1G>A 2 25,462,020 NA C A 38 12 NA 0 DSMT3A NM_022552:exon20:c.G2387T:p.G796V 2 25,462,024 NA A G 37 10 NA 0 DSMT3A NM_022552:exon20:c.T2383C:p.W795R 2 25,462,032 NA C T 36 7 720761 4 DSMT3A NM_022552:exon20:c.G2375A:p.R792H 720762 2 25,462,068 rs370751539 A G 33 7 1583121 1 DSMT3A NM_022552:exon20:c.T2539C:p.I780T 2 25,462,077 NA G C 19 13 NA 0 DSMT3A NM_022552:exon20:c.T2330G:p.P777R 2 25,462,085 NA C T 22 11 NA 0 DSMT3A NM_022552:exon21:c.2323−1G>A 2 25,463,174 NA GAGAAATC G 152 19 NA 0 DSMT3A NM_022552:exon19:c.2305_2319C GCCAGAT 2 25,463,182 NA G A 144 23 231563 4 DSMT3A NM_022552:exon19:c.T2311T:p.R771X 2 25,463,184 NA G T 169 36 1583106 1 DSMT3A NM_022552:exon19:c.T2309A:p.S770X 2 25,463,187 NA A G 183 45 NA 0 DSMT3A NM_022552:exon19:c.T2306C:p.I769T 2 25,463,195 NA CTT C 61 33 NA 0 DSMT3A NM_022552:exon19:c.2296_2298G 2 25,463,212 NA T C 84 89 NA 0 DSMT3A NM_022552:exon19:c.A2281G:p.M761V 2 25,463,225 NA C A 173 42 NA 0 DSMT3A NM_022552:exon19:c.G2268T:p.E756D 2 25,463,229 NA A G 126 20 NA 0 DSMT3A NM_022552:exon19:c.T2264C:p.F755S 2 25,463,229 NA A G 43 16 NA 0 DSMT3A NM_022552:exon19:c.T2264C:p.F755S 2 25,463,234 NA C G 105 30 NA 0 DSMT3A NM_022552:exon19:c.G2259C:p.W753C 2 25,463,241 NA A C 193 31 NA 0 DSMT3A NM_022552:exon19:c.T2252G:p.F751C 2 25,463,248 NA G A 153 47 239133 4 DSMT3A NM_022552:exon19:c.T2245T:p.R749C 2 25,463,248 NA G A 90 23 219333 4 DSMT3A NM_022552:exon19:c.T2245T:p.R749C 2 25,463,286 rs139293773 C T 137 25 1318940 6 DSMT3A NM_022552:exon19:c.G2207A:p.R736H 133737 2 25,463,286 rs139293773 C T 44 36 1318940 6 DSMT3A NM_022552:exon19:c.G2207A:p.R736H 133737 2 25,463,286 rs139293773 C T 55 32 1318940 6 DSMT3A NM_022552:exon19:c.G2207A:p.R736H 133737 2 25,463,286 rs139293773 C T 84 12 1318940 6 DSMT3A NM_022552:exon19:c.G2207A:p.R736H 133737 2 25,463,287 NA G A 71 18 231560 5 DSMT3A NM_022552:exon19:c.C2206T:p.R736C 2 25,463,289 rs147828672 T C 100 25 133126 4 DSMT3A NM_022552:exon19:c.A2204G:p.Y735C 2 25,463,289 rs147828672 T C 76 21 133126 4 DSMT3A NM_022552:exon19:c.A2204G:p.Y735C 2 25,463,289 rs147828672 T C 84 13 133126 4 DSMT3A NM_022552:exon19:c.A2204G:p.Y735C 2 25,463,289 rs147829872 T C 90 13 133126 4 DSMT3A NM_022552:exon19:c.A2204G:p.Y735C 2 25,463,295 NA T C 66 10 NA 0 DSMT3A NM_022552:exon19:c.A2198G:p.E733G 2 23,463,296 NA CAA C 79 20 NA 0 DSMT3A NM_022552:exon19:c2195 2197G 2 25,463,296 NA C CA 23 3 NA 0 DSMT3A NM_022552:exon19:c.2197_2197delinsTG 2 25,463,296 NA C CA 48 19 NA 0 DSMT3A NM_022552:exon19:c.2197_2197delinsTG 2 25,463,297 NA AAAG A 107 26 1583117 8 DSMT3A NM_022552:exon19:c.2193_2196T 99742 2 25,463,297 NA AAAG A 138 35 1593117 8 DSMT3A NM_022552:exon19:c.2193_2196T 99742 2 25,463,297 NA AAAG A 77 20 1583117 8 DSMT3A NM_022552:exon19:c.2193_2196T 99742 2 25,463,297 NA AAAG A 92 22 1583117 8 DSMT3A NM_022552:exon19:c.2193_2196T 99742 2 25,463,298 NA A C 101 18 NA 0 DSMT3A NM_022552:exon19:c.T2195G:p.F732C 2 25,463,308 rs200018028 G A 58 70 1318937 4 DSMT3A NM_022552:exon19:c.C2185T:p.R729W 249142 2 25,463,308 rs200018028 G A 61 22 1338937 4 DSMT3A NM_022552:exon19:c.C2185T:p.R729W 249142 2 25,463,541 rs367909007 G C 124 21 442677 11 DSMT3A NM_022552:exon18:c.C2141G:p.S714C 87011 2 25,463,541 rs367909007 G C 164 29 442677 11 DSMT3A NM_022552:exon18:c.C2141G:p.S714C 87011 2 25,463,541 rs367909007 G C 172 28 442677 11 DSMT3A NM_022552:exon18:c.C2141G:p.S714C 87011 2 25,463,554 NA A T 79 16 249803 1 DSMT3A NM_022552:exon18:c.T2128A:p.C710S 2 25,463,565 NA C T 117 31 NA 0 DSMT3A NM_022552:exon18:c.G2117A:p.G706E 2 25,463,566 NA CA C 62 9 NA 0 DSMT3A NM_022552:exon18:c.2115_2116G 2 25,463,574 NA AG A 71 24 NA 0 DSMT3A NM_022552:exon18:c.2107_2108T 2 25,463,578 NA C T 117 18 NA 0 DSMT3A NM_022552:exon18:c.G2104A:p.D702N 2 25,463,593 NA C A 38 24 NA 0 DSMT3A NM_022552:exon18:c.G2089T:p.E697X 2 25,463,595 NA TG T 137 18 1583101 1 DSMT3A NM_022552:exon18:c.2086_2087A 2 25,464,430 NA C T 33 13 NA 0 DSMT3A NM_022552:exon18:c.2082+1G>A 2 25,464,430 NA C T 46 9 NA 0 DSMT3A NM_022552:exon18:c.2082+1G>A 2 25,464,430 NA C T 51 8 NA 0 DSMT3A NM_022552:exon18:c.2082+1G>A 2 25,464,450 rs369713081 C T 42 5 NA 0 DSMT3A NM_022552:exon17:c.G2063A:p.R688H 2 25,464,450 rs369713081 C T 43 35 NA 0 DSMT3A NM_022552:exon17:c.G2063A:p.R688H 2 25,464,459 NA C T 29 7 1690275 0 DSMT3A NM_022552:exon17:c.C2054A:p.G685E 1690276 2 25,464,470 NA GA G 38 7 NA 0 DSMT3A NM_022552:exon17:c.2042_2043C 2 25,464,470 NA G C 58 11 NA 0 DSMT3A NM_022552:exon17:c.C2043G:p.I681M 2 25,464,471 NA A T 43 8 NA 0 DSMT3A NM_022552:exon17:c.T2042A:p.I681N 2 25,464,486 NA C A 35 24 NA 0 DSMT3A NM_022552:exon17:c.G2027T:p.R676L 2 25,464,507 NA GAGTCCT G 40 7 NA 0 DSMT3A NM_022552:exon17:c.2000_2006C 2 25,464,520 NA C A 41 21 NA 0 DSMT3A NM_022552:exon17:c.G1993T:p.V665L 2 25,464,529 NA C T 42 23 NA 0 DSMT3A NM_022552:exon17:c.G1984A:p.A662T 2 25,464,544 rs368961181 C T 17 5 NA 0 DSMT3A NM_022552:exon17:c.G1969A:p.V657M 2 25,464,544 rs368961181 C T 33 11 NA 0 DSMT3A NM_022552:exon17:c.G1969A:p.V657M 2 25,464,544 rs368961181 C T 34 10 NA 0 DSMT3A NM_022552:exon17:c.G1969A:p.V657M 2 25,464,549 NA A T 28 7 133136 1 DSMT3A NM_022552:exon17:c.T1964A:p.I655N 3 25,467,023 NA C A 58 9 NA 0 DSMT3A NM_022552:exon16:c.C1851+1G>T 2 25,467,029 NA C A 89 15 NA 0 DSMT3A NM_022552:exon15:c.G1846T:p.E616X 2 25,467,034 NA TC T 81 28 NA 0 DSMT3A NM_022552:exon15:c.1840_1841A 2 25,467,038 NA G C 44 21 NA 0 DSMT3A NM_022552:exon15:c.C1837G:p.H613D 2 25,467,061 NA A G 62 22 NA 0 DSMT3A NM_022552:exon15:c.T1814C:p.L605P 2 25,467,064 NA C T 40 25 NA 0 DSMT3A NM_022552:exon15:c.G1811A:p.R604Q 2 25,467,078 NA C A 30 19 NA 0 DSMT3A NM_022552:exon15:c.G1797T:p.E599D 2 25,467,078 NA C A 39 23 NA 0 DSMT3A NM_022552:exon15:c.G1797T:p.E599D 2 25,467,078 NA C A 58 42 NA 0 DSMT3A NM_022552:exon15:c.G1797T:p.E599D 2 25,467,078 NA C A 63 54 NA 0 DSMT3A NM_022552:exon15:c.G1797T:p.E599D 2 25,467,083 NA G A 49 18 133736 4 DSMT3A NM_022552:exon15:c.G1792T:p.R598X 2 25,467,086 NA G A 39 34 NA 0 DSMT3A NM_022552:exon15:c.G1789T:p.R597W 3 25,467,133 NA CAGGGGT C 34 5 NA 0 DSMT3A NM_022552:exon15:c1736_1742G 2 23,467,136 NA G C 7 14 NA 0 DSMT3A NM_022552:exon15:c.C1739G:p.R580R 2 23,467,169 NA G A 13 14 NA 0 DSMT3A NM_022552:exon15:c.C1706T:p.P569L 2 25,467,410 NA T C 53 33 NA 0 DSMT3A NM_022552:exon14:c.A1666G:p.R556G 2 25,467,428 NA C T 67 12 256035 4 DSMT3A NM_022552:exon14:c.G1648A:p.G550R 2 25,467,449 NA C A 53 8 87002 10 DSMT3A NM_022552:exon14:c.G1627T:p.G543C 2 25,467,481 NA CCGT C 37 13 1583078 1 DSMT3A NM_022552:exon14:c.1592_1595G 2 23,467,490 NA T A 69 17 NA 0 DSMT3A NM_022552:exon14:c.A1586T:p.D529V 2 25,467,516 NA G T 67 12 NA 0 DSMT3A NM_022552:exon14:c.C1560A:p.C520X 2 25,468,120 NA A C 60 20 NA 0 DSMT3A NM_022552:exon14:c.1554+2T>G 2 25,468,121 NA C T 103 12 NA 0 DSMT3A NM_022552:exon14:c.1554+1G>A 2 25,468,121 NA C T 63 10 NA 0 DSMT3A NM_022552:exon14:c.1554+1G>A 2 25,462,138 NA A AT 46 11 NA 0 DSMT3A NM_022552:exon13:c.1538 1538delinsAT 2 25,468,174 rs149738328 T C 37 32 231571 3 DSMT3A NM_022552:exon13:c.A1502G:p.N501S 2 25,468,174 rs149738328 T C 50 32 231571 3 DSMT3A NM_022552:exon13:c.A1502G:p.N501S 2 25,468,186 NA C T 23 7 1318925 3 DSMT3A NM_022552:exon13:c.G1490A:p.C497Y 1318926 2 25,468,888 NA C T 105 43 NA 0 DSMT3A NM_022552:exon13:c.1474+1G>A 2 25,468,912 NA C T 65 1 NA 0 DSMT3A NM_022552:exon12:c.G1451A:p.R484Q 2 25,468,922 NA A C 55 3 NA 0 DSMT3A NM_022552:exon12:c.T1441G:p.Y481D 2 25,469,053 NA C A 125 28 NA 0 DSMT3A NM_022552:exon11:c.G1405T:p.E469X 2 25,469,060 NA CT C 133 25 NA 0 DSMT3A NM_022552:exon11:c.1397_1398G 2 25,469,080 NA T C 106 90 NA 0 DSMT3A NM_022552:exon11:c.A1378G:p.S460G 2 23,469,100 NA G A 104 89 NA 0 DSMT3A NM_022552:exon11:c.C1358T:p.P453L 2 23,469,100 NA G A 77 99 NA 0 DSMT3A NM_022552:exon11:c.C1358T:p.P453L 2 25,469,139 NA C T 179 38 NA 0 DSMT3A NM_022552:exon11:c.G1319A:p.W440X 2 25,469,142 NA A G 153 102 NA 0 DSMT3A NM_022552:exon11:c.T1316C:p.M439T 2 25,469,142 NA A G 80 66 NA 0 DSMT3A NM_022552:exon11:c.T1316C:p.M439T 2 25,469,174 NA CT C 167 24 NA 0 DSMT3A NM_022552:exon11:c.1283_1284G 2 25,469,501 NA C G 52 70 NA 0 DSMT3A NM_022552:exon10:c.G1267C:p.E423Q 2 25,469,614 NA G A 109 73 NA 0 DSMT3A NM_022552:exon10:c.C1154T:p.P385L 2 25,469,614 NA G A 61 39 NA 0 DSMT3A NM_022552:exon10:c.C1154T:p.P385L 2 25,469,614 NA G A 97 62 NA 0 DSMT3A NM_022552:exon10:c.C1154T:p.P385L 2 25,469,633 NA G A 83 14 NA 0 DSMT3A NM_022552:exon10:c.C1135T:p.R379C 2 25,469,647 NA T G 149 18 NA 0 DSMT3A NM_022552:exon11:c.1123−2A>C 2 25,469,927 NA A G 23 14 NA 0 DSMT3A NM_022552:exon9:c.T1115C:p.V372A 2 25,469,928 rs371677904 C T 21 20 NA 0 DSMT3A NM_022552:exon9:c.G1114A:p.V372I 2 25,469,951 NA A G 30 17 NA 0 DSMT3A NM_022552:exon9:c.T1091C:p.M364T 2 25,464,987 rs139053291 C T 24 11 133129 1 DSMT3A NM_022552:exon9:c.G1055A:p.S352N 2 25,469,988 NA TGC TT 53 9 NA 0 DSMT3A NM_022552:exon9:c.1052_1054AA 2 25,470,011 NA A G 17 6 NA 0 DSMT3A NM_022552:exon9:c.T1031C:p.L344P 2 25,470,019 NA A AAC 23 9 NA 0 DSMT3A NM_022552:exon9:c.1023_1023delinsGTT 2 25,470,028 NA CT C 21 6 NA 0 DSMT3A NM_022552:exon10:c.1015_splice 2 25,470,479 NA C T 147 30 477212 0 DSMT3A NM_022552:exon8:c.G995A:p.G332E 2 25,470,480 NA C T 102 48 NA 0 DSMT3A NM_022552:exon8:c.G994A:p.G332R 2 25,470,484 NA C T 150 21 249799 1 DSMT3A NM_022552:exon8:c.G990A:p.W330X 2 25,470,484 NA C T 72 11 249799 1 DSMT3A NM_022552:exon8:c.G990A:p.W330X 2 25,470,498 NA G A 90 17 NA 0 DSMT3A NM_022552:exon8:c.C976T:p.R326C 2 25,470,516 NA G A 108 17 1318922 4 DSMT3A NM_022552:exon8:c.C958T:p.R320X 133721 133724 2 25,470,516 NA G A 98 16 1318922 4 DSMT3A NM_022552:exon8:c.C958T:p.R320X 133721 133724 2 25,470,532 NA C T 83 30 NA 0 DSMT3A NM_022552:exon8:c.G942A:p.W314X 2 25,470,554 NA G A 77 16 NA 0 DSMT3A NM_022552:exon8:c.C920T:p.P307L 2 25,470,554 NA G C 51 6 221579 1 DSMT3A NM_022552:exon8:c.C920G:p.P307R 2 25,470,554 NA G C 86 18 221579 1 DSMT3A NM_022552:exon8:c.C920G:p.P307R 2 25,470,556 NA C T 60 10 NA 0 DSMT3A NM_022552:exon8:c.G918A:p.W306X 2 25,470,588 NA C T 60 13 NA 0 DSMT3A NM_022552:exon8:c.G886A:p.V296M 2 25,470,588 NA C T 83 15 NA 0 DSMT3A NM_022552:exon8:c.G886A:p.V296M 2 25,470,588 NA C T 86 18 NA 0 DSMT3A NM_022552:exon8:c.G886A:p.V296M 2 25,470,591 NA G C 48 10 NA 0 DSMT3A NM_022552:exon8:c.C883G:p.L295V 2 25,470,599 NA A G 70 19 NA 0 DSMT3A NM_022552:exon8:c.T875C:p.I292T 2 25,470,599 NA A G 99 17 NA 0 DSMT3A NM_022552:exon8:c.T875C:p.I292T 2 25,471,024 NA G GC 71 18 NA 0 DSMT3A NM_022552:exon7:c.737 737delinsGC 2 25,471,064 NA GC G 58 22 NA 0 DSMT3A NM_022552:exon7:c.696_697C 2 198,266,834 NA T C 148 16 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 50 12 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 50 16 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 53 6 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 60 17 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 66 10 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 79 14 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 91 8 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,266,834 NA T C 97 11 84677 230 SFB31 NM_12433:exon15:c.A2098G:p.K700E 2 198,267,359 rs377023736 C A 207 27 131557 13 SFB31 NM_12433:exon14:c.G1998T:p.K666N 2 198,267,359 rs377023736 C G 66 22 132937 9 SFB31 NM_12433:exon14:c.G1998T:p.K666N 2 198,267,360 NA T G 61 11 131556 8 SFB31 NM_12433:exon14:c.A1997C:p.K666T 2 197,267,491 NA C G 106 15 132938 7 SFB31 NM_12433:exon14:c.G1866C:p.E622D 3 38,182,641 rs387907272 T C 91 21 85990 1027 MYD38 NM_002468:exon5:c.7941T>C:p.L265P 4 106,155,544 NA G T 29 14 3428018 0 TET2 NM_017628:exon3:c.G445T:p.E149X 3428019 4 106,155,915 NA GC G 24 12 NA 0 TET2 NM_017628:exon3:c.816_817G 4 106,156,079 NA C G 97 18 NA 0 TET2 NM_017628:exon3:c.C980G:p.S327X 4 106,156,409 NA A AC 73 12 NA 0 TET2 NM_017628:exon3:c.1310_1310delinsAC 4 106,156,441 NA G T 38 9 NA 0 TET2 NM_017628:exon3:c.G1342T:p.E448X 4 106,156,564 NA GA G 106 23 NA 0 TET2 NM_017628:exon3:c.1465_1466G 4 106,156,623 NA GT G 50 11 NA 0 TET2 NM_017628:exon3:c.1524_1525G 4 106,156,747 NA C T 119 13 1318629 26 TET2 NM_017628:exon3:c.C1678T:p.R550X 41644 4 106,156,758 NA G GC 152 31 43490 3 TET2 NM_017628:exon3:c.1659_1659delinsGC 4 106,157,162 NA A AT 105 16 NA 0 TET2 NM_017628:exon3:c.2063_20639delinsAT 4 106,157,332 NA CAG C 39 28 NA 0 TET2 NM_017628:exon3:c.2233_2235C 4 106,157,335 NA C T 53 10 87099 1 TET2 NM_017628:exon3:c.C2236T:p.Q746X 4 106,157,367 NA AC A 75 39 NA 0 TET2 NM_017628:exon3:c.2268_2269A 4 106,157,467 NA C T 53 10 43416 1 TET2 NM_017628:exon3:c.C2368T:p.Q790X 4 106,157,503 NA GT G 66 14 NA 0 TET2 NM_017628:exon3:c.2404_2405G 4 106,157,525 NA TA T 68 11 NA 0 TET2 NM_017628:exon3:c.2426_2477T 4 106,157,542 NA A T 55 22 NA 0 TET2 NM_017628.exon3:c.A2443T:p.R815X 4 106,157,608 NA AAT A 53 20 NA 0 TET2 NM_017628:exon3:c.2509_2511A 4 106,157,638 NA C T 38 8 NA 0 TET2 NM_017628.exon3:c.C2539T:p.Q847X 4 106,157,761 NA C T 54 11 NA 0 TET2 NM_017628.exon3:c.C2662T:p.Q888X 4 106,157,842 NA G GCT 31 10 NA 0 TET2 NM_017628:exon3:c.2743_ 2743delinsGCT 4 106,158,224 NA AC A 97 19 NA 0 TET2 NM_017628:exon3:c.3125 3126A 4 106,158,349 NA CA C 77 12 NA 0 TET2 NM_017628:exon3:c.3250_3251C 4 106,158,359 NA CTT C 42 10 NA 0 TET2 NM_017628:exon3:c.3260_3262C 4 106,158,378 NA C CA 18 3 NA 0 TET2 NM_017628:exon3:c.3279_3279delinsCA 4 106,158,378 NA C CA 40 9 NA 0 TET2 NM_017628:exon3:c.3279_3279delinsCA 4 106,158,442 NA C CT 55 17 NA 0 TET2 NM_017628:exon3:c.3343 3343delinsCT 4 106,158,485 NA AT A 69 22 NA 0 TET2 NM_017628:exon3:c.3386_3387A 4 106,159,509 NA G A 75 24 87117 1 TET2 NM_001127208:exon3:c.3409+1G>2A 4 106,158,579 NA A AT 32 23 NA 0 TET2 NM_017628:exon3:c.3480_3480delinsAT 4 106,158,595 NA T A 54 22 NA 0 TET2 NM_017628:exon3:c.T3496A:p.X1166K 9 5,073,770 rs386626619 G T 101 20 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,770 rs386626619 G T 115 14 12600 30,637 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,770 rs386626619 G T 117 18 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 125 11 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 126 14 12600 30,687 JAK2 NM 004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 126 21 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 175 16 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 31 59 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 45 56 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 47 73 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 49 57 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 63 53 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 64 17 12600 30,687 JAK2 NM 004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 66 23 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 69 15 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 70 9 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 73 10 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 79 9 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 81 7 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 81 9 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 84 13 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 87 19 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 88 28 12600 30,687 JAK2 NM_004972:exon14:c.G1849T:p.V617F 9 5,073,170 rs386626619 G T 88 42 12600 30,687 JAK2 NM 004972:exon14:c.G1849T:p.V617F 11 108,236,087 NA G A 81 7 11396011 8 ATM NM_000051:exon63:c.G9023A:p.R3008H 21626 11 119,148,891 rs267606706 T C 30 8 34052 24 CBL NM_005188:exon8:c.T1111C:p.Y371H 11 119,149,251 rs267606708 G A 109 18 34077 11 CBL NM_005188:exon9:c.G1259A:p.R420Q 11 119,349,251 rs267606708 G A 125 13 34077 11 CBL NM_005188:exon9:c.G1259A:p.R420Q 15 90,631,935 NA G A 81 11 41877 10 IDH2 NM_002188:exon4:c.C418T:p.R140W 17 7,577,538 rs11540652 C T 79 25 10662 71 TP53 NM_000546:exon7:c.G743A:p.R248Q 1640830 3356964 99020 99021 99602 17 7,577,538 rs11540652 C T 83 15 10662 71 TP53 NM_000546:exon7:c.G743A:p.R248Q 1640830 3356964 99020 99021 99602 17 7,577,568 NA C T 63 29 11059 8 TP53 NM_000546:exon7:c.G713A:p.C238Y 1649400 179811 179812 179813 3388191 17 7,578,190 NA T C 26 17 10758 23 TP53 NM_000546:exon6:c.A659G:p.Y220C 1644277 3355993 99718 99719 99720 17 40,474,482 NA T A 188 18 1155243 45 STAT3 NM_003150:exon21:c.A1919T:p.Y640F 17 58,678,121 NA G GC 11 5 NA 0 PPMID NM_003620:exon1:c.346_346delinsGC 17 58,725,309 NA GAC G 37 41 NA 0 PPMID NM_003620:exon4:c.883_ 885G 17 58,734,163 NA T A 68 31 NA 0 PPMID NM_003620:exon5:c.T1221A:p.C407X 17 58,740,374 NA TG T 106 22 NA 0 PPMID NM_003620:exon6:c.1279_1280T 17 58,740,467 NA C T 42 37 NA 0 PPMID NM_003620:exon6:c.C1372T:p.R458X 17 58,740,467 NA C T 73 55 NA 0 PPMID NM_003620:exon6:c.C1372T:p.R458X 17 58,740,507 NA CA C 98 31 NA 0 PPMID NM_003620:exon6:c.1412_1413C 17 58,740,525 NA AT A 82 32 NA 0 PPMID NM_003620:exon6:c.1430 1431A 17 58,740,532 NA T TA 40 66 NA 3 PPMID NM_003620:exon6:c.1437_1437delinsTA 17 58,740,543 NA C CT 97 31 NA 0 PPMID NM_003620:exon6:c.1448_1448delinsCT 17 58,749,560 NA TC T 79 18 NA 0 PPMID NM_003620:exon6:c.1465_1466T 17 58,740,623 NA C CA 71 21 NA 0 PPMID NM_003620:exon6:c.1528_1528delinsCA 17 58,740,668 NA G T 62 19 982224 0 PPMID NM_003620:exon6:c.G1573T:p.E525X 17 58,740,713 NA G T 47 12 NA 0 PPMID NM_003620:exon6.c.G1618T:p.E540X 17 58,740,809 NA C T 60 10 NA 0 PPMID NM_003620:exon6.c.C1714T:p.R572X 17 74,732,935 NA CGGCGGCT C 30 6 1318446 23 SRSF2 NM_003016:exon1:c.284_308G GTGGTGTG 146289 AGTCCGGG G 17 74,732,935 NA CGGCGGCT C 86 9 1318446 23 SRSF2 NM_003016:exon1:c.284_308G GTGGTGTG 146289 AGTCCGGG G 17 74,732,959 NA G C 41 22 211661 30 SRSF2 NM_003016:exon1:c.C284G:p.P95R 17 74,732,959 NA G C 48 19 211661 30 SRSF2 NM_003016:exon1:c.C284G:p.P95R 17 74,732,959 NA G C 50 19 211661 30 SRSF2 NM_003016:exon1:c.C284G:p.P95R 17 74,732,959 NA G T 34 15 211029 84 SRSF2 NM_003016:exon1:c.C284A:p.P95H 211504 211505 17 74,732,959 NA G T 37 10 211029 84 SRSF2 NM_003016:exon1:c.C284A:p.P95H 211504 211505 20 31,019,423 NA CA C 35 30 NA 0 ASXL1 NM_015338:exon9:c.920_921C 20 31,021,158 NA T A 52 14 NA 0 ASXL1 NM_015338:exon11:c.T1157A:p.L386X 20 31,021,295 NA C T 71 21 NA 0 ASXL1 NM_015338:exon11:c.C1294T:p.Q432X 20 31,021,542 NA CTG C 194 33 NA 0 ASXL1 NM_015338:exon11:c.1541_1543C 20 31,021,565 NA C T 160 104 NA 0 ASXL1 NM_015338:exon11:c.C1564T:p.Q522Y 20 31,021,622 NA C CGGCT 170 25 NA 0 ASXL1 NM_015338:exon11:c.1621_1621delinsCGGCT 20 31,022,286 NA T TA 74 15 36166 9 ASXL1 NM_015338:exon12:c.1771_1771delinsTA 20 31,022,402 NA TCACCACT T 12 6 36165 61 ASXL1 NM 015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 13 7 36165 61 ASXL1 NM_015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 16 8 36165 61 ASXL1 NM 015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 29 3 36165 61 ASXL1 NM 015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 29 3 36165 61 ASXL1 NM_015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 30 5 36165 61 ASXL1 NM_015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,402 NA TCACCACT T 39 8 36165 61 ASXL1 NM_015338:exon12:c.1887_1910T GCCATAGA 41597 GAGGCGGC 51200 20 31,022,414 NA TAG T 14 6 NA 0 ASXL1 NM_015338:exon12:c.1899_1901T 20 31,022,485 NA A AG 7 4 NA 0 ASXL1 NM_015338:exon12:c.1970_1970delinsAG 20 31,022,572 NA AGT A 35 9 146261 2 ASXL1 NM_015338:exon12:c.2057_2059A 20 31,022,592 rs373221034 C T 30 5 51388 11 ASXL1 NM_015338:exon12:c.C2077T:p.693X 20 31,022,592 rs373221034 C T 38 5 51388 11 ASXL1 NM_015338:exon12:c2077T:p.693X 20 31,022,624 NA TG T 43 11 266052 0 ASXL1 NM_015338:exon12:c.2109_2110T 20 31,022,624 NA T TC 60 14 1155825 1 ASXL1 NM_015338:exon12:c.2109_2109delinsTC 20 31,022,688 NA A T 24 8 NA 0 ASXL1 NM_015338:exon12:c.A2173T:p.R725X 20 31,022,708 NA AC A 30 10 NA 0 ASXL1 NM_015338:exon12:c.2193_2194A 20 31,022,898 NA TC T 39 11 1716903 4 ASXL1 NM_015338:exon12:c.2383_2384T 34212 20 31,022,922 NA C T 84 19 96380 1 ASXL1 NM_015338:exon12:c.C2407T:p.Q803X 20 31,022,981 NA AT A 96 71 NA 0 ASXL1 NM_015338:exon12:c.2466_2467A 20 31,022,991 NA G T 117 18 NA 0 ASXL1 NM_015338:exon12:c.G2476T:p.G826X 20 31,023,045 NA A AC 247 47 1411087 1 ASXL1 NM_015338:exon12:c.2530_2530delinsAC 41712 20 31,023,083 NA C A 306 65 NA 0 ASXL1 NM_015338:exon12:c.C2568A:p.C856X 20 31,023,209 NA G A 50 13 NA 0 ASXL1 NM_015338:exon12:c.G2694A:p.W898X 20 31,023,408 NA C T 52 14 267971 3 ASXL1 NM_015338:exon12:c.C2893T:p.R965X 20 31,023,473 NA C CGT 92 20 NA 0 ASXL1 NM_015338:exon12:c.2958_2958delinsCGT 20 31,023,717 NA C T 92 26 41715 4 ASXL1 NM_015338:exon12:c.C3202T:p.R1068X 20 31,024,273 NA G GC 40 38 NA 0 ASXL1 NM_015338:exon12:c.3758_3758delinsGC 20 31,025,057 NA CAT C 60 49 NA 0 ASXL1 NM_015338:exon12:c.4542_4544C 21 44,524,456 rs371769427 G A 26 5 1142948 33 U2AF1 NM_006758:exon2:c.C101T:p.S34F 166866
(165) TABLE-US-00004 TABLE S4 Cysteine mutations in the DNMT3A gene. DNMT3A mutations leading to the formation of new cysteine residues and predicted de novo disulfide bond formation. Number Disulfide Disulfide Mutation of subjects bonds Bond Score* G543C 1 524-543 0.99676 S714C 3 541-714 0.99651 F732C 1 497-732 0.97115 Y735C 4 520-735 0.30687 R736C 1 520-736 0.99095 R749C 2 749-818 0.99843 F751C 1 524-751 0.99811 W753C 1 554-753 0.72528 R882C 6 494-882 0.8412 L889C 1 818-889 0.99797 *DiANNA: unified software for Cysteine state and Disulfide Bond partner predition Note: Catalytic ADD-Domain amino acids 472-610
(166) TABLE-US-00005 TABLE S5 Counts for subjects with one putative somatic mutation and no candidate drivers (one mut.), subjects with exactly two putative somatic mutations and no candidate drivers (two muts.), subjects with clonal hematopoiesis with unknown drivers (CH-UD), subjects with clonal hematopoiesis with candidate drivers (CH-CD), and subjects with clonal hematopoiesis with candidate or unknown drivers (CH). Subjects were counted across all individuals for whom both age at sampling information and sequencing data of sufficient quality for detection of putative somatic mutations were available, with the exception of subject with CH-CD for whom only age at sampling information was required. Age one mut. two muts. CH-UD CH-CD CH 19-30 18/174 1/174 0/174 1/196 1/174 11-35 36/349 5/349 2/349 2/371 3/349 36-40 48/661 13/661 1/661 5/708 5/661 41-45 93/1081 15/1081 5/1081 6/1154 9/1081 46-50 120/1303 12/1303 5/1303 18/1378 22/1303 51-55 148/1597 28/1597 10/1597 76/1695 32/1597 56-60 190/1725 41/1725 19/1725 41/1815 58/1725 61-65 187/1608 40/1608 35/1608 56/1659 88/1608 66-70 141/1105 36/1105 32/1105 44/1140 76/1105 71-75 77/600 29/600 29/600 48/619 75/600 76-80 57/355 15/355 32/355 25/356 58/355 81-93 13/73 5/73 5/73 7/73 12/73
(167) TABLE-US-00006 TABLE S6 Subjects with clonal hematopoiesis and a diagnosis of hematologic malignancy after DNA sampling. There were 37 subjects diagnosed with hematologic malignancies after DNA sampling. Of these, 15 had showed clonal hematopoiesis in their initial DNA sample. Diagnoses of hematologic malignancies in these subjects followed DNA sampling by an average of 17 months (range: 2-36 months). Subjects with additional sequence generated to identify the malignancy are highlighted in bold. First diagnosis Subject Mutations Months Sex Age Died Candidate drivers Passengers after Type Male 62 Yes NA 3 32 Unspecified B-cell lymphoma, unspecified site Male 64 No NA 3 7 Multiple myeloma Male 70 Yes SF3B1 p.K700E 3 20 Chronic lymphocytic leukemia of B-cell type Female 63 No NA 3 11 Chronic lymphocytic leukemia of B-cell type Male 63 No NA 10 9 Chronic lymphocytic leukemia of B-cell type Female 72 Yes TP53 p.R248Q 3 34 Acute myeloblastic leukemia.sup.3 Male 73 Yes SRSF2 p.P95H 6 21 Acute mycloblastic leukemia Female 71 No SRSF2 p.P95H 1 9 Chronic myelomonocytic leukemia Male 64 No NA 3 2 Acute leukemia of unspecified cell type.sup.2 Female 73 Yes DNMT3A 0 36 Chronic leukemia of unspecified cell type p.V372A Female 61 Yes DNMT3A 1 11 Other myelodysplastic syndromes p.P904L Male 85 Yes SRSF2 p.P95H 13 2 Other myelodysplastic syndromes.sup.1 Male 69 No JAK2 p.V617F 2 35 Chronic myeloproliferative disease Female 76 No JAK2 p.V617F 4 13 Chronic myeloproliferative disease Male 57 No DNMT3A 0 14 Monoclonal gammopathy p.H613D .sup.1Subject #1 .sup.2Subject #2 (later progressed to acute myeloblastic leukemia) .sup.3Subject #3
(168) TABLE-US-00007 TABLE S7 Subjects with clonal hematopoiesis and a diagnosis of hematologic malignancy before DNA sampling. There were 55 subjects with a previous diagnosis of hematologic malignancy up to 12 years before DNA sampling. Of these, 14 showed clonal hematopoiesis. Previous history of hematologic malignancy was a strong risk factor for clonal hematopoiesis (OR = 6.0; 95% CI 3.1 to 12; P < 0.001, adjusting for age and sex using a linear regression model). First diagnosis Subject Mutations Months Sex Age Died Candidate drivers Passengers before Type Female 64 No NA 6 95 Hodgkin lymphoma, unspecified Female 72 Yes NA 18 148 Hodgkin lymphoma, unspecified Female 72 No DNMT3A 2 17 Follicular lymphoma, unspecified p.R556G Male 63 No DNMT3A 0 12 Diffuse large B-cell lymphoma p.R597W Male 76 Yes NA 3 52 Other non-follicular lymphoma, unspecified site Male 61 No DNMT3A 0 13 Other specified types of non-Hodgkin p.E907K lymphoma PPM1D frameshift Female 61 No DNMT3A 2 145 Acute leukemia of unspecified cell type p.G543C Male 57 No NA 3 1 Polycythemia vera Male 51 No JAK2 p.V617F 3 49 Polycythemia vera Male 70 No JAK2 p.V617F 1 46 Polycythemia vera Male 61 No JAK2 p.V617F 1 25 Polycythemia vera Male 77 Yes CBL p.Y371H 9 46 Other myelodysplastic syndromes U2AF1 p.S34F Male 57 No JAK2 p.V617F 5 4 Chronic myeloproliferative disease Female 56 No JAK2 P.V617F 0 20 Essential (hemorrhagic) thrombocytbemia
(169) TABLE-US-00008 TABLE S8 Subjects with clonal hematopoiesis at DNA sampling who died during follow-up. Subjects with additional sequence generated to identify the malignancy are highlighted in bold. Subject Mutations Death Sex Age Candidate Drivers Passengers Months after Cause Male 73 NA 3 7 Malignant neoplasm of sigmoid colon Male 67 DNMT3A 0 65 Malignant neoplasm of prostate p.Y908C Male 74 ASXL1 p.Q803X 1 30 Malignant neoplasm of prostate Male 76 NA 3 17 Unspecified B-cell lymphoma Female 72 NA 18 3 Unspecified Non-Hodgkin lymphoma Female 61 DNMT3A 1 18 Acute myeloblastic leukaemia [AML] p.P904L Female 72 TP53 p.R248Q 3 36 Acute myeloblastic leukaemia [AML] Male 73 SRSF2 p.P95R 6 26 Acute myeloblastic leukaemia [AML] Male 85 SRSF2 p.P95H 13 16 Unspecified leukemia Male 77 CBL p.Y371H 9 16 Myelodysplastic syndrome, unspecified U2AF1 p.S34F Male 78 NA 3 19 Anemia, unspecified Male 63 DNMT3A 0 6 Haemophagocytic syndrome, infection- frameshift associated Male 68 NA 4 14 Diabetes mellitus type 2 with renal complications Male 76 NA 5 6 Unspecified diabetes mellitus without complications Male 59 ASXL1 frameshift 0 6 Unspecified diabetes mellitus without complications Male 72 PPM1D p.E540X 5 4 Parkinson disease Male 66 NA 3 5 Anoxic brain damage, not elsewhere classified Female 64 JAK2 p.V617F 4 45 Acute myocardial infarction, unspecified Female 82 NA 5 37 Acute myocardial infarction, unspecified Male 59 DNMT3A 0 30 Acute myocardial infarction, unspecified p.E599D Female 74 NA 3 9 Atherosclerotic heart disease Female 64 DNMT3A 2 40 Pulmonary heart disease, unspecified p.F751C Male 73 SF3B1 p.K666T 8 12 Acute and subacute infective endocarditis TET2 frameshift Male 77 NA 5 10 Endocarditis, valve unspecified Male 77 NA 7 27 Heart failure, unspecified Female 80 NA 4 10 Heart failure, unspecified Female 65 PPM1D p.R458X 0 10 Cardiomegaly Female 75 TET2 frameshift 2 19 Subarachnoid haemorrhage unspecified Female 88 ASXL1 p.R965X 3 7 intracerebral haemorrhage, unspecified Male 67 NA 4 34 Stroke, not specified as haemorrhage or infarction Female 64 DNMT3A 3 24 Other specified cerebrovascular diseases p.C520X Male 81 DNMT3A 0 42 Sequelae of other and unspecified p.L344P cerebrovascular diseases Female 70 DNMT3A 0 48 Generalized and unspecified atherosclerosis p.R882H Male 54 NA 3 39 Generalized and unspecified atherosclerosis Male 66 DNMT3A 4 32 Generalized and unspecified atherosclerosis p.I681M Female 75 DNMT3A 0 7 Unspecified chronic bronchitis p.Q816X Male 68 NA 3 11 Chronic obstructive pulmonary disease, unspecified Female 62 ASXL1 frameshift 2 57 Chronic obstructive pulmonary disease, unspecified Male 74 NA 5 39 Chronic obstructive pulmonary disease, unspecified Female 65 DNMT3A 7 26 Chromc obstructive pulmonary disease, p.E733G unspecificd JAK2 p.V617F Male 57 NA 3 34 Gastro-oesophageal reflux disease with oesophagitis Male 51 DNMT3A 6 28 Other ill-defined and unspecified causes of p.M761V mortality Female 65 TET2 frameshift 3 15 Other ill-defined and unspecified causes of mortality Male 77 NA 4 26 Unspecified drowning and submersion Female 74 DNMT3A 3 44 Unknown p.P307R Male 64 SF3B1 p.K666N 7 48 Unknown Male 70 NA 3 52 Unknown Male 62 NA 3 43 Unknown Male 70 SF3B1 p.K700E 3 43 Unknown Male 74 ASXL1 frameshift 4 41 Unknown Female 73 DNMT3A 0 42 Unknown p.V372A Male 67 JAK2 p.V617F 4 44 Unknown Male 72 1DH2 p.R140W 2 37 Unknown SRSF2 frameshift Female 75 DNMT3A 0 27 Unknown p.Y735C
(170) TABLE-US-00009 TABLE S9 Somatic mutations for Subject #1. List of putative somatic mutations and candidate driver somatic mutations from whole-exome sequencing (WES) data and high coverage whole-genome sequencing (WGS) data of blood. Candidate driver somatic mutations are highlighted in bold. Subject #1 (diagnosed with myeloid malignancy 2 months after DNA sampling) Reference Alternate Reference Count Count Count Position dbSNP 138 or Refence Alternate (WES (WES (WGS Chromosome (GRCh37) COSMIC ID Allele Allele blood) blood) blood) 1 197,070,852 NA A G 82 23 115 2 242,178,077 NA T G 196 79 109 3 38,519,942 NA G A 65 18 107 3 46,306,703 NA T A 52 8 126 3 52,437,754 rs150524807 G A 52 8 124 4 106,162,527 NA T TTA 0 0 111 4 106,164,929 NA A G 0 0 126 4 158,284,236 NA C T 79 21 107 5 54,404,054 NA G A 74 10 108 6 50,696,983 COSM3354285 C T 160 44 105 11 67,265,009 NA C T 198 25 146 13 23,909,533 rs9552930 T C 75 17 107 14 92,472,207 NA G C 154 30 118 15 43,668,387 NA A T 110 32 139 17 74,732,959 COSM211029 G T 37 30 100 COSM211504 COSM211505 20 1,107,965 NA A G 196 29 103 20 31,022,441 COSM34210 A AG 10 3 89 COSM1411076 COSM1658769 21 36,259,198 COSM24719 AG A 57 3 127 COSM24728 X 123,191,828 NA G A 28 2 33 Alternate Count (WGS Chromosome blood) Gene Annotation 1 19 ASPM NM_018136:exon18:c.T7529C:p.12510T 2 33 HDLBP NM 005336:exon20:c.A2736C:p.R912S 3 35 ACPR2B NM_001106:exon5:c.G599A:p.R200H 3 21 CCR3 NM_001837:exon3:c.T54A:p.D18H 3 31 BAP1 NM_004656:exon13:c.C1407T:p.S469S 4 17 TET2 NM_001127208:exon4:c.3441_3441delinsTTA 4 22 TET2 NM_001127208:exon6:c.A3797G:p.N1266S 4 22 GRIA2 NA 5 23 GZMA NN_006144:exon4:c.G459A:p.W153X 6 28 TFAF2D NM_172233:exon5:c.C841T:p.R281W 11 18 PITPNM1 NM_004910:exon13:c.G1924A:p.E642K 13 38 SACS NM_014363:exon10:c.A8482G:p.S2828G 14 16 TRIP11 NM_004239:exon11:c.C2113G:p.L705V 15 31 TUBGCP4 NM_014444:exon2:c.A170T:p.E57V 17 20 SRSF2 NM_003016:exon1:c.C284A:p.P95H 20 20 PSMF1 NA 20 31 ASXL1 NM_015338:exon12:c.1926_1926delinsAG 21 17 RUNX1 NM_001754:exon4:c.292_293T X 12 STAG2 NM_001042750:exon15:c.1416+1G>A
(171) TABLE-US-00010 TABLE S10 Somatic mutations for Subject #2. List of putative somatic mutations and candidate driver somatic mutations from whole-exome sequencing (WES) data of blood, high coverage whole- genome sequencing (WGS) data of blood, and whole-exome sequencing data for bone marrow biopsy at the time of first diagnosis. Candidate driver somatic mutations are highlighted in bold. Table S10 discloses SEQ ID NO: 4 in the alternate allele column and SEQ ID NO: 5 in the annotation column. Subject #2 (diagnosed with AML 2 months after DNA sampling) Ref- Al- Ref- Al- Ref- Al- er- ter- dbSNP Ref- Al- er- ter- er- ter- ence nate 138 er- ter- ence nate ence nate Count Count Chro- or ence nate Count Count Count Count (WES (WES mo- Position COSMIC Al- Al- (WES (WES (WGS (WGS bone bone some (GRCh37) ID lele lele blood) blood) blood) blood) marrow) marrow) Gene Annotation 11 123,811,251 NA G A 91 20 150 11 36 14 OR5D5 NM_001001965: exon1: c.G928A: p.G310S 19 10,090,052 NA G A 182 52 149 14 140 38 COL5A3 NM_05719: exon38: c.C2754T: p.V918V 19 33,792,380 COSM27466 A ACCTT 42 0* 55 13* 85 7.sup.Λ CEBPA NM_004364: CTGCT exon1: GCGTC c.941_941 TCCAC delins GTTGC CCAAGCAGCGCA GCTGC ACGTGGAGACGC TTGG AGCAGAAGGT 19 33,793,111 COSM18539 CG C 0 0 92 7.sup.Λ 26 3* CEBPA NM_004364: COSM29127 exon1: COSM29220 c.210_211G 20 43,129,883 NA C T 109 18 136 16 110 22 SERINC3 NM_006811: exon9: c.G1114A: p.V372I *due to the size of this insertion, alternate allele count is dependent on sequencing reads length, 76 for WES blood, 151 for WGS blood, and 101 for WES bone marrow .sup.Λthis mutation was not automatically genotyped by Haplotype Caller from the Genome Analysis Toolkit due to low allelic count
(172) Having thus described in detail preferred embodiments of the present invention, it is to be understood that the invention defined by the above paragraphs is not to be limited to particular details set forth in the above description as many apparent variations thereof are possible without departing from the spirit or scope of the present invention.