1. Patent Search
  2. Details

REAGENT COMBINATION AND KIT FOR DETECTING LIVER CANCERS, AND USE THEREOF

20230102121 · 2023-03-30

    Inventors

    Cpc classification

    International classification

    Abstract

    Provided is a reagent combination for detecting a liver cancer. The reagent combination includes any one of the following detection reagents: 1) a detection reagent for detecting methylation level of at least one of the following methylation sites of the PAK1 gene: cg17202086, cg26996201, and cg18309286; 2) a detection reagent for detecting methylation level of at least one of the following methylation sites of the OTX1 gene: cg23229261 and cg10122865; and 3) a detection reagent for detecting methylation level of methylation site cg16657538 of the ZNF397OS gene. Further provided are use of the reagent combination, and a kit including the reagent combination. By using the reagent combination of the present invention, a liver cancer may be detected clinically in a tissue sample with fewer markers, and may be detected sensitively and specifically by detecting methylation sites, and the cost and time are saved.

    Claims

    1. A reagent combination for detecting a liver cancer, comprising any one of the following detection reagents: 1) a detection reagent for detecting methylation level of at least one of the following methylation sites of a PAK1 gene: cg17202086, cg26996201, and cg18309286; 2) a detection reagent for detecting methylation level of at least one of the following methylation sites of a OTX1 gene: cg23229261 and cg10122865; and 3) a detection reagent for detecting methylation level of cg16657538 methylation site of ZNF397OS gene.

    2. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 1), the reagent combination further comprises a detection reagent for detecting methylation level of at least one of the following genes: GRASP, ZNF397OS, PPFIA1, and OTX1.

    3. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 1), the reagent combination further comprises a detection reagent for detecting methylation level of at least two of the following genes: GRASP, ZNF3970S, PPFIA1, and OTX1.

    4. The reagent combination according to claims 1, wherein when the reagent combination comprises the above detection reagent 1), the reagent combination further comprises a detection reagent for detecting methylation level of at least three of the following genes: GRASP, ZNF397OS, PPFIA1, and OTX1.

    5. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 1), the reagent combination further comprises a detection reagent for detecting methylation level of the following four genes: GRASP, ZNF397OS, PPFIA1, and OTX1.

    6. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination comprises a detection reagent for detecting methylation level of the following methylation sites of the OTX1 gene: cg23229261, cg10122865.

    7. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination further comprises a detection reagent for detecting methylation level of the methylation site cg21472506 of the OTX1 gene.

    8. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination further comprises a detection reagent for detecting methylation level of at least one of the following genes: GRASP, PAK1, PPFIA1, and ZNF397OS.

    9. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination further comprises a detection reagent for detecting methylation level of at least two of the following genes: GRASP, PAK1, PPFIA1, and ZNF397OS.

    10. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination further comprises a detection reagent for detecting methylation level of at least three of the following genes: GRASP, PAK1, PPFIA1, and ZNF397OS.

    11. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 2), the reagent combination further comprises a detection reagent for detecting methylation level of the following four genes: GRASP, PAK1, PPFIA1, and ZNF397OS.

    12. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 3), the reagent combination further comprises a detection reagent for detecting methylation level of at least one of the following genes: GRASP, PAK1, PPFIA1, and OTX1.

    13. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 3), the reagent combination further comprises a detection reagent for detecting methylation level of at least two of the following genes: GRASP, PAK1, PPFIA1, and OTX1.

    14. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 3), the reagent combination further comprises a detection reagent for detecting methylation level of at least three of the following genes: GRASP, PAK1, PPFIA1, and OTX1.

    15. The reagent combination according to claim 1, wherein when the reagent combination comprises the above detection reagent 3), the reagent combination further comprises a detection reagent for detecting methylation level of the following four genes: GRASP, PAK1, PPFIA1, and OTX1.

    16. The reagent combination according to claim 1, wherein the detection reagent is any one or more selected from the group consisting of: a nucleic acid primer, a sequencing Tag sequence, a methylation chip, and a nucleic acid probe.

    17. The reagent combination according to claim 1, wherein the reagent combination further comprises a reagent for extracting plasma cell-free DNA.

    18. A method for detecting a liver cancer, comprising detecting methylation level of sample DNA using the reagent combination according to claim 1.

    19. A kit for detecting a liver cancer, comprising the reagent combination according to claim 1.

    20. The kit according to claim 19, wherein the kit further comprises other reagents, comprising at least one selected from the group consisting of: a reagent for extracting nucleic acid, a reagent for purifying nucleic acid, bisulfate, T4 polymer nucleotide kinase and T4 ligase.

    Description

    BRIEF DESCRIPTION OF DRAWINGS

    [0083] FIG. 1 shows the ROC diagram of ZNF397OS single gene for identification of cancer and non-cancer in peripheral blood cell-free DNA samples;

    [0084] FIG. 2 shows the methylation levels of ZNF397OS gene in different groups of peripheral blood cell-free DNA samples;

    [0085] FIG. 3 shows the methylation levels of the methylation sites of ZNF397OS gene in different groups of tissue samples;

    [0086] FIG. 4 shows the methylation levels of methylation sites outside the target region of ZNF397OS gene (as comparative examples) in different groups of tissue samples;

    [0087] FIG. 5 is the ROC diagram of PAK1 methylation sites for identification of cancer and non-cancer in cell-free DNA samples of peripheral blood;

    [0088] FIG. 6 shows the methylation levels of PAK1 gene methylation sites in different groups in cell-free DNA samples of peripheral blood;

    [0089] FIG. 7 shows the methylation levels of PAK1 gene methylation sites in different groups of tissue samples;

    [0090] FIG. 8 shows the methylation levels of methylation sites outside the target region of PAK1 gene (as comparative examples) in different groups of tissue samples;

    [0091] FIG. 9 is the ROC diagram of OTX1 gene methylation sites for identification of cancer and non-cancer in cell-free DNA samples of peripheral blood;

    [0092] FIG. 10 shows the methylation levels of OTX1 gene methylation sites in different groups of peripheral blood cell-free DNA samples;

    [0093] FIG. 11 is a comparison of methylation levels of OTX1 gene methylation sites in different groups of tissue samples;

    [0094] FIG. 12 is a comparison of methylation levels of methylation sites outside the target region of OTX1 gene (as comparative examples) in different groups of tissue samples.

    DETAILED DESCRIPTION OF EMBODIMENTS

    [0095] The present invention will be described in detail below with reference to particular embodiments and examples, and the advantages and various effects of the present invention will be more clearly presented therefrom. It should be understood by those skilled in the art that these particular embodiments and examples are used to illustrate the present invention, but not to limit it.

    Example 1: Screening of Methylated Genes

    [0096] In present invention, the methylation data of a total of 785 cancer tissues, 461 para-cancer tissues or normal control tissues, and 656 healthy whole blood samples are collected from the TCGA database (https://tcga.xenahubs.net) of the UCSC Xena website and the GEO database of the American National Center for Biotechnology Information (NCBI). Differential analysis of the data of the liver cancer and control is performed, annotating the physical location and gene information of the differential sites. In order to ensure that the screening fragments have consistent methylation levels, the screening of methylated gene fragments needs to meet the following requirements: 1) the selected gene fragments are required to have no less than 2 adjacent sites with consistent methylation levels; 2) differential analysis is performed between the liver cancer and para-cancer tissue or normal control tissue, and methylation gene fragments with high consistency and difference in liver cancer samples are selected as candidate target genes; 3) differential analysis on the methylation detection data of whole blood of liver cancer and healthy samples, and the gene fragments that are differentially hypermethylated in liver cancer are selected; 4) finally, the methylation sites are analyzed one by one, thereby obtaining the candidate methylation sites.

    Example 2: Detection of Gene Methylation Levels in Clinical Samples

    [0097] 10 ml of peripheral blood for each sample is collected for detection and analysis of methylation level of DNA methylation markers in the samples. The experimental procedure is as follows:

    [0098] 1. Sample Preparation

    [0099] The sample of the present invention is prepared by extracting 4 ml of plasma with MagMAX™ Cell-Free DNA Isolation Kit, and eluting with 45 μL of eluent. The extracted cell-free nucleic acids must meet the following quality control conditions: the total amount of extracted nucleic acids is greater than 20 ng.

    [0100] 2. Library Preparation

    [0101] In the present invention, all cell-free nucleic acids qualified for quality control are treated with bisulfite by using EZ DNA Methylation-Lightning™ Kit (Zymo Research, Irvine, Calif., USA). Subsequently, the sample DNA after bisulfite treatment is used to construct a pre-library by the single-strand library construction method. After passing the quality inspection of the pre-library, the target region is captured and enriched by liquid chip hybridization to complete the construction of the final library.

    [0102] Construction steps of pre-library: 1) phosphorylation: T4 polynucleotide kinase is used to phosphorylate the 5′-end of the bisulfite-treated DNA; 2) SS1 ligation: T4 DNA Ligase (Rapid) is used to ligate the SS1 linker to the 5′-end of phosphorylated DNA; 3) nucleic acid purification: 2 volumes of Agencourt AMPure XP system (Beckman CouLter, CA, USA) is used to remove the remaining linkers; 4) SS2 ligation: T4 DNA Ligase (Rapid) is used to ligate the SS2 linker to the 3′-end of phosphorylated DNA; 5) nucleic acid purification: 2 volumes of Agencourt AMPure XP system (Beckman CouLter, CA, USA) is used to remove the remaining linkers; 6) amplification: NEBNext Q5U Master Mix, primer1.0 (universal primers) and Bacard sequences are used to amplify the nucleic acids in the previous step; 7) nucleic acid purification: 1.2 volume of Agencourt AMPure XP system (Beckman CouLter, CA, USA) is used to remove primer dimers and excess primers; 8) quality inspection: the pre-library after purification is checked the total amount of the library by Qubit® dsDNA HS Assay Kit (Life Technologies, CA, USA), and the quality inspection of the library fragment distribution is performed by LabChip GXII Touch.

    [0103] Capture steps of chip (Twist Bioscience) hybridization: 1) chip hybridization: 1.5 μg of the library mixed well after passing the quality inspection is subjected to vacuum concentration into powder in advance, then mixing well with Panel, Hybridization Mix, Blocker Solution, Universal Blockers, and Hybridization Enhancer reagents (the reagents used for chip hybridization are all provided by Twist Bioscience), placing in a PCR amplifier to incubate for 16 h overnight at 70 degrees (the temperature of the hot lid is 85 degrees); 2) magnetic bead capture: Streptavidin Binding Buffer is used to wash the capture magnetic beads 3 times in advance, adding the hybridized product to the capture magnetic beads to incubate for 30 min, washing once with Wash Buffer I, washing 3 times with Wash Buffer 2, then eluting with 42 μl of ultra-pure water; 3) amplification: KAPA HiFi HotStart ReadyMix and universal primers are used to amplify the captured library; 4) purification: 1 volume of Agencourt AMPure XP system (Beckman CouLter, CA, USA) is used to remove primer dimers and excess primers.

    [0104] As for the purified library, the quality inspection of the total amount of nucleic acid, fragment distribution and primer-dimer ratio in the library is performed by Qubit® dsDNA HS Assay Kit (Life Technologies, CA, USA) and LabChip GXII Touch.

    [0105] 3. Sequencing

    [0106] The libraries after passing the quality inspection of the total amount of library, fragment size distribution of the amplified products and primer-dimer ratio are used as the libraries to be tested and mixed according to the mass of 1:1, Qubit® dsDNA HS Assay Kit (Life Technologies, CA, USA) is used to perform accurate quantification of the mixed library, the library is denatured and diluted, then sequencing on the NextSeq500 desktop sequencer by using PE75.

    [0107] 4. Establishment and Evaluation of Liver Cancer Classification Model

    [0108] For the raw fastq data obtained by sequencing, after filtering the raw data, bismark methylation analysis software is used to perform methylation analysis on the fragments captured by the chip, thereby obtaining methylation level of a single site of the candidate genes and methylation level of the gene fragments. Methylation level of a single site of candidate genes and methylation level of gene fragments are used to perform differential analysis and model construction for liver cancer and the control samples. In the present invention, liver cancer classification model is constructed and evaluated by Logistic regression analysis of the data.

    Example 3: Detection of Methylation Level of Clinical Samples by the Reagent Combination of the Present Invention

    [0109] Samples from 63 patients with primary liver cancer, 25 patients with cirrhosis, 15 patients with hepatitis and 7 healthy people are collected, and methylation level of the methylation site cg16657538 of ZNF397OS gene, and methylation levels of the combinations of the ZNF397OS gene and the rest of the genes in the samples are detected and analyzed according to the method described in Example 2, thereby verifying the effect of detecting a liver cancer. The test results are shown in Table 1, FIG. 1 and FIG. 2.

    TABLE-US-00007 TABLE 1 Prediction performance of the combinations of ZNF397OS gene and other genes in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 5 0.9410 0.9100 0.9500 0.8948 0.7619 0.9574 1 + 5 0.9603 0.9185 0.9393 0.8997 0.8095 0.9574 2 + 5 0.9396 0.8994 0.9176 0.8862 0.8095 0.8936 3 + 5 0.9438 0.8994 0.9197 0.8845 0.8413 0.9362 4 + 5 0.9598 0.9134 0.9501 0.9206 0.873 0.9787 1 + 2 + 5 0.9679 0.9261 0.9544 0.9098 0.8413 0.9787 1 + 3 + 5 0.9647 0.9236 0.9631 0.9119 0.8254 0.9787 1 + 4 + 5 0.9663 0.9325 0.961 0.9348 0.873 0.9787 2 + 3 + 5 0.944 0.8981 0.9219 0.8855 0.8413 0.8936 2 + 4 + 5 0.9634 0.9134 0.9436 0.9237 0.8571 0.9149 3 + 4 + 5 0.9667 0.921 0.9544 0.9254 0.8571 0.9362 1 + 2 + 3 + 5 0.9682 0.9248 0.961 0.9092 0.8413 0.9787 1 + 2 + 4 + 5 0.9723 0.9414 0.9653 0.9433 0.8889 0.9574 1 + 3 + 4 + 5 0.9719 0.9338 0.9675 0.9331 0.873 0.9787 2 + 3 + 4 + 5 0.9669 0.9197 0.9544 0.9264 0.8571 0.9149 1 + 2 + 3 + 4 + 5 0.9739 0.9376 0.9696 0.9541 0.8889 0.9574

    [0110] Wherein, 1 represents the GRASP gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 5); 2 represents the PAK1 gene (detecting the average methylation of the fragment containing four methylation sites cg17202086, cg26996201, cg12269002 and cg18309286); 3 represents the PPFIA1 gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 6); 4 represents the OTX1 gene (detecting the average methylation of the fragment containing three methylation sites cg21472506, cg23229261 and cg10122865); and 5 represents the ZNF3970S gene (detecting the average methylation level of the fragment containing methylation site cg16657538).

    [0111] As can be seen from Table 1 that, in tissue samples, all reagent combinations are able to predict liver cancer with a specificity of at least 0.9176 and a sensitivity of 0.8981 and an area under the curve of 0.9396. while in plasma cell-free DNA samples, all reagent combinations are able to predict liver cancer with a specificity of at least 0.8936 and a sensitivity of 0.7619 and an area under the curve of 0.8845. Therefore, the reagent combinations of the present invention have a good predictive effect on liver cancer, and in particular when plasma cell-free DNA is used as a sample, they have excellent specificity and sensitivity.

    Comparative Example 1

    [0112] In order to investigate whether other methylation sites on ZNF397OS gene may uniformly be used as markers for detecting a liver cancer, another methylation site in ZNF397OS gene is selected as a comparative example (all of which are located outside the currently selected target fragment of ZNF397OS gene), comparing the methylation differences in cancer and non-cancer samples; the results are shown in FIG. 3 and FIG. 4. It can be seen from FIG. 3 and FIG. 4 that, the methylation sites upstream and downstream of the selected target fragment of ZNF397OS gene have no significant difference between liver cancer and non-cancer tissues, and are not suitable to be used as markers for detecting a liver cancer.

    Comparative Example 2

    [0113] In order to investigate whether the ZNF397OS gene may be used as a marker for detecting a liver cancer, the genes PLAC8 and ATXN1, which are closely related to liver cancer known in the art, are further selected (see Xu RH, Wei W, Krawczyk M, Wang W, Luo H, Flagg K, et al. al. Circulating tumor DNA methylation markers for diagnosis and prognosis for diagnosis and prognosis of hepatocellular carcinoma. Nature Materials, 2017, and Chinese patent CN106947830B). The detection is also performed according to the method of the above example, and the detection results are shown in Table 3 below.

    TABLE-US-00008 TABLE 3 Prediction performance of the comparative genes and their combinations in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 6 0.578 0.944 0.004 0.750 0.683 0.681 7 0.858 0.813 0.711 0.722 0.746 0.574 6 + 7 0.872 0.827 0.722 0.883 0.810 0.809

    [0114] Wherein, 6 represents the PLAC8 gene (detecting the methylation level of cg11606215 of this gene), and 7 represents the ATXN1 gene (detecting the methylation level of cg24067911 of this gene).

    [0115] As can be seen from Table 3 that, in tissue samples, the highest area under the curve of a single comparative gene is 0.858, while in cell-free DNA samples, the highest area under the curve is 0.75, which are lower than the area under the curve of ZNF397OS of the present invention. The area under the curve of the combination is also lower than the value of the area under the curve of the combination of the agents of the present invention.

    Example 4: Detection of Methylation Level of Clinical Samples by the Reagent Combination of the Present Invention

    [0116] Samples from 63 patients with primary liver cancer, 25 patients with liver cirrhosis, 15 patients with hepatitis, and 7 healthy people are collected, and the methylation sites of PAK1 gene in the samples are detected and analyzed according to the method described in Example 2, thereby verifying their effect in detecting a liver cancer. The diagnosis of hepatocellular carcinoma, liver cirrhosis, hepatitis and healthy people is based on the final hospital pathological diagnosis. The test results are shown in Table 4, FIG. 5 and FIG. 6.

    TABLE-US-00009 TABLE 4 Prediction performance of methylation sites of PAK1 gene in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Methylation site AUC Sensitivity Specificity AUC Sensitivity Specificity cg17202086 0.8607 0.865 0.655 0.8352 0.7619 0.8936 cg26996201 0.8194 0.903 0.360 0.8333 0.7778 0.8936 cg18309286 0.8729 0.871 0.625 0.8286 0.7460 0.8511 cg17202086 + 0.853 0.843 0.662 0.833 0.778 0.894 cg26996201 cg17202086 + 0.876 0.870 0.651 0.838 0.778 0.872 cg18309286 cg26996201 + 0.884 0.870 0.716 0.837 0.794 0.894 cg18309286 cg17202086 + 0.887 0.868 0.729 0.839 0.778 0.872 cg26996201 + cg18309286

    [0117] As can be seen from Table 4 that, in tissues, each methylation site of the PAK1 gene may predict liver cancer with a specificity of at least 0.360, a sensitivity of 0.843, and an area under the curve of 0.8194. while in plasma cell-free DNA, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.8511, a sensitivity of 0.7460, and an area under the curve of 0.833.

    Example 5: Detection of methylation Level of Clinical Samples by the Reagent Combinations of the Present Invention

    [0118] Samples from 63 patients with primary liver cancer, 25 patients with liver cirrhosis, 15 patients with hepatitis and 7 healthy people are collected, and the methyl levels of the combinations of PAK1 gene and other genes in the samples are detected and analyzed according to the method described in Example 2, thereby verifying their effect on detecting a liver cancer. The test results are shown in Table 5 below.

    TABLE-US-00010 TABLE 5 Prediction performance of the combinations of PAK1 gene and other genes in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 1 + 2 0.9508 0.9096 0.961 0.8625 0.7937 0.9787 2 + 5 0.9396 0.8994 0.9176 0.8862 0.8095 0.8936 2 + 3 0.8944 0.893 0.6594 0.8578 0.8095 0.9149 2 + 4 0.9398 0.8904 0.9024 0.9132 0.8571 0.9149 1 + 2 + 5 0.9679 0.9261 0.9544 0.9098 0.8413 0.9787 1 + 2 + 3 0.9523 0.9108 0.9675 0.874 0.7937 0.9787 1 + 2 + 4 0.963 0.9223 0.9631 0.9301 0.873 0.9787 2 + 3 + 5 0.944 0.8981 0.9219 0.8855 0.8413 0.8936 2 + 4 + 5 0.9634 0.9134 0.9436 0.9237 0.8571 0.9149 2 + 3 + 4 0.9461 0.8955 0.9067 0.9176 0.8571 0.9149 1 + 2 + 3 + 5 0.9682 0.9248 0.961 0.9092 0.8413 0.9787 1 + 2 + 4 + 5 0.9723 0.9414 0.9653 0.9433 0.8889 0.9574 1 + 2 + 3 + 4 0.9657 0.9248 0.9696 0.9362 0.873 0.9787 2 + 3 + 4 + 5 0.9669 0.9197 0.9544 0.9264 0.8571 0.9149 1 + 2 + 3 + 4 + 5 0.9739 0.9376 0.9696 0.9541 0.8889 0.9574

    [0119] Wherein, 1 represents the GRASP gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 5); 2 represents the PAK1 gene (detecting the methylation levels of the three methylation sites cg17202086, cg26996201, and cg18309286); 3 represents the PPFIA1 gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 6); 4 represents the OTX1 gene (detecting the methylation levels of three methylation sites cg21472506, cg23229261 and cg10122865); and 5 represents the ZNF397OS gene (detecting the methylation level of the methylation site cg16657538).

    [0120] It can be seen in Table 5 that, in tissue samples, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.9176, a sensitivity of 0.8904, and an area under the curve of 0.8944; while in plasma cell-free DNA samples, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.7937, a sensitivity of 0.8578, and an area under the curve of 0.8578. Therefore, the reagent combinations of the present invention have a good predictive effect on liver cancer, and in particular when plasma cell-free DNA is used as a sample, they have excellent specificity and sensitivity.

    Comparative Example 3

    [0121] In order to investigate whether other methylation sites on PAK1 gene may uniformly be used as markers for detecting a liver cancer, the methylation sites upstream and downstream of the selected target fragment in PAK1 gene are selected, comparing their methylation differences in cancer and non-cancer samples; the results are shown in FIG. 7 and FIG. 8. It can be seen from FIG. 7 and FIG. 8 that, the methylation sites upstream and downstream of the selected target fragment in the PAK1 gene have no significant difference between cancer and non-cancer tissues, and are not suitable to be used as markers for detecting a liver cancer.

    Comparative Example 4

    [0122] In order to investigate whether the PAK1 gene may be used as a marker for detecting a liver cancer, we further selected the common liver cancer-related genes PLAC8 and ATXN1 known in the art (see Xu RH, Wei W, Krawczyk M, Wang W, Luo H, Flagg K, et al. al. Circulating tumor DNA methylation markers for diagnosis and prognosis for diagnosis and prognosis of hepatocellular carcinoma. Nature Materials, 2017, and Chinese patent CN106947830B). The detection is also performed according to the method of the above example, and the detection results are shown in Table 6 below.

    TABLE-US-00011 TABLE 6 Prediction performance of the comparative genes and their combinations in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 6 0.578 0.944 0.004 0.750 0.683 0.681 7 0.858 0.813 0.711 0.722 0.746 0.574 6 + 7 0.872 0.827 0.722 0.883 0.810 0.809

    [0123] Wherein, 6 represents the PLAC8 gene (detecting the methylation level of cg11606215 of this gene), and 7 represents the ATXN1 gene (detecting the methylation level of cg24067911 of this gene).

    [0124] It can be seen from Table 6 that, in tissues, the highest area under the curve of the single control gene is 0.858, and the highest area under the curve of the cell-free DNA is 0.75, which are lower than the area under the curve of the PAK1 gene of the present invention. The area under the curve of the combination is also lower than the value of the area under the curve of the combination of the agents of the present invention.

    Example 6: Detection of methylation Level of Clinical Samples by the Reagent Combination of the Present invention

    [0125] Samples from 63 patients with primary liver cancer, 25 patients with liver cirrhosis, 15 patients with hepatitis, and 7 healthy people are collected, and methylation level of the OTX1 gene in the samples is detected and analyzed according to the method described in Example 2, thereby verifying its effect on detecting a liver cancer. Diagnosis of hepatocellular carcinoma, cirrhosis, hepatitis and healthy people is based on the final hospital pathological diagnosis. The test results are shown in Table 7, FIG. 9 and FIG. 10.

    TABLE-US-00012 TABLE 7 Prediction performance of OTX1 gene methylation sites in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Methylation site AUC Sensitivity Specificity AUC Sensitivity Specificity cg23229261 0.929 0.884 0.885 0.8828 0.7937 0.9574 cg10122865 0.914 0.860 0.846 0.8823 0.8095 0.8298 cg23229261 + 0.934 0.882 0.885 0.888 0.810 0.957 cg10122865 cg21472506 + 0.941 0.890 0.892 0.887 0.778 0.957 cg23229261 cg21472506 + 0.938 0.887 0.907 0.889 0.810 0.936 cg10122865 cg21472506 + 0.942 0.887 0.896 0.888 0.794 0.957 cg23229261 + cg10122865

    [0126] It can be seen from Table 7 that, in tissues, each methylation site of the OTX1 gene may predict a liver cancer with a specificity of at least 0.846, a sensitivity of 0.860, and an area under the curve of 0.914; while in plasma cell-free DNA, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.8298, a sensitivity of 0.778, and an area under the curve of 0.8823.

    Example 7: Detection of methylation Level of Clinical Samples by the Reagent Combination of the Present Invention

    [0127] Samples from 63 patients with primary liver cancer, 25 patients with liver cirrhosis, 15 patients with hepatitis and 7 healthy people are collected, and methylation levels of the combinations of OTX1 gene and other genes in the samples are detected and analyzed according to the method described in Example 2, thereby verifying their effect on detecting a liver cancer. The test results are shown in Table 8 below:

    TABLE-US-00013 TABLE 8 Prediction performance of the combinations of OTX1 gene and other genes in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 1 + 4 0.9625 0.9236 0.9631 0.9237 0.873 0.9787 2 + 4 0.9398 0.8904 0.9024 0.9132 0.8571 0.9149 3 + 4 0.9433 0.8968 0.9024 0.9169 0.873 0.8723 4 + 5 0.9598 0.9134 0.9501 0.9206 0.873 0.9787 1 + 2 + 4 0.963 0.9223 0.9631 0.9301 0.873 0.9787 1 + 3 + 4 0.966 0.9248 0.9696 0.922 0.873 1 1 + 4 + 5 0.9663 0.9325 0.961 0.9348 0.873 0.9787 2 + 3 + 4 0.9461 0.8955 0.9067 0.9176 0.8571 0.9149 2 + 4 + 5 0.9634 0.9134 0.9436 0.9237 0.8571 0.9149 3 + 4 + 5 0.9667 0.921 0.9544 0.9254 0.8571 0.9362 1 + 2 + 3 + 4 0.9657 0.9248 0.9696 0.9362 0.873 0.9787 1 + 2 + 4 + 5 0.9723 0.9414 0.9653 0.9433 0.8889 0.9574 1 + 3 + 4 + 5 0.9719 0.9338 0.9675 0.9331 0.873 0.9787 2 + 3 + 4 + 5 0.9669 0.9197 0.9544 0.9264 0.8571 0.9149 1 + 2 + 3 + 4 + 5 0.9739 0.9376 0.9696 0.9541 0.8889 0.9574

    [0128] Wherein, 1 represents the GRASP gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 5); and 2 represents the PAK1 gene (detecting methylation levels of four methylation sites cg17202086, cg26996201, cg12269002, and cg18309286); 3 represents the PPFIA1 gene (detecting the average methylation level of the fragment represented by SEQ ID NO: 6); 4 represents the OTX1 gene (detecting the average methylation levels of the three methylation sites cg21472506, cg23229261, cg10122865); 5 represents the ZNF3970S gene (detecting methylation level of the methylation site cg16657538).

    [0129] It can be seen from the table that, in tissue samples, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.9024, a sensitivity of 0.8904 and an area under the curve of 0.9398; while in plasma cell-free DNA samples, all reagent combinations are able to predict a liver cancer with a specificity of at least 0.8723, a sensitivity of 0.8571, and an area under the curve of 0.9132. Therefore, the reagent combinations of the present invention have a good predictive effect on detecting a liver cancer, and in particular when plasma cell-free DNA is used as a sample, they have excellent specificity and sensitivity.

    Comparative Example 5

    [0130] In order to investigate whether other methylation sites on OTX1 gene may uniformly be used as markers for the detection of liver cancer, the methylation sites upstream of the selected target fragment in OTX1 gene are selected, comparing the methylation differences in cancer and non-cancer samples; and the results are shown in FIG. 11 and FIG. 12. It can be seen from FIG. 11 and FIG. 12 that, the methylation sites upstream of the selected target fragment in the OTX1 gene have no significant difference between cancerous and non-cancerous tissues, and they are not suitable to be used as markers for detecting a liver cancer.

    Comparative Example 6

    [0131] In order to investigate whether the OTX1 gene may be used as a marker for detecting a liver cancer, we further select the common liver cancer-related genes PLAC8 and ATXN1 known in the art (see Xu RH, Wei W, Krawczyk M, Wang W, Luo H, Flagg K, et al. al. Circulating tumor DNA methylation markers for diagnosis and prognosis for diagnosis and prognosis of hepatocellular carcinoma. Nature Materials, 2017, and Chinese patent CN106947830B). The detection is also performed according to the method of the above example, and the detection results are shown in Table 9 below.

    TABLE-US-00014 TABLE 9 Prediction performance of the comparative genes and their combinations in the Logistics liver cancer classification model Data of the tissue Data of the cell-free DNA Gene combination AUC Sensitivity Specificity AUC Sensitivity Specificity 6 0.578 0.944 0.004 0.750 0.683 0.681 7 0.858 0.813 0.711 0.722 0.746 0.574 6 + 7 0.872 0.827 0.722 0.883 0.810 0.809

    [0132] Wherein, 6 represents the PLAC8 gene (detecting methylation level of cg11606215 of this gene), and 7 represents the ATXN1 gene (detecting methylation level of cg24067911 of this gene).

    [0133] It can be seen from Table 9 that, in the tissue samples, the highest area under the curve of the single control gene is 0.858, while in the cell-free DNA samples, the highest area under the curve is 0.75, which are lower than the area under the curve of the OTX1 gene of the present invention. The area under the curve of the gene combination model is also lower than the value of the area under the curve of the reagent-related combination model of the present invention.