Method for screening methylation-detection-based marker, and model

Abstract

The present application provides a method for screening a methylation-detection-based marker. The method comprises: the step of acquiring first target data, wherein the first target data is sequencing data for a first population; the step of acquiring second target data, wherein the second target data is sequencing data for a second population; the step of determining a candidate region, involving determining M CpG-enriched candidate regions on the basis of the first target data and the second target data; the step of acquiring third target data, wherein the third target data comprises a methylation haplotype of each candidate region among the M candidate regions in the first target data and the second target data, and the number of times that each haplotype appears; and the step of determining a methylation-detection-based marker from the candidate regions, involving: on the basis of a difference in the third target data between the first population and the second population, determining from the candidate regions a marker that can be used for distinguishing between the first population and the second population, wherein the first population and the second population are two types of populations having different features that do not overlap.

Description

A method and model for screening markers based on methylation detection Technical Field

[0001] The present application relates to the field of biomedicine, and in particular to a method for screening markers based on methylation detection, a method for constructing a model for distinguishing a first population from a second population using markers, a marker for detecting liver cancer, and a model for screening markers based on methylation detection. Background Art

[0002] Gene expression in eukaryotes is regulated by a variety of mechanisms and at multiple levels. Changes in gene expression levels and functions without changing the DNA sequence of a gene can be heritable, a phenomenon known as epigenetic inheritance. Epigenetic regulatory mechanisms primarily include DNA methylation, histone modification, and the action of non-coding RNA, among which DNA methylation is the most well-studied epigenetic modification. There are three types of methylation states in eukaryotic cells: persistent hypomethylation (such as methylation of housekeeping genes), induced demethylation (such as modification of some developmental stage-specific genes), and hypermethylation (such as methylation of the constricted and inactivated X chromosome in human female cells). The biological functions of DNA methylation involve multiple aspects. For example: 1) DNA methylation affects gene expression regulation. DNA methylation provides an effective inhibitory mechanism for the long-term silencing of non-coding regions (such as introns). Methylation of CpG sites within gene promoter regions affects gene transcriptional activity in three ways: direct DNA sequence methylation impedes the binding of transcription factors; methyl-CpG binding proteins bind to methylated CpG sites and interact with other transcriptional repressors; and condensation of chromatin structure hinders the binding of transcription factors to their regulatory sequences. Second, DNA methylation influences tumorigenesis. Altered DNA methylation status is commonly found in tumors, characterized by decreased global methylation levels and increased localized methylation levels. In tumor cells, oncogenes are activated by hypomethylation, while tumor suppressor genes are repressed by hypermethylation. Recently, DNA methylation has been shown to be tissue-specific, potentially enabling early disease detection. Regions in the human genome with CpG sequence densities 10-20 times higher than average, GC content greater than 50%, and lengths greater than 200 base pairs are termed CpG islands. There are approximately 28,890 CpG islands in the human genome, with most chromosomes containing 5-15 CpG islands per megabase (Mb), for an average of 10.5 CpG islands per Mb. These high-CG regions involve the promoter regions of many genes and regulate the expression of downstream genes through methylation and demethylation.

[0003] Summary of the Invention

[0004] As the cost of high-throughput sequencing decreases, methylation detection in the NGS (Next Generation Sequencing) field is becoming more and more popular. Based on the second generation sequencing, the methylation data of all CpGs in the target region can be obtained. Conventional methods for calculating methylation,

[0005] This method ignores the role of methylation haplotypes in disease development and progression, and is extremely dependent on sample size, achieving relatively stable results only with large sample sizes. In the field of cancer detection, cfDNA (circulating free DNA) often contains very small amounts of ctDNA (circulating tumor DNA), further complicating detection.

[0006] The technical solution of this application is as follows:

[0007] 1. A method for screening markers based on methylation detection, comprising:

[0008] a step of acquiring first target data, wherein the first target data is sequencing data for a first population;

[0009] a step of acquiring second target data, wherein the second target data is sequencing data for a second population;

[0010] a step of determining candidate regions, determining M CpG-enriched candidate regions based on the first target data and the second target data;

[0011] The step of obtaining third target data includes:

[0012] The methylation haplotype of each candidate region in the M candidate regions in the first target data and the second target data, and the number of occurrences of each haplotype;

[0013] a step of determining markers based on methylation detection from the candidate regions, wherein the markers that can be used to distinguish the first population from the second population are determined from the candidate regions based on the difference between the third target data in the first population and the second population;

[0014] The first population or the second population is two groups of people with different characteristics that do not overlap.

[0015] 2. The method according to claim 1, wherein

[0016] The sequencing data is methylation sequencing data;

[0017] Preferably, the sequencing data is whole-genome methylation sequencing data.

[0018] 3. The method according to items 1 to 2, wherein

[0019] Methylation haplotype refers to the situation in which all n CpG sites in any candidate region are in methylated or unmethylated state; the number of occurrences of each methylation haplotype in any candidate region is 2 n indivual.

[0020] 4. The method according to items 1 to 3, wherein

[0021] The candidate region is a region where the distance between any two adjacent CpG sites does not exceed 1 bp;

[0022] Preferably, L is a positive integer less than or equal to 100.

[0023] 5. The method according to items 1 to 4, wherein

[0024] The step of determining markers based on methylation detection from candidate regions utilizes a random forest algorithm to determine methylation haploids with significant differences between the first population and the second population, and uses the candidate intervals of methylation haploids with the significant differences as markers based on methylation detection.

[0025] 6. The method according to items 1 to 5, wherein

[0026] The first group is the population suffering from the target disease, and the second group is the population not suffering from the target disease.

[0027] 7. The method according to claim 6, wherein

[0028] wherein the disease is selected from one or more of cancer, liver disease, nonalcoholic steatohepatitis, metabolic disease, autoimmune disease, cardiovascular disease, hypertension, systemic hypertension, pulmonary hypertension, vascular disease, atherosclerosis, diabetic vascular disease, vascular calcification, arrhythmia, heart failure, hypertrophy, fibrosis, cardiomyocyte proliferation and apoptosis, inflammation, diabetic cardiomyopathy, angiogenesis, neurological disorder, viral disease, drug addiction, conditions caused by drug addiction, increased adipose tissue, insulin resistance, and osteoporosis;

[0029] Preferably,

[0030] The cancer is selected from any one of liver cancer, lung cancer, pancreatic cancer, breast cancer, cervical cancer, endometrial cancer, colorectal cancer, nasopharyngeal cancer, ovarian cancer, prostate cancer, skin cancer, bladder cancer, stomach cancer, prostate cancer, colon cancer, bone cancer, brain cancer, rectal cancer, esophageal cancer, tongue cancer, kidney cancer, uterine body cancer, testicular cancer, urinary cancer, melanoma, astrocytic carcinoma, meningioma, Hodgkin's lymphoma, non-Hodgkin's lymphoma, acute lymphocytic leukemia, chronic lymphocytic leukemia, acute myeloid leukemia, chronic myeloid leukemia, adult T-cell leukemia lymphoma, hepatocellular carcinoma, bronchogenic carcinoma, multiple myeloma, basal cell tumor, seminoma, chondrosarcoma, myosarcoma, and fibrosarcoma.

[0031] 8. The method according to claim 7, wherein

[0032] The first group is people with liver cancer, and the second group is people without liver cancer.

[0033] 9. The method according to items 1 to 8, wherein

[0034] The marker is one or more of chr1, chr2 and chr3;

[0035] in,

[0036] The nucleotide sequence of chr1 is shown in SEQ ID NO: 1, the nucleotide sequence of chr2 is shown in SEQ ID NO: 2, and the nucleotide sequence of chr3 is shown in SEQ ID NO: 3.

[0037] 10. A method for constructing a model for distinguishing a first population from a second population using the markers obtained by any one of the methods of items 1 to 9, wherein:

[0038] Calculate the Mepm1 value of the marker for the first population;

[0039] Calculating the Mepm1 value of the marker for the second population;

[0040] Based on the Mepm1 values ​​of the markers in the first population and the Mepm1 values ​​of the markers in the second population, a risk discrimination model was established using the generalized linear regression method;

[0041] in,

[0042] The calculation formula of Mepm1 value is: Formula (1)

[0043] Among them, e refers to: the number of reads of a certain haplotype in the candidate region;

[0044] Interval length refers to: the size of the candidate region;

[0045] The total number of aligned reads refers to the number of reads that match the human reference gene from the sequencing data.

[0046] 11. A marker for detecting liver cancer, wherein the marker is one or more of chr1, chr2 and chr3;

[0047] in,

[0048] The nucleotide sequence of chr1 is shown in SEQ ID NO: 1, the nucleotide sequence of chr2 is shown in SEQ ID NO: 2, and the nucleotide sequence of chr3 is shown in SEQ ID NO: 3.

[0049] 12. A probe composition, wherein the probe composition comprises a probe targeting the methylation of the marker described in item 11.

[0050] 13. A model for screening markers based on methylation detection, comprising:

[0051] Obtaining a model for first target data, where the first target data is sequencing data for a first population;

[0052] Obtaining a model for second target data, where the second target data is sequencing data for a second population;

[0053] Determine a model of the candidate region, and determine M CpG-enriched candidate regions based on the first target data and the second target data;

[0054] A model for obtaining third target data, wherein the third target data includes:

[0055] The methylation haplotype of each candidate region in the M candidate regions in the first target data and the second target data; and the number of occurrences of each haplotype;

[0056] Determining a model of markers based on methylation detection from the candidate regions, and determining markers from the candidate regions that can be used to distinguish the first population from the second population based on the difference between the third target data in the first population and the second population;

[0057] The first population or the second population is two groups of people with different characteristics that do not overlap.

[0058] Compared with the prior art, the present invention has the following advantages:

[0059] This application uses high-throughput sequencing to capture the methylation haplotype signals of DNA rich in high CpG regions and develop disease prediction models, which has broad application prospects.

[0060] This application begins with high-throughput methylation sequencing, enriching high-CpG regions and extracting the characteristics of methylation haplotypes. Through model construction, disease risk prediction can be achieved. The data collection method proposed in this application takes into account the impact of high-CpG regions and methylation haplotypes on disease development and progression, and is expected to improve the accuracy of existing screening methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] FIG1 shows a schematic diagram of a high CpG enrichment region. DETAILED DESCRIPTION

[0062] The present application is further described below with reference to examples. It should be understood that the examples are only used to further illustrate and explain the present application and are not intended to limit the present application.

[0063] Unless otherwise defined, technical and scientific terms used in this specification have the same meaning as those commonly understood by those skilled in the art. Although methods and materials similar or identical to those described herein may be used in experiments or practical applications, the materials and methods are described herein below. In the event of a conflict, the present specification, including definitions, will prevail. In addition, the materials, methods, and examples are provided for illustrative purposes only and are not intended to be limiting. The present application is further described below with reference to specific examples, which are not intended to limit the scope of this application.

[0064] In this application, methylation is an important modification of proteins and nucleic acids that regulates gene expression and shutdown. It is closely related to many diseases such as cancer, aging, and Alzheimer's disease, and is one of the important research topics in epigenetics. The most common methylation modifications are DNA methylation and histone methylation.

[0065] In this application, DNA methylation refers to the methylation of the carbon atom at the 5th position of cytosine in CpG dinucleotides. As a relatively stable modification state, it can be inherited to new daughter DNA during DNA replication under the action of DNA methyltransferases. It is an important epigenetic mechanism. During DNA methylation, methylation of gene promoter regions can lead to transcriptional silencing of tumor suppressor genes, and is therefore closely related to the occurrence of tumors. Abnormal methylation includes hypermethylation of tumor suppressor genes and DNA repair genes, hypomethylation of repetitive DNA sequences, and loss of imprinting of certain genes, which are associated with the occurrence of various tumors.

[0066] Methylation Enhanced Gene Expression (MEGE) refers to the increase in gene expression due to methylation of certain sites within the gene. In principle, overexpression of the gene can lead to the occurrence of corresponding cancers. In this application, MEGE refers to a screening method for methylation markers within genes, which aims to screen sequence fragments in which methylation occurs at corresponding sites within genes in cancer tissues. By detecting the methylation status of the fragments, the occurrence of cancer can be prompted.

[0067] Normal tissue refers to adjacent tissue. For example, when screening for lung cancer markers, normal tissue refers to normal tissue in the lungs. Adjacent tissue refers to tissue surrounding cancerous tissue that has been pathologically identified and is different from cancerous tissue in terms of morphology, atypia, and histological typing. In some cases, because the data information of cancer tissue and adjacent tissue in the database is not one-to-one corresponding, normal tissue can also be used to replace the corresponding adjacent tissue. For example, when screening for lung cancer markers, adjacent tissue refers to tissue surrounding lung cancer tissue that does not have the characteristics of lung cancer tissue or normal lung tissue.

[0068] Specificity refers to the proportion of negative test results in samples from patients without a specific clinical disease.

[0069] Sensitivity refers to the proportion of samples from patients with established clinical disease that test positive.

[0070] The present application provides a method for screening markers based on methylation detection, which comprises:

[0071] a step of acquiring first target data, wherein the first target data is sequencing data for a first population; and a step of acquiring second target data, wherein the second target data is sequencing data for a second population;

[0072] The step of determining candidate regions is to determine M CpG-enriched candidate regions based on the first target data and the second target data; the step of obtaining third target data is to obtain the third target data, wherein the third target data includes: the methylation haplotype of each of the M candidate regions in the first target data and the second target data, and the number of occurrences of each haplotype; the step of determining markers based on methylation detection from the candidate regions is to determine markers that can be used to distinguish between the first and second populations based on the difference between the third target data in the first and second populations; the first population or the second population is two populations with different characteristics and the characteristics do not overlap.

[0073] In some embodiments of the present application, candidate CpG-enriched regions are regions enriched in high CpGs. The entire genome is scanned, requiring that a 5 bp interval contain at least two CpGs, and if the two segments differ by no more than 5 bp, a high CpG-enriched interval set is formed.

[0074] DNA methylation refers to the covalent attachment of a methyl group to the 5-carbon position of cytosine in genomic CpG dinucleotides by DNA methyltransferases (DNMTs). Most gene promoters are rich in high-CpG regions. Changes in methylation levels at these sites are correlated with the expression of downstream genes. Therefore, capturing changes in methylation haplotypes in high-CpG regions provides a theoretical basis for rapidly identifying disease-causing changes.

[0075] In some embodiments of the present application, the first population is a population suffering from the target disease, and the second population is a population not suffering from the target disease.

[0076] In some embodiments of the present application, the first population is a case group, that is, a population suffering from the target disease, and the second population is a control group, that is, a population not suffering from the target disease.

[0077] In some embodiments of the present application, the disease is selected from one or more of cancer, liver disease, non-alcoholic fatty liver disease, metabolic disease, autoimmune disease, cardiovascular disease, hypertension, systemic hypertension, pulmonary hypertension, vascular disease, atherosclerosis, diabetic vascular disease, vascular calcification, arrhythmia, heart failure, hypertrophy, fibrosis, myocardial cell proliferation and apoptosis, inflammation, diabetic cardiomyopathy, angiogenesis, neurological disorders, viral diseases, drug addiction, conditions caused by drug addiction, increased adipose tissue, insulin resistance, and osteoporosis; preferably, the cancer is selected from liver cancer, lung cancer, Any of pancreatic cancer, breast cancer, cervical cancer, endometrial cancer, colorectal cancer, nasopharyngeal cancer, ovarian cancer, prostate cancer, skin cancer, bladder cancer, stomach cancer, prostate cancer, colon cancer, bone cancer, brain cancer, rectal cancer, esophageal cancer, tongue cancer, kidney cancer, uterine corpus cancer, testicular cancer, urinary tract cancer, melanoma, astrocytic carcinoma, meningioma, Hodgkin lymphoma, non-Hodgkin lymphoma, acute lymphocytic leukemia, chronic lymphocytic leukemia, acute myeloid leukemia, chronic myeloid leukemia, adult T-cell leukemia / lymphoma, hepatocellular carcinoma, bronchogenic carcinoma, multiple myeloma, basal cell tumor, seminoma, chondrosarcoma, myeloid fibrosis, and fibrosarcoma.

[0078] In some embodiments of the present application, the sequencing data is methylation sequencing data; preferably, the sequencing data is whole-genome methylation sequencing data.

[0079] In some embodiments of the present application, a methylation haplotype refers to a situation in which all n CpG sites in any candidate region are in a methylated state or an unmethylated state; the number of occurrences of each methylation haplotype in any candidate region is 2 n For example, when there are 5 CpG sites in a candidate region, that is, n=5, the 5 CpG sites are either in a methylated state or in an unmethylated state, with a total of 2 5 possibilities.

[0080] In some embodiments of the present application, the total number of occurrences of different methylation haplotypes in the M candidate regions is counted as follows:

[0081] First, count the number of occurrences of different methylation haplotypes in each candidate region, which are 2 n1 , 2 n2 , 2 n3 , 2 n4 , 2 n5 ……2 nM , where n1, n2, n3, n4, n5...nM represent the number of CpG sites in each CpG-enriched region, and n1, n2, n3, n4, n5...nM are positive integers greater than 0;

[0082] The sum of the occurrences of different methylation haplotypes in the M candidate regions is 2 n1 +2 n2 +2 n3 +2 n4 +2 n5 +……+2 nM .

[0083] In some embodiments of the present application, the candidate region is a region where the distance between any two adjacent CpG sites does not exceed Lbp; preferably, L is a positive integer less than or equal to 100; for example, L can be 1bp, 2bp, 3bp, 4bp, 5bp, 6bp, 7bp, 8bp, 9bp, 10bp, 11bp, 12bp, 13bp, 14bp, 15bp, 16bp, 17bp, 18bp, 19bp, 20bp, 21bp, 22bp, 23bp, 24bp, 25bp, 26bp, 27bp, 28bp, 29bp, 30bp, 31bp, 32bp, 33bp, 34bp, 35bp, 36bp, 37bp, 38bp, 39bp, 40bp, 41bp, 42bp, 43bp, 44bp, 45bp, 46bp, 47bp, 48bp, 49bp, 50bp, 51bp, 52bp, 53bp, 54bp, 55bp, 56bp, 57bp, 58bp, 59bp, 60bp, 61bp, 62bp, 63bp, 64bp, 65bp, 66bp, 67bp, 68bp, 69bp, 70bp, 71bp, 72bp, 73bp, 74bp, 75bp, 76bp, 77bp, 78bp, 79bp, 80bp 6bp, 47bp, 48bp, 49bp, 50bp, 51bp, 52bp, 53bp, 54bp, 55bp, 56bp, 57bp, 58bp, 59bp, 60bp, 61bp, 62bp, 63bp, 64bp, 65bp, 66bp, 67bp, 68bp, 69bp, 70bp, 71bp, 72bp, 73bp, 74bp, 75bp, 76bp, 77bp, 78bp, 79bp, 80bp, 81bp, 82bp, 83bp, 84bp, 85bp, 86bp, 87bp, 88bp, 89bp, 90bp, 91bp, 92bp, 93bp, 94bp, 95bp, 96bp, 97bp, 98bp, 99bp, 100bp or any range therebetween. Current sequencing methods primarily utilize next-generation sequencing, with PE150 (paired-end sequencing 150) being the most common. This yields sequencing reads of up to 150 base pairs. When calculating haplotypes, it's best to use reads that cover the entire targeted region. Therefore, a larger L value reduces the number of available reads spanning the targeted region. This significantly reduces the available data, leading to greater bias in the true results. Therefore, L should not be too large.

[0084] In some embodiments of the present application, the step of determining markers based on methylation detection from candidate regions utilizes a random forest algorithm to determine methylation haploids having significant differences between the first population and the second population, and the candidate intervals of methylation haploids having the significant differences are used as markers based on methylation detection.

[0085] In some embodiments of the present application, the first population is a population suffering from liver cancer, and the second population is a population not suffering from liver cancer.

[0086] In some embodiments of the present application, the marker is one or more of chr1, chr2 and chr3; wherein the nucleotide sequence of chr1 is shown in SEQ ID NO: 1, the nucleotide sequence of chr2 is shown in SEQ ID NO: 2, and the nucleotide sequence of chr3 is shown in SEQ ID NO: 3.

[0087] The present application provides a model for screening markers based on methylation detection, which includes:

[0088] Obtaining a model for first target data, where the first target data is sequencing data for a first population;

[0089] Obtaining a model for second target data, where the second target data is sequencing data for a second population;

[0090] Determine a model for the candidate region, and determine M CpG-enriched candidate regions based on the first target data and the second target data; obtain a model for third target data, wherein the third target data includes: a methylation haplotype of each of the M candidate regions in the first target data and the second target data; and the number of occurrences of each haplotype; determine a model for a marker based on methylation detection from the candidate region, and determine a marker that can be used to distinguish between the first and second populations from the candidate region based on the difference between the third target data in the first and second populations; the first population or the second population is two populations with different characteristics that do not overlap.

[0091] The present application provides a method for constructing a model for distinguishing a first population from a second population using the markers obtained by the above method, wherein:

[0092] Calculate the Mepm1 value of the marker for the first population;

[0093] Calculating the Mepm1 value of the marker for the second population;

[0094] Based on the Mepm1 values ​​of the markers in the first population and the Mepm1 values ​​of the markers in the second population, a risk discrimination model was established using the generalized linear regression method;

[0095] in,

[0096] The calculation formula of Mepm1 value is: Formula (1)

[0097] Among them, e refers to: the number of reads of a certain haplotype in the candidate region;

[0098] Interval length refers to: the size of the candidate region;

[0099] The total number of aligned reads refers to the number of reads that match the human reference gene from the sequencing data.

[0100] Mepm1 is an important part of the algorithm described in this application. The Mepm1 value in this application mainly emphasizes the calculation method of haplotype. The calculation formula of Mepm1 value is the calculation method of haplotype. This application screens markers by enriching CG regions and checking haplotype data.

[0101] Using the proportion of haplotype methylation as an indicator for screening markers can ignore the inconsistency of individual test data, accurately reflect the true situation of the sample, and has high universal adaptability.

[0102] This application uses the Mepm1 calculation method for regions with high CpG enrichment, so the resulting markers are more easily converted to PCR. In addition, the use of haplotypes allows for more precise design of probe primers based on haplotypes, thus improving accuracy.

[0103] The present application provides a marker for detecting liver cancer, wherein the marker is one or more of chr1, chr2 and chr3; wherein the nucleotide sequence of chr1 is shown in SEQ ID NO: 1, the nucleotide sequence of chr2 is shown in SEQ ID NO: 2, and the nucleotide sequence of chr3 is shown in SEQ ID NO: 3.

[0104] The present application provides a probe composition, wherein the probe composition comprises a probe targeting the methylation of the above-mentioned marker.

[0105] Example

[0106] The inventors of this application utilized epigenomic and bioinformatics technologies, using the extraction protocol described herein, to analyze high-throughput methylation data from cancers, such as liver cancer, and identified a methylation haplotype pattern associated with cancer (e.g., liver cancer). Using this target sequence of a methylated gene, they were able to sensitively and specifically detect the methylation status of that gene, enabling the detection of cell-free DNA in peripheral blood.

[0107] 1) Construct high-CpG-rich regions and count all methylation pattern combinations within each region:

[0108] The whole genome is cut to find CpG-rich regions. The distance between two CpG sites must not exceed Lbp. The candidate region region 1 is obtained. Due to the sequencing read length, it is recommended to set L to a number less than 100 to obtain a shorter CG-rich region. Assuming that there are 5 CpG sites in a 20bp region, each site has two possibilities: methylation and unmethylation. Then the total number of methylation combination patterns in this interval is 25 (32) species.

[0109] As shown in Figure 1, a high-CpG-rich region is constructed. For example, when L = 5 bp, the entire genome is scanned, requiring that there are at least two CpGs in a 5 bp interval, and if the difference between the two segments does not exceed 5 bp, a target interval file is formed.

[0110] 2) Data Collection

[0111] The target cfDNA was captured using the Six Cancer NGS Methylation Detection Kit from Boercheng (Beijing) Technology Co., Ltd., and a methylation library was constructed and high-throughput sequencing was performed. Each sample generated at least 40GB of data, achieving a sequencing depth of 14,000X (approximately 150X non-repetitive sequencing depth).

[0112] 3) Data preprocessing

[0113] Bioinformatics methods were used to perform the following steps on the sequencing data of each sample: data quality preprocessing and assessment (fastp software), genome alignment (Bismark software), and removal of duplicate data caused by samples / experimental techniques (deduplicate_bismark software).

[0114] 2) Feature set extraction:

[0115] For all reads falling within region 1, count the number of occurrences of each methylation haplotype pattern within that region. Count the number of reads (e) that match a particular haplotype within these regions and calculate the corresponding Mepm1 value (which is used to determine the difference between the samples). If there are 32 total haplotype combinations in the interval and only 10 occur, the Mepm1 value for the remaining haplotype combinations is 0.

[0116] Among them, e refers to: the number of reads of a certain haplotype in the candidate region;

[0117] Interval length refers to: the size of the candidate region;

[0118] The total number of aligned reads refers to the number of reads that match the human reference gene from the sequencing data.

[0119] 3) Candidate marker selection:

[0120] 3.1): Screen variables with important contributions based on the random forest algorithm.

[0121] 4) Classification model construction:

[0122] 4.1): A disease risk discrimination model was established using the generalized linear regression method: the Mepm1 value of the candidate region was used as input, and the prediction score was calculated by the model to determine whether the subject was sick.

[0123] 5) Model evaluation and validation:

[0124] The constructed model was used to calculate disease risk scores and discrimination results for the training and test data sets. The classification accuracy of the model was evaluated using the AUC, sensitivity, and specificity, as well as the confusion matrix.

[0125] For example, when L is 20 bp, the genome is cut at 20 bp intervals according to the method described above. A total of 1,694,900 intervals are obtained. Due to the use of double-end PE150 sequencing, appropriately short intervals are selected in terms of interval length to ensure that the sequenced reads can cover the target interval as much as possible. In the experimental data, intervals with a length of less than 100 are uniformly selected, resulting in a total of 1,646,829 intervals. In the statistical analysis of valid sequencing reads, reads that can completely cover the target interval are selected. If the interval cannot be completely covered by reads, the reads are discarded. Since each interval contains different haplotypes depending on the CG sites it contains, the number of reads covering the haplotype in each interval is counted, and the corresponding Mepm1 value is calculated. Finally, the haplotype intervals that differ between the two groups of samples are selected to build a model. There are many ways to find differential intervals. You can choose T-test to view the P value, or you can filter features based on the model, such as SVM (support vector machine) and random forest, which will not be described in detail here.

[0126] Example 1

[0127] 1.1.cfDNA extraction and purification:

[0128] 1.1.1. Preparation of plasma samples:

[0129] Centrifuge the blood sample at 2000 g for 10 minutes at 4°C and transfer the plasma to a fresh centrifuge tube. Centrifuge the plasma sample at 16000 g for 10 minutes at 4°C and proceed to the next step depending on the type of collection tube used. The collection tube used in this experiment was other.

[0130] 1.1.2. Lysis and binding:

[0131] 1.1.2.1. Pipette 100 μL, 200 μL, 300 μL, 400 μL, and 500 μL of QIAGEN Proteinase K into 50 mL centrifuge tubes.

[0132] 1.1.2.2. Add 1 mL, 2 mL, 3 mL, 4 mL, or 5 mL of plasma or serum to the above 50 mL centrifuge tube.

[0133] 1.1.2.3. Add 0.8 mL, 1.6 mL, 2.4 mL, 3.2 mL, and 4.0 mL of Buffer ACL (containing 1.0 μg carrier RNA), cover, and vortex for 30 seconds. Note: Mix thoroughly to ensure complete lysis. Proceed immediately to the next step.

[0134] 1.1.2.4. Incubate at 60°C for 30 min.

[0135] 1.1.2.5. Take out the centrifuge tube, place it on the test bench, and unscrew the tube cap.

[0136] 1.1.2.6. Add 1.8 mL, 3.6 mL, 5.4 mL, 7.2 mL, and 9.0 mL of Buffer ACB to a 50 mL centrifuge tube; cover and mix for 15-30 seconds.

[0137] 1.1.2.7. Incubate the lysis mixture on ice for 5 minutes.

[0138] 1.1.2.8. Insert the QIAamp Mini column into the vacuum pump connector adapter and insert the 20 mL tube extender into the column. Note: Ensure that the tube extender is firmly inserted into the QIAamp Mini column to prevent sample leakage.

[0139] 1.1.2.9. Carefully add the lysate-buffer ACB mixture from step 7 to the tube extender of the QIAamp Mini column and turn on the vacuum pump. When all the lysate has been completely extracted from the Mini column, turn off the vacuum pump and release the pressure to 0 mbar. Carefully remove and discard the tube extender.

[0140] 1.1.3. Washing:

[0141] 1.1.3.1. Add 600 μL of Buffer ACW1 to the Mini column. With the cap open, start the vacuum pump. After all the liquid has passed through the column membrane, turn off the vacuum pump and release the pressure to 0 mbar.

[0142] 1.1.3.2. Add 750 μL of Buffer ACW2 to the Mini column. With the cap open, start the vacuum pump. After all the liquid has passed through the column membrane, turn off the vacuum pump and release the pressure to 0 mbar.

[0143] 1.1.3.3. Add 750 μL of ethanol (96-100%) to the Mini column, open the lid, and start the vacuum pump. After all the liquid has passed through the column membrane, turn off the vacuum pump and release the pressure to 0 mbar.

[0144] 1.1.3.4. Cover the QIAamp Mini column cap, remove it from the adapter, and discard the VacConnector. Place the QIAamp Mini column in a clean 2 mL collection tube and centrifuge at high speed (20,000 g; 14,000 rpm) for 3 minutes.

[0145] 1.1.3.5. Place the QIAamp Mini column in a new 2 mL collection tube. Open the tube cap and incubate at room temperature for 5 minutes to allow the membrane to completely dry.

[0146] 1.1.4. Elution of cfDNA:

[0147] Place the QIAamp Mini column in a 1.5 mL elution tube and discard the 2 mL collection tube from step 14. Add 20-150 μL of Buffer AVE to the center of the Mini membrane; cap the tube and incubate at room temperature for 3 minutes.

[0148] 1.1.4.2. Centrifuge at full speed (20,000 g; 14,000 rpm) for 1 min to elute the nucleic acid.

[0149] For cfDNA samples, AgiLent2100 performs fragment detection and Qubit is directly used for subsequent experiments.

[0150] 1.2. Bisulfite conversion and purification:

[0151] 1.2.1..Prepare CT Conversion Reagent:

[0152] 1.2.1.1. Add 700 μL NF water, 300 μL M-Dilution Buffer, and 50 μL M-Dissolving Buffer to a tube of CT conversion reagent. Mix thoroughly by vortexing or shaking frequently for 10 minutes at room temperature.

[0153] 1.2.1.2. After mixing, aliquot the solution into 10 reaction volumes.

[0154] 1.2.2. Perform bisulfite conversion on the DNA library and prepare the reaction system according to the table below.

[0155] Table 1

[0156] 1.2.3. Adjust the pipette volume to 100 μL, gently pipette up and down 6 times to mix, then divide the tubes into two and place them on the PCR instrument.

[0157] 1.2.4. Set the following program to perform the reaction on a PCR instrument: heated lid temperature 105°C.

[0158] Table 2

[0159] 1.2.5. Take a new 1.5mL centrifuge tube and add 600μL of M-Binding Buffer.

[0160] 1.2.6. After PCR, briefly centrifuge and transfer the two tubes of the same sample into the corresponding 1.5 mL centrifuge tubes mentioned above and mix thoroughly.

[0161] 1.2.7. Add the mixed sample to the Zymo-Spin TM Invert the IC Column to mix thoroughly, and centrifuge at 10,000 x g for 30 seconds.

[0162] 1.2.8. Add 100 μL of M-Wash Buffer to the column and centrifuge at 10,000 x g for 30 seconds.

[0163] 1.2.9. Add 200 μL of M-Desulphonation Buffer to the column, incubate at room temperature for 15-20 min, and centrifuge at 10,000 x g for 30 s.

[0164] 1.2.10. Add 200 μL of M-Wash Buffer to the column and centrifuge at 10,000 x g for 30 seconds.

[0165] 1.2.11. Repeat the previous step once.

[0166] 1.2.12. Place the column in a new collection tube and centrifuge again at 10,000 x g for 30 seconds.

[0167] 1.2.13. Place the recovery column in a new 1.5 mL EP tube, add 15 μL of low EDTA buffer to the center of the column membrane, and centrifuge at 10,000 x g for 30 seconds.

[0168] 1.3. Transformation:

[0169] Preheat the PCR instrument to 95°C.

[0170] 1.3.2. Set the following program to perform the reaction on a PCR instrument: heated lid temperature 105°C.

[0171] Table 3

[0172] 1.3.3. After incubation, immediately place the tube on ice for 2 minutes.

[0173] 1.4. Adapter ligation and purification:

[0174] 1.4.1. Configure the reaction system according to the following table:

[0175] Table 4

[0176] 1.4.2. Set the following program to perform the reaction on a PCR instrument: heated lid temperature 105°C.

[0177] Table 5

[0178] 1.5. Sample extension and purification:

[0179] 1.5.1. Configure the reaction system according to the following table:

[0180] Table 6

[0181] 1.5.2. Set the following program to perform the reaction on a PCR instrument: heated lid temperature 105°C.

[0182] Table 7

[0183] 1.5.3. When DNA protection buffer is added, the liquid turns blue. Gently pipette to mix thoroughly, then divide the tubes into two and place them in the PCR instrument.

[0184] 1.5.4. Set the following program and run: Heated lid at 105°C.

[0185] Table 8

[0186] 1.5.5. Prepare the purification system according to the following table:

[0187] Table 9

[0188] 1.5.6. Add the above proportion of magnetic beads to each sample for recovery, shake and mix thoroughly, then centrifuge immediately.

[0189] 1.5.7. Incubate at room temperature for 5 minutes.

[0190] 1.5.8. Vortex and mix thoroughly. Centrifuge briefly and place on a magnetic rack for adsorption until the solution becomes clear (~2 min). After the solution is clear, aspirate the supernatant.

[0191] 1.5.9. Add 200 μL of 80% ethanol to wash the beads for 30 seconds. Discard the supernatant and carefully remove any remaining ethanol from the inner wall of the dropper.

[0192] 1.5.10. Repeat the above steps.

[0193] 1.5.11. Add the optimal volume of Low EDTATE buffer recommended in the table above for elution and vortex to mix.

[0194] Incubate at room temperature for 2 minutes.

[0195] 1.5.13. Place on a magnetic stand for adsorption until the solution becomes clear (~2 min). Once the solution is clear, aspirate the supernatant.

[0196] 1.5.14. Transfer the entire eluate to a new 0.2 mL PCR tube, ensuring that the eluate does not contain the magnetic beads.

[0197] 1.6. Adapter ligation and purification:

[0198] 1.6.1. Prepare the library reaction system according to the following table:

[0199] Table 10

[0200] 1.6.2. Set up the following program and run: Heated lid 0°C:

[0201] Table 11

[0202] Prepare the purification system according to the following table:

[0203] Table 12

[0204] 1.6.4. Add the above proportion of magnetic beads to each sample for recovery, shake and mix thoroughly, then centrifuge immediately.

[0205] 1.6.5. Incubate at room temperature for 5 minutes.

[0206] 1.6.6. Vortex and mix thoroughly. Centrifuge briefly and place on a magnetic rack for adsorption until the solution becomes clear (~2 min). Once the solution is clear, aspirate the supernatant.

[0207] 1.6.7. Add 200 μL of 80% ethanol to wash the beads for 30 seconds. Discard the supernatant and carefully remove any remaining ethanol from the inner wall of the dropper.

[0208] 1.6.8. Repeat the above steps.

[0209] 1.6.9. Add the optimal volume of Low EDTATE buffer recommended in the table above for elution and vortex to mix.

[0210] 1.6.10. Incubate at room temperature for 2 min.

[0211] 1.6.11. Place on a magnetic rack for adsorption until the solution becomes clear (~2 min). Once the solution is clear, aspirate the supernatant.

[0212] 1.6.12. Transfer the entire eluate to a new 0.2 mL PCR tube, ensuring that the eluate does not contain the magnetic beads.

[0213] 1.7. Library amplification and purification:

[0214] 1.7.1. Prepare the library reaction system according to the following table:

[0215] Table 13

[0216] 1.7.2. Set up the following program and run: Heated lid at 105°C:

[0217] Table 14

[0218] 1.7.3. The recommended number of cycles is as follows:

[0219] Table 15

[0220] 1.7.4. Prepare the purification system according to the following table:

[0221] Table 16

[0222] 1.7.5. Transfer the PCR product into a 1.5 mL centrifuge tube.

[0223] 1.7.6. Add the above proportion of magnetic beads to each sample for recovery, shake and mix thoroughly, then centrifuge immediately.

[0224] 1.7.7. Incubate at room temperature for 5 minutes.

[0225] 1.7.8. Vortex and mix thoroughly. Centrifuge briefly and place on a magnetic rack for adsorption until the solution becomes clear (~2 min). Once the solution is clear, aspirate the supernatant.

[0226] 1.7.9. Add 500 μL of 80% ethanol to wash the beads for 30 seconds. Discard the supernatant and carefully remove any remaining ethanol from the inner wall of the dropper.

[0227] 1.7.10. Repeat the above steps.

[0228] 1.7.11. Place on the magnetic stand for 5-10 minutes until the beads are dry (avoid over-drying, as over-drying may result in reduced DNA recovery).

[0229] 1.7.12. Add the optimal volume of Low EDTATE buffer recommended in the table above for elution and vortex to mix.

[0230] Incubate at room temperature for 2 minutes.

[0231] 1.7.14. Place on a magnetic rack for adsorption until the solution becomes clear (~2 min). Once the solution is clear, aspirate the supernatant.

[0232] 1.7.15. Transfer the entire eluate to a new 0.2 mL PCR tube, ensuring that the eluate does not contain the magnetic beads.

[0233] 1.7.16. Aspirate 1 μL of sample for qubit calibration and perform 2100 quality control.

[0234] 1.8. Hybridization of sample and probe:

[0235] 1.8.1. Mixed samples:

[0236] 1.8.1.1. DNA library dosage refers to the table below. The total dosage can exceed 1500ng, but not exceed 4ug;

[0237] Table 17

[0238] 1.8.1.2. Calculate the amount of different samples and mix them evenly in a centrifuge tube.

[0239] 1.8.1.3. Add the following prehybridization reagents to the mixed sample and mix thoroughly, avoiding any bubbles.

[0240] Table 18

[0241] 1.8.1.4. Dry the mixed pre-hybridization reagents in a vacuum concentrator at room temperature (if heating is required, please use low temperature).

[0242] 1.8.2. Hybridization:

[0243] Incubate the Fast Hybridization Mix at 65°C for 10 min or until all precipitates have dissolved. Quickly vortex and add 20 μL of the Fast Hybridization Mix to the lyophilized sample from the previous step to resuspend the sample (do not allow the hybridization solution to return to room temperature). Mix gently with your fingertips to avoid creating bubbles.

[0244] 1.8.2.2. Centrifuge quickly to remove air bubbles and add 30 μL of Hybridization Enhancer to the surface of the above reagents.

[0245] 1.8.2.3. Place the PCR tube in a preheated PCR instrument for hybridization.

[0246] 1.8.2.4. Set the following program and run: Heated lid 85°C.

[0247] Table 19

[0248] 1.8.3. Combined:

[0249] 1.8.3.1. Vortex the pre-equilibrated streptavidin magnetic beads until thoroughly mixed. Add 100 μL of magnetic beads to a 1.5 mL centrifuge tube.

[0250] 1.8.3.2. Add 200 μL of binding buffer and mix thoroughly by pipetting.

[0251] 1.8.3.3. Place the centrifuge tube on a magnetic rack for 1 minute or until the solution is clear, discard the supernatant, and remove the centrifuge tube.

[0252] 1.8.3.4. Repeat the above washing steps 2 times, for a total of 3 times.

[0253] 1.8.3.5. After the final wash, add 200 μL of binding buffer and resuspend by vortexing to mix thoroughly.

[0254] 1.8.3.6. After hybridization is completed, open the lid of the PCR instrument and quickly transfer all the hybridization solution to the equilibrated magnetic beads.

[0255] 1.8.3.7. Mix the magnetic beads with hybridization solution thoroughly on a shaker, rocker, or rotator at room temperature for 30 minutes.

[0256] 1.8.3.8. Remove the centrifuge tube from the mixer, centrifuge rapidly, and place on a magnetic rack for 1 minute. Remove the supernatant and remove the tube.

[0257] 1.8.3.9. Add 200uL of preheated Wash Buffer 1 and mix thoroughly.

[0258] 1.8.3.10. Incubate at 63°C or 65°C for 5 min.

[0259] 1.8.3.11. Place the centrifuge tube on the magnetic rack for 1 minute, remove the supernatant, and remove the tube.

[0260] Repeat the above steps, add 200 μL of preheated Wash Solution 1 again, and mix thoroughly.

[0261] Incubate at 63°C or 65°C for 5 min.

[0262] 1.8.3.14. Transfer the liquid to a new tube; place it on a magnetic rack for 1 minute, discard the supernatant, and remove the tube.

[0263] 1.8.3.15. Add 200 μL of preheated wash buffer 2 and mix thoroughly using a pipette tip.

[0264] 1.8.3.16. Incubate at 48°C for 5 min.

[0265] 1.8.3.17. Place on the magnetic rack for 1 minute, remove the supernatant and take out the tube.

[0266] 1.8.3.18. Repeat (steps 3.15-3.17) and wash twice, for a total of three times.

[0267] 1.8.3.19. For the final time, use a 10uL pipette tip to aspirate the wash solution.

[0268] 1.8.3.20. Add 45 μL of water, mix well, and incubate the solution on ice.

[0269] 1.8.4. Post-capture PCR amplification, purification, and quality control:

[0270] 1.8.4.1. Set the following program and run: Heated lid 105°C.

[0271] Table 20

[0272] 1.8.4.2. Mix the magnetic bead mixture from 1.3 and pipette 22.5 μL into a 0.2 mL PCR tube.

[0273] 1.8.4.3. Add 2.5 μL of amplification primers and 25 μL of KAPA HiFi HotStart ReadyMix to a 0.2 mL PCR tube, for a total of 50 μL reaction volume.

[0274] 1.8.4.4. Mix gently with a pipette tip, centrifuge quickly, place in a PCR instrument, and start amplification.

[0275] 1.8.4.5. Vortex to thoroughly mix the pre-equilibrated DNA purification magnetic beads.

[0276] 1.8.4.6. Add 90 μL (1.8*) of DNA purification magnetic beads to the amplified PCR product and vortex to mix thoroughly.

[0277] 1.8.4.7. Incubate at room temperature for 5 minutes.

[0278] 1.8.4.8. Place the centrifuge tube on a magnetic rack for 1 minute and discard the supernatant after the solution becomes clear.

[0279] 1.8.4.9. Without removing the centrifuge tube from the magnetic rack, add 200 μL of freshly prepared 80% ethanol, incubate for 1 minute, and discard the supernatant. Repeat the 80% ethanol wash once (2 times in total), keeping the centrifuge tube on the magnetic rack.

[0280] 1.8.4.10. Carefully remove any remaining ethanol using a 10uL pipette tip and allow to stand at room temperature for 5-10 minutes or until the beads are dry. Be careful not to overdry the beads.

[0281] Remove the tube from the magnetic rack and add 32 μL of water. Mix thoroughly by pipetting with a pipette. Incubate at room temperature for 2 minutes.

[0282] 1.8.4.12. Place the centrifuge tube on a magnetic rack for 3 minutes or until the solution is clear.

[0283] 1.8.4.13. Transfer 30uL of supernatant to a clean 0.2mL centrifuge tube.

[0284] 1.8.4.14. Take 1 μL of library and quantify using Qubit and record the library concentration.

[0285] 1.8.4.15. Take 1 μL of sample and use AgiLent2100 to measure the library fragment length.

[0286] 1.8.4.16. Sequencing was performed using the Illumina high-throughput sequencing platform.

[0287] 1.9. The methylation bioinformatics analysis process is as follows: Use fastp quality control software to check sequencing quality, remove low-quality reads, and then use Bismark alignment software to align the clean, quality-controlled data to the reference genome. According to the analysis plan, extract reads within region 1, i.e., e, and count the occurrences of all methylation haplotypes within each region to form a feature set.

[0288] Example 2: Constructing a cancer risk assessment classification model

[0289] Borcheng collected 10-20 mL of peripheral blood from 55 liver cancer patients and 60 healthy subjects, separated the plasma samples, and extracted and purified cfDNA. The detection method in Example 1 was used to obtain the number of occurrences of all methylation haplotypes in each region. 40 liver cancer data and 40 healthy subjects were selected as training sets, and three features were selected as candidate markers using the random forest method. The detailed sequences of these three features are as follows:

[0290] >chr1:226926646-226926713_ITPKB

[0291] >chr2:131129718-131129730_PTPN18

[0292] >chr11:77122736-77122763_PAK1

[0293] After sulfite conversion, it was found that the hypermethylated haplotypes in these three intervals (i.e., all CG sites are in a methylated state) were clearly distinguished between liver cancer and healthy people. We used the remaining 15 liver cancer data and 20 healthy people as the test set. The model was constructed using generalized linear regression. If the values ​​corresponding to the hypermethylated haplotypes of the above three sequences are mepm_ITPKB, mepm_PTPN18, and mepm_PAK1, respectively. The model constructed using generalized linear regression is:

[0294] logp / (1-p) = 0.0785*mepm_ITPKB + 0.0664*mepm_PTPN18 + 0.0629*mepm_PAK1 - 3.187. The model classified a subject as having cancer when the predicted score exceeded 0.5. The model achieved 100% sensitivity, 100% specificity, and an AUC of 1 in the training set. In the test set, the model achieved 86.67% sensitivity, 1% specificity, and an AUC of 0.98.

[0295] It can be seen that the markers obtained by the screening method of the present application have high accuracy and strong sensitivity when used to predict diseases.

[0296] The classification model in Example 2 calculated the cancer risk of 15 subjects and the discrimination results were as follows:

[0297] Table 1

[0298] Samples No. 1 through No. 15 are patients with liver cancer, while samples No. 16 through No. 40 do not have cancer. If the model's calculated score exceeds 0.5, the subject is considered to have liver cancer; otherwise, the subject is considered to not have cancer.

[0299] Table 2

[0300] Although the present invention has been disclosed above with reference to the embodiments, they are not intended to limit the present invention. Anyone with ordinary knowledge in the art may make slight changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the scope of the patent application attached hereto.

Claims

1. A method for screening markers based on methylation detection, wherein: include: A step of acquiring first target data, wherein the first target data is sequencing data for a first population; a step of acquiring second target data, wherein the second target data is sequencing data for a second population; a step of determining candidate regions, determining M CpG-enriched candidate regions based on the first target data and the second target data; The step of acquiring third target data includes: The methylation haplotype of each candidate region in the M candidate regions in the first target data and the second target data, and the number of occurrences of each haplotype; A step of determining a marker based on methylation detection from the candidate region, wherein the marker that can be used to distinguish the first population from the second population is determined from the candidate region based on the difference between the third target data in the first population and the second population; The first population or the second population is two groups of people with different characteristics and the characteristics do not overlap.

2. The method according to claim 1, wherein: The sequencing data is methylation sequencing data; Preferably, the sequencing data is whole genome methylation sequencing data.

3. The method according to claims 1 to 2, wherein: A methylation haplotype refers to a situation in which all n CpG sites in any candidate region are in a methylated state or an unmethylated state; the number of occurrences of each methylation haplotype in any candidate region is 2 n indivual.

4. The method according to claims 1 to 3, wherein: The candidate region is a region where the distance between any two adjacent CpG sites does not exceed Lbp; Preferably, L is a positive integer less than or equal to 100.

5. The method according to claims 1 to 4, wherein: The step of determining markers based on methylation detection from candidate regions utilizes a random forest algorithm to determine methylation haploids with significant differences between the first population and the second population, and uses the candidate intervals of methylation haploids with the significant differences as markers based on methylation detection.

6. The method according to claims 1 to 5, wherein: The first group is the people suffering from the target disease, and the second group is the people who do not suffer from the target disease.

7. The method according to claim 6, wherein: wherein the disease is selected from one or more of cancer, liver disease, nonalcoholic steatohepatitis, metabolic disease, autoimmune disease, cardiovascular disease, hypertension, systemic hypertension, pulmonary hypertension, vascular disease, atherosclerosis, diabetic vascular disease, vascular calcification, arrhythmia, heart failure, hypertrophy, fibrosis, cardiomyocyte proliferation and apoptosis, inflammation, diabetic cardiomyopathy, angiogenesis, neurological disorders, viral diseases, drug addiction, conditions caused by drug addiction, increased adipose tissue, insulin resistance, and osteoporosis; Preferably, The cancer is selected from any one of liver cancer, lung cancer, pancreatic cancer, breast cancer, cervical cancer, endometrial cancer, colorectal cancer, nasopharyngeal cancer, ovarian cancer, prostate cancer, skin cancer, bladder cancer, stomach cancer, prostate cancer, colon cancer, bone cancer, brain cancer, rectal cancer, esophageal cancer, tongue cancer, kidney cancer, uterine body cancer, testicular cancer, urinary cancer, melanoma, astrocytic carcinoma, meningioma, Hodgkin's lymphoma, non-Hodgkin's lymphoma, acute lymphocytic leukemia, chronic lymphocytic leukemia, acute myeloid leukemia, chronic myeloid leukemia, adult T-cell leukemia lymphoma, hepatocellular carcinoma, bronchial cancer, multiple myeloma, basal cell tumor, seminoma, chondrosarcoma, myosarcoma, and fibrosarcoma.

8. The method according to claim 7, wherein: The first group is people with liver cancer, and the second group is people without liver cancer.

9. The method according to claims 1 to 8, wherein: The marker is one or more of chr1, chr2 and chr3; in, The nucleotide sequence of chr1 is shown in SEQ ID NO:1, the nucleotide sequence of chr2 is shown in SEQ ID NO:2, and the nucleotide sequence of chr3 is shown in SEQ ID NO:

3.

10. A method for constructing a model for distinguishing a first population from a second population using the markers obtained by the method of any one of claims 1 to 9, wherein: Calculate the Mepm1 value of the marker for the first population; Calculating the Mepm1 value of the marker for the second population; Based on the Mepm1 values ​​of the markers of the first population and the Mepm1 values ​​of the markers of the second population, a risk discrimination model was established using the generalized linear regression method; in, The calculation formula of Mepm1 value is: Formula (1) Among them, e refers to: the number of reads of a certain haplotype in the candidate region; The interval length refers to: the size of the candidate region; The total number of aligned reads refers to the number of reads obtained from sequencing data that match the human reference gene.

11. A marker for detecting liver cancer, wherein: The marker is one or more of chr1, chr2 and chr3; in, The nucleotide sequence of chr1 is shown in SEQ ID NO:1, the nucleotide sequence of chr2 is shown in SEQ ID NO:2, and the nucleotide sequence of chr3 is shown in SEQ ID NO:

3.

12. A probe composition, wherein: The probe composition comprises a probe targeting the methylation of the marker according to claim 11.

13. A model for screening markers based on methylation detection, wherein: include: Obtaining a model of first target data, where the first target data is sequencing data for a first population; Obtaining a model for second target data, where the second target data is sequencing data for a second population; Determine a model of the candidate region, and determine M CpG-enriched candidate regions based on the first target data and the second target data; A model for obtaining third target data, wherein the third target data includes: The methylation haplotype of each candidate region in the M candidate regions in the first target data and the second target data; and the number of occurrences of each haplotype; Determine a model of markers based on methylation detection from the candidate regions, and determine markers from the candidate regions that can be used to distinguish the first population from the second population based on the difference between the third target data in the first population and the second population; The first population or the second population is two groups of people with different characteristics and the characteristics do not overlap.

Images