Marker screening method and system based on cfDNA methylation mode, electronic equipment, marker combination and application

Abstract

The marker screening method based on the cfDNA methylation mode comprises the following steps: acquiring corresponding methylation intervals of two crowd samples, and respectively determining methylation state information of the two methylation intervals; based on methylation state information, methylation mode recognition results of two methylation intervals are obtained respectively, methylation interval markers of susceptible diseases in different methylation states can be determined, and the detection rate of cfDNA of different diseases can be increased; based on the methylation pattern recognition result, two cfDNA quantitative results are obtained respectively to serve as potential biomarker features of sample classification; based on a cfDNA quantitative result, screening a difference characteristic interval in the two samples from the methylation interval, and determining the difference characteristic interval as a marker so as to identify disease information contained in various cfDNAs in the plasma and realize early screening and multi-screening of various diseases.

Description

Technical Field

[0001] The present application belongs to the technical field of disease classification, and specifically, relates to a marker screening method, system, electronic device and marker combination and application based on cfDNA methylation pattern. Background Art

[0002] There are many corresponding detection methods in clinical practice for various types of diseases, but the methods that can be used as the gold standard for detection are usually costly and invasive, which greatly affects the willingness of the early screening population to test. This is also an important reason for the low prevalence of early screening for metabolic diseases, tumors, and autoimmune diseases. In recent years, some successful applications of plasma free DNA (cfDNA) in the fields of non-invasive prenatal diagnosis and disease detection have shown its great prospects as a new generation of molecular markers. Among them, abnormal cfDNA methylation patterns are highly potential molecular markers for disease diagnosis. Changes in the methylation pattern of cfDNA from specific tissues can lead to abnormalities in the final cfDNA methylation pattern. Therefore, identifying abnormal methylation patterns from patients' cfDNA can effectively diagnose their disease status and locate their diseased tissues or organs.

[0003] However, the extremely low cfDNA content in the early plasma of some diseases, as well as the heterogeneity within the same disease and between multiple diseases, poses a huge challenge to cfDNA-based cancer early screening. Since there is only an extremely low relevant cfDNA content in the plasma in the early stages of many diseases, traditional detection methods such as high-depth sequencing or small panels can only capture a very small amount of cfDNA fragments, resulting in extremely high false negatives. In addition, the traditional methylation quantification method, namely the beta value, is defined as the proportion of the number of methylated reads at a CpG site. It is very insensitive to changes in methylation levels caused by trace amounts of cfDNA, and it is difficult to distinguish these cfDNA from random noise.

[0004] Therefore, how to effectively improve the ability to detect early disease onset signals in plasma through specific high-depth detection of target areas and the use of methylation quantification methods at the cfDNA fragment level is a key issue that needs to be addressed in the current cfDNA marker screening. In view of the methylation patterns contained in methylated CpG sites in cfDNA that can reveal their association with the early evolution of the disease, this patent proposes a method for screening methylation intervals with specific methylation pattern recognition results as markers. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the purpose of this application is to provide a marker screening method based on cfDNA methylation pattern. Through the screening of target detection areas, specific high-depth detection, and the use of methylation quantification methods at the cfDNA fragment level, methylation markers that effectively improve the ability to detect early disease signals in plasma are obtained.

[0006] Specifically, this application involves the following aspects:

[0007] According to one aspect of the present application, a marker screening method based on cfDNA methylation pattern is proposed, comprising:

[0008] A collection step, obtaining corresponding methylation intervals of the first population sample and the second population sample, and determining methylation status information of the methylation intervals of the first population sample and the second population sample respectively;

[0009] A pattern recognition quantitative step, based on the methylation status information, respectively obtaining methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample; and

[0010] Based on the methylation pattern recognition results, the cfDNA quantitative results of the first population samples and the second population samples are obtained respectively;

[0011] In the characteristic interval screening step, based on the cfDNA quantification results, characteristic intervals that are different between the first population sample and the second population sample are screened out from the methylation intervals and determined as markers.

[0012] According to one embodiment of the present application,

[0013] The corresponding methylation intervals of the first population sample and the second population sample are obtained respectively:

[0014] Merge CpG sites with relative distance ≤ d1 to obtain the merged interval;

[0015] The merged intervals were further screened to obtain methylated intervals.

[0016] According to one embodiment of the present application,

[0017] The combined intervals were further screened to obtain methylation intervals including:

[0018] Use one or more of the following methods to further filter the merged intervals:

[0019] Method 1: stipulate that the interval length of the merged interval is ≤ d2;

[0020] Method 2: Select the merged interval containing CpG sites ≥ i;

[0021] Method 3: Select the merged interval located within the methylation island (CGI); and

[0022] Method 4: Select the merged intervals within the range of known disease-related genes.

[0023] According to one embodiment of the present application,

[0024] Determining the methylation status information of the methylation intervals of the first population sample and the second population sample respectively includes:

[0025] The methylation status of all CpG sites in the corresponding methylation intervals of the first population sample and the second population sample are obtained respectively.

[0026] According to one embodiment of the present application,

[0027] Based on the methylation status information, respectively obtaining methylation pattern recognition results of the first population sample and the second population sample includes:

[0028] Based on the methylation status information of the first population sample and the second population sample, respectively calculate the methylation rate α of the methylation interval of the first population sample and the second population sample; and

[0029] Based on the methylation rate α, the methylation threshold α of the methylation interval of the first population sample and the second population sample is determined respectively. cut and methylation direction m d .

[0030] According to one embodiment of the present application,

[0031] Based on the methylation pattern recognition results, obtaining cfDNA quantitative results of the first population sample and the second population sample respectively includes:

[0032] Based on the methylation pattern recognition results, the relative quantitative abundance of cfDNA in the corresponding methylation intervals of the first population samples and the second population samples was calculated respectively.

[0033] According to one embodiment of the present application,

[0034] The characteristic intervals that are screened out from the methylation intervals and that are different between the first population sample and the second population sample include:

[0035] Methylation threshold α based on methylation interval cut , methylation direction m d and cfDNA relative quantitative abundance, to obtain the sensitivity value s of the methylation interval in the first population sample or the second population sample;

[0036] For different methylation directions m dThe methylation intervals are arranged in descending order according to the sensitivity value s; and

[0037] Select the first m different methylation directions m respectively d The methylation interval of

[0038] According to one embodiment of the present application,

[0039] The methylation pattern recognition results determine the differential methylation intervals of high methylation and / or low methylation in the first population sample and the second population sample to improve the detection rate of disease-related cfDNA.

[0040] According to another aspect of the present application, a marker screening system based on cfDNA methylation pattern is proposed, comprising:

[0041] A collection module, which obtains corresponding methylation intervals of the first population sample and the second population sample, and determines methylation status information of the methylation intervals of the first population sample and the second population sample respectively;

[0042] A pattern recognition quantitative module, based on the methylation status information, obtains methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample respectively; and

[0043] Based on the methylation pattern recognition results, the cfDNA quantitative results of the first population samples and the second population samples are obtained respectively;

[0044] The characteristic interval screening module, based on the cfDNA quantitative results, screens out characteristic intervals that are different between the first population samples and the second population samples from the methylation intervals, and determines them as markers.

[0045] According to one embodiment of the present application, a marker screening system based on cfDNA methylation patterns performs a marker screening method based on cfDNA methylation patterns.

[0046] According to another aspect of the present application, an electronic device is also provided, comprising:

[0047] Processor; and

[0048] A memory stores computer program instructions, which, when executed by a processor, cause the processor to execute a marker screening method based on a cfDNA methylation pattern.

[0049] The present application also proposes a marker for cancer detection, wherein the marker comprises one or more of the following:

[0050] The genomic region from 41284213 to 41284519 of chromosome 1, the genomic region from 156814769 to 156815069 of chromosome 1, the genomic region from 30453757 to 30454057 of chromosome 2, the genomic region from 13589071 to 13589195 of chromosome 3, the genomic region from 50243493 to 50243657 of chromosome 4 The genomic region from positions 151177304 to 151177622 of chromosome 7, the genomic region from positions 99691599 to 99691943 of chromosome 7, the genomic region from positions 101838935 to 101839237 of chromosome 7, the genomic region from positions 17767751 to 17767994 of chromosome 8, the genomic region from positions 112081154 to 112081515 of chromosome 9, The genomic region from 126772025 to 126772354 of chromosome 11, the genomic region from 3313213 to 3313341 of chromosome 10, the genomic region from 48413903 to 48414022 of chromosome 10, the genomic region from 69516983 to 69517300 of chromosome 11, the genomic region from 120435185 to 120435492 of chromosome 11, The genomic region from 90921341 to 90921645 of chromosome 14, the genomic region from 103011099 to 103011399 of chromosome 14, the genomic region from 27893261 to 27893592 of chromosome 17, the genomic region from 32610596 to 32610787 of chromosome 17, and the genomic region from 5059596 to 5059899 of chromosome 19.

[0051] The present application also proposes the use of markers in preparing a kit for cancer detection, wherein the markers include one or two or more of the following:

[0052] The genomic region from 41284213 to 41284519 of chromosome 1, the genomic region from 156814769 to 156815069 of chromosome 1, the genomic region from 30453757 to 30454057 of chromosome 2, the genomic region from 13589071 to 13589195 of chromosome 3, the genomic region from 50243493 to 50243657 of chromosome 4 The genomic region from positions 151177304 to 151177622 of chromosome 7, the genomic region from positions 99691599 to 99691943 of chromosome 7, the genomic region from positions 101838935 to 101839237 of chromosome 7, the genomic region from positions 17767751 to 17767994 of chromosome 8, the genomic region from positions 112081154 to 112081515 of chromosome 9, The genomic region from 126772025 to 126772354 of chromosome 11, the genomic region from 3313213 to 3313341 of chromosome 10, the genomic region from 48413903 to 48414022 of chromosome 10, the genomic region from 69516983 to 69517300 of chromosome 11, the genomic region from 120435185 to 120435492 of chromosome 11, The genomic region from 90921341 to 90921645 of chromosome 14, the genomic region from 103011099 to 103011399 of chromosome 14, the genomic region from 27893261 to 27893592 of chromosome 17, the genomic region from 32610596 to 32610787 of chromosome 17, and the genomic region from 5059596 to 5059899 of chromosome 19.

[0053] The present application also proposes a composition for cancer detection, wherein the composition comprises:

[0054] A nucleic acid for detecting the methylation status of the above markers,

[0055] The methylation status of the marker is characterized by the methylation of the target sequence of the marker.

[0056] The present application also proposes a kit, which comprises reagents for detecting the above-mentioned markers.

[0057] Beneficial Effects

[0058] 1. Provide cost-effective early screening. The cfDNA markers of various diseases obtained by the method disclosed in this application can accurately distinguish between diseased individuals and healthy individuals at an early stage, especially the detection of diseased individuals. This application is based on the method of screening multiple binary markers based on cfDNA methylation patterns, obtaining markers and the combination and application of markers, which can be conveniently applied to early non-invasive screening of diseases, and can also ensure that the markers have high sensitivity and specificity in the detection, thereby reducing the misdiagnosis rate.

[0059] 2. To solve the defects of existing cfDNA markers in early screening. The previous method of dividing methylation blocks only considered the distance between adjacent CpGs. However, in DNA sequences, even adjacent CpG sites will show different methylation directions in diseased samples and healthy samples. Therefore, merging them into the same block based only on the distance between CpGs may cause the delta values ​​of diseased samples and healthy samples between CpG sites with different methylation directions to cancel each other out, and some different methylation intervals are filtered out. The method of the present application not only considers the distance between CpG sites, but also considers the directionality between CpG sites based on the methylation status of the interval where the CpG sites are located, and then considers the difference in methylation levels between the sites, and calculates the characteristic values ​​of the difference intervals corresponding to diseased samples and healthy samples, so as to identify the disease information contained in various types of cfDNA in plasma.

[0060] 3. To improve the treatment effect and prognosis of early screening diseases, this application uses cfDNA markers with specific methylation patterns for detection, which can be used for early detection and diagnosis of various human lesions. Since many early diseases, such as tumors and autoimmune diseases, are easy to carry out early prevention and cure, early screening and multiple screening of diseases through cfDNA markers can effectively avoid the deterioration of the disease and improve the subsequent treatment effect and survival rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a schematic diagram of the model prediction value of the test set of Example 3 of the present application;

[0062] Figure 2 This is a schematic diagram of the model prediction value of the training set of Example 3 of the present application;

[0063] Figure 3 This is the ROC curve diagram of the test set of Example 3 of the present application. DETAILED DESCRIPTION

[0064] The present application is further described below in conjunction with examples. It should be understood that the examples are only used to further describe and illustrate the present application and are not used to limit the present application.

[0065] Unless otherwise defined, the technical and scientific terms in this specification have the same meaning as those generally understood by those skilled in the art. Although similar or identical methods and materials as described herein may be used in experiments or practical applications, materials and methods are described herein below. In the event of a conflict, the present specification including the definitions therein shall prevail. In addition, materials, methods and examples are provided for illustration only and are not restrictive. The present application is further described below in conjunction with specific embodiments, but is not intended to limit the scope of the present application.

[0066] definition

[0067] DNA methylation

[0068] DNA methylation refers to the methylation process of the 5th carbon atom on cytosine in CpG dinucleotides. As a relatively stable modification state, it can be inherited to new offspring DNA during DNA replication under the action of DNA methyltransferase. It is an important epigenetic mechanism. When DNA is methylated, methylation of the gene promoter region can lead to transcriptional silencing of tumor suppressor genes, so it is closely related to the occurrence of tumors. Abnormal methylation includes high methylation of tumor suppressor genes and DNA repair genes, low methylation of repetitive sequence DNA, and loss of imprinting of certain genes, which are related to the occurrence of various tumors.

[0069] Specificity

[0070] Specificity refers to the proportion of negative test results among samples from patients without a specific clinical disease.

[0071] Sensitivity

[0072] Sensitivity refers to the proportion of samples from patients with established clinical disease that have a positive test result.

[0073] Methylation status

[0074] DNA methylation refers to the process in which a methyl group is covalently bonded to the 5th carbon position of cytosine in the genomic CpG dinucleotide under the action of DNA methyltransferase. This modification can affect genetic expression without changing the DNA sequence. The methylation state refers to the methylation of the CpG site. When the CpG site is not methylated, it is not in the methylation state. The methylation state information refers to the information containing whether the CpG site is in the methylation state. The generation of the methylation state is closely related to the regulation of gene expression, genetic imprinting, and the occurrence of diseases.

[0075] Random Forest

[0076] Random forest is essentially a branch of machine learning - Ensemble Learning, which is a method of integrating many decision trees into a forest and using it to predict the final result. Random forest, as the name implies, is to build a forest in a random way. The forest consists of many decision trees, and there is no correlation between each decision tree in the random forest. After obtaining the forest, when a new input sample enters, each decision tree in the forest is allowed to make a judgment to see which category the sample should belong to (for classification algorithms), and then see which category is selected the most, and predict the sample as that category.

[0077] ROC Curve

[0078] The receiver operating characteristic curve (ROC curve) is also called the sensitivity curve. It is named so because each point on the curve reflects the same sensitivity. They are all responses to the same signal stimulus, but the results are obtained under several different judgment criteria. The receiver operating characteristic curve is a coordinate graph composed of the false alarm probability as the horizontal axis and the hit probability as the vertical axis, and a curve drawn by the different results obtained by the subjects under specific stimulus conditions due to different judgment criteria.

[0079] In this application, the ROC curve can reflect the classification effect of the classifier to a certain extent. AUC is actually the area under the ROC curve. AUC intuitively reflects the classification ability expressed by the ROC curve.

[0080] The marker screening method based on cfDNA methylation pattern disclosed in the present application includes:

[0081] The acquisition step obtains the corresponding methylation intervals of the first population sample and the second population sample, and determines the methylation state information of the methylation intervals of the first population sample and the second population sample respectively. Specifically, the distribution characteristics of CpG sites in the whole genome range, such as the adjacent relationship of CpG sites, can be used to define the methylation interval. The present application can select hg19 (GRCh37) as the reference genome, and based on the adjacent relationship of CpG sites in the reference genome sequence, merge adjacent CpG sites to obtain methylation intervals containing multiple CpG sites. It will be appreciated by those skilled in the art that the reference genome that can be selected is not limited to hg19, and the marker screening method disclosed in the present application can also be used to obtain methylation intervals containing multiple CpG sites on other reference genomes.

[0082] In the present application, the first population and the second population are two groups of people with different characteristics and non-overlapping characteristics; specifically, the first population can represent a patient population suffering from any disease that is positively correlated with high methylation levels of cfDNA, and the second population can represent a healthy population without disease. Among them, the first population samples are selected from samples of patients with rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis, organ transplant rejection, schizophrenia, depression, bipolar disorder, autism spectrum disorder, attention deficit hyperactivity disorder, liver cancer, lung cancer, pancreatic cancer, breast cancer, cervical cancer, endometrial cancer, nasopharyngeal cancer, ovarian cancer, bladder cancer, gastric cancer, prostate cancer, colon cancer, bone cancer, rectal cancer, esophageal cancer, testicular cancer, 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, multiple myeloma, chondrosarcoma, myosarcoma, fibrosarcoma, coronary heart disease, hypertension, arrhythmia, diabetes, hypothyroidism or hyperthyroidism; the second population samples are selected from samples of healthy people.

[0083] In the present application, respectively obtaining the corresponding methylation intervals of the first population sample and the second population sample may include the following methods:

[0084] Merge CpG sites with a relative distance ≤ d1 to obtain a merged interval; and further screen the merged interval to obtain a methylated interval. By adjusting the size of d1, the density of CpG sites in the merged interval can be changed; the value range of d1 is 10-100bp, for example, d1 can be 10bp, 20bp, 30bp, 40bp, 50bp, 60bp, 70bp, 80bp, 90bp or 100bp, preferably, d1=100bp can be set, at which time the relative distance of CpG sites in the merged interval obtained by merging adjacent CpG sites does not exceed 100bp; on this basis, if the value of d1 is reduced, the relative distance of CpG sites in the merged interval is further shortened, the length of each merged interval obtained on the same length of the genomic sequence fragment may be reduced, and the proportion of CpG sites is higher. If the value of d1 is increased, the relative distance of CpG sites in the merged interval increases, the length of each merged interval obtained on the same length of the genomic sequence fragment may increase, and the proportion of CpG sites is lower.

[0085] Among them, further screening the merged interval to obtain the methylation interval may include:

[0086] Use one or more of the following methods to further filter the merged intervals:

[0087] Method 1: stipulate that the interval length of the merged interval is ≤ d2;

[0088] Method 2: Select the merged interval containing CpG sites ≥ i;

[0089] Method 3: Select the merged interval located within the methylation island (CGI); and

[0090] Method 4: Select the merged intervals within the range of known disease-related genes.

[0091] It should be noted that in the present application, for the merged interval obtained by merging adjacent CpG sites at a specified distance, any one of methods 1 to 4 can be selected for further screening, or methods 1 to 4 can be arranged and combined in sequence to further screen the merged interval. The methylation intervals obtained by the above screening methods can be used to determine their methylation status information in the first population sample and the second population sample.

[0092] In particular, d2=300bp, i=3 can be set, and the merged intervals can be screened in the order of methods 1 to 4, so that the length of the merged interval does not exceed 300bp, and then the intervals containing at least 3 CpG sites, the intervals located in the CGI, and the intervals located in the range of known genes related to hepatocellular tumors are selected in turn; in this way, methylation intervals containing rich CpG sites for obtaining hepatocellular tumor markers can be obtained. The known disease-related genes can come from paper documents, patent documents, and public information of enterprises, but their scope is not limited to the known related gene range of hepatocellular tumors in the above examples, and can be determined according to the specific range of the markers screened in this application.

[0093] In the present application, determining the methylation status information of the methylation interval of the first population sample and the second population sample, respectively, may include: obtaining the methylation status of all CpG sites in the corresponding methylation interval of the first population sample and the second population sample, respectively. Specifically, cfDNA samples are extracted from the first population sample and the second population sample, and then the cfDNA samples are subjected to panel capture and methylation library sequencing, and multiple sequencing reads are obtained from the cfDNA samples, and the sequencing reads are entered into the methylation interval. The methylation state of a CpG site may refer to whether the CpG site in each sequencing read is methylated, and the proportion of methylated CpG sites in all CpG sites in each sequencing read. It will be appreciated by those skilled in the art that whether methylation occurs at a CpG site can be determined by methods such as methylation-specific PCR, bisulfite sequencing, or high-resolution melting curve method.

[0094] The marker screening method based on cfDNA methylation pattern disclosed in the present application also includes:

[0095] A pattern recognition quantitative step, based on the methylation status information, respectively obtaining methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample; and

[0096] Based on the methylation pattern recognition results, the cfDNA quantitative results of the first population samples and the second population samples were obtained respectively.

[0097] In the present application, the methylation pattern recognition result may refer to the result of comparing the methylation rate of a methylation interval with the methylation threshold in the high methylation direction / low methylation direction specified for the interval to determine whether the methylation rate of the interval is greater than / less than the specified methylation threshold. Specifically, the methylation pattern recognition result of the methylation interval may include: methylation rate α, methylation threshold α cut Or methylation direction m d One or more of. The methylation rate α may be the ratio of the methylated CpG sites on each sequencing read segment of the methylation interval of the first population sample and the second population sample to all CpG sites on the sequencing read segment. For example, a sequencing read segment in a methylation interval contains a total of n CpG sites, of which n1 CpG sites are detected to be methylated and n2 CpG sites are detected to be unmethylated. Therefore, α of the sequencing read segment = n1 / n1+n2.

[0098] Furthermore, the methylation threshold α cut It can be a specified methylation rate threshold α, which represents the minimum / maximum methylation rate of a methylation interval to be identified as a disease-related cfDNA DNA fragment in the high / low methylation direction. d It can be stipulated that a DNA fragment of cfDNA is in a high methylation rate or a low methylation rate, that is, it is assumed that a DNA fragment of cfDNA is located in a DNA methylation interval with a high methylation rate and / or a DNA methylation interval with a low methylation rate. By stipulating the methylation direction in this way, the screening method of the present application can search for markers in DNA methylation intervals with different methylation rate gradients, or search for markers separately in DNA methylation intervals with high / low methylation rates. Therefore, the screening method of the present application is more flexible than the existing marker screening method, and can be adapted to screening markers in a wider range of methylation intervals, rather than being limited to screening markers associated with high or low methylation levels, thereby improving the accuracy of methylation markers from cfDNA for disease detection.

[0099] Therefore, based on the methylation status information of the first population sample and the second population sample, the methylation rate α of the methylation interval of the first population sample and the second population sample can be calculated respectively. In this way, the methylation rate of all sequencing reads in the methylation interval can be obtained to further obtain the methylation rate of the methylation interval.

[0100] In this application, the methylation threshold α of the methylation interval cut and methylation direction m d It can be artificially specified according to the different diseases corresponding to the markers. Specifically, the methylation direction m of the methylation interval d It can include hypermethylation direction and hypomethylation direction. Those skilled in the art can understand that DNA methylation markers can characterize abnormal DNA methylation changes in diseased cells. This abnormal methylation runs through the entire process of disease occurrence and development. For example, the high methylation state of the promoter region of tumor suppressor genes leads to silencing of tumor suppressor genes, thereby causing cells to break away from the normal cell cycle and become cancerous; for another example, proto-oncogenes are more likely to be activated when they are in a low methylation state, which also increases the risk of tumor occurrence.

[0101] When the present application specifies the methylation direction m of the methylation interval d When the methylation direction m of the methylation interval is specified in the present application, the methylation pattern recognition result can be used to determine the markers of the differential methylation interval of high methylation in the first population sample and the second population sample; similarly, when the methylation direction m of the methylation interval is specified in the present application, d When α is hypo, the methylation pattern recognition result can be used to determine the markers of the differential methylation intervals of low methylation in the first population sample and the second population sample. Therefore, the methylation threshold α of the methylation interval in the two samples can be further determined based on the methylation rate α of the methylation interval. cut , to determine the methylation rate cutoff value of the differential methylation interval of high / low methylation in the first population sample and the second population sample. For example, specifying the methylation direction m of a certain methylation interval d is hypo, and its methylation threshold α cut =0.25, when the methylation rate α of a sequencing read segment of the methylation interval in the first population sample or the second population sample is ≤0.25, it can be considered that the sequencing read segment contains the differential information of the first population sample or the second population sample.

[0102] In this application, by setting the methylation direction m of different methylation intervals d Type and methylation threshold α cutThe value of can be used to determine the methylation pattern of the methylation interval, and the methylation interval markers of different susceptible diseases in high methylation state, low methylation state and intermediate methylation state can be determined to improve the detection rate of cfDNA with high / low methylation expression related to different diseases. This application can also determine different methylation patterns in DNA intervals of different types of cfDNA for the same disease to implement comprehensive detection of multiple methylation marker mining for the disease, or determine different methylation patterns in DNA intervals of the same cfDNA at different stages of disease development to achieve long-term tracking and evaluation of specific disease states.

[0103] It should be noted that since the methylation rate α is between 0 and 1, the methylation threshold α cut The value of is also between 0 and 1, but the methylation threshold α cut The value is between 0 and 1, and this application can be adjusted according to the situation without further restriction.

[0104] In the present application, after respectively obtaining the methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample, the cfDNA quantitative results of the first population sample and the second population sample can be calculated and obtained. In particular, the cfDNA quantitative results can be: the relative quantitative abundance of one or more cfDNAs, and the relative quantitative abundance of one or more cfDNAs can refer to the ratio of one cfDNA to other DNAs in each sample, which can characterize the development of the disease. For example, the quantitative results of ctDNA (tumor circulating DNA) of the first population sample and the second population sample are calculated and obtained to characterize the heterogeneity of tumor cells in the sample plasma. It should be noted that ctDNA is one type of cfDNA. When the diseases to which the markers detected in this application belong are different, the quantitative results of cfDNA are the quantitative results of one or more specific cfDNAs of different diseases, and are not limited to ctDNA.

[0105] In the present application, the cfDNA quantitative result is the relative quantitative abundance of one or more specific cfDNA methylation intervals. The calculation of the relative quantitative abundance of the cfDNA methylation interval is based on the methylation direction m of the previously determined methylation interval. d , methylation threshold α cut Specifically, the relative quantitative abundance of the methylation interval of cfDNA can be the ratio of the sequencing reads that meet the specified methylation pattern in all sequencing reads. For example, the methylation pattern of a certain methylation interval is specified as m d is hypo, α cut=0.25, and a total of m sequencing reads are detected in the methylation interval for a first population sample or a second population sample, where the self-methylation rate α≤α cut If there are m1 sequencing reads, the relative quantitative abundance of cfDNA in this methylation interval of the sample is m1 / m; in this way, the relative quantitative abundance of cfDNA in all methylation intervals with a specific methylation pattern can be obtained.

[0106] The relative quantitative abundance of cfDNA in the methylation intervals of the two samples can include the different methylation states exhibited by the methylation intervals of the two samples during the development of the disease. Therefore, the relative quantitative abundance of cfDNA can be used as a potential biomarker feature for sample classification and for learning the classification model. Therefore, the classification model finally obtained in this application can be used for the prediction of multiple diseases, which depends on the determination of the methylation pattern of the methylation interval.

[0107] The marker screening method based on cfDNA methylation pattern disclosed in the present application also includes:

[0108] In the characteristic interval screening step, based on the cfDNA quantification results, characteristic intervals that are different between the first population sample and the second population sample are screened out from the methylation intervals and determined as markers.

[0109] In the present application, screening out characteristic intervals that differ between the first population sample and the second population sample from the methylation intervals may include:

[0110] Methylation threshold α based on methylation interval cut , methylation direction m d and cfDNA relative quantitative abundance, and obtain the sensitivity value s of the methylation interval in the first population sample or the second population sample; for different methylation directions m d The methylation intervals are arranged in descending order according to the sensitivity value s; and the first m different methylation directions m are selected respectively. d The methylation interval is taken as the characteristic interval. Those skilled in the art will understand that in the methylation direction m d When set to hyper or hypo, the sensitivity value s of the methylation interval in the first population sample and the second population sample may be completely different. Therefore, this application independently obtains the two methylation directions m d The high-sensitivity value s methylation interval under the condition is selected so that the obtained feature interval contains classification information of two methylation directions, so as to comprehensively consider the methylation markers of the disease to be detected and improve the disease detection rate.

[0111] The sensitivity value s proposed in the present application may be the ratio of the first population samples / the second population samples to the total number of first population samples / the second population samples when the cfDNA quantitative result of a methylation interval in the first population samples or the second population samples is higher than the specified threshold. The methylation interval with the sensitivity value s has a greater potential to become a methylation marker than the methylation interval whose cfDNA quantitative result is not higher than the specified threshold; further, the higher the sensitivity value s, the stronger the ability of the methylation interval to detect diseased samples or healthy samples as a methylation marker, so the present application selects the methylation interval with the highest sensitivity value s as the characteristic interval and determines it as a marker.

[0112] It should be noted that the number of feature intervals obtained by the above screening method can be determined according to the number of methylation intervals previously obtained. When the number of methylation intervals with high sensitivity values ​​s is large, the dimension of the feature intervals can be appropriately increased. This application does not limit the specific number of feature intervals.

[0113] The present application also includes a marker screening system based on cfDNA methylation pattern, which includes:

[0114] A collection module, which obtains corresponding methylation intervals of the first population sample and the second population sample, and determines methylation status information of the methylation intervals of the first population sample and the second population sample respectively;

[0115] A pattern recognition quantitative module, based on the methylation status information, obtains methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample respectively; and

[0116] Based on the methylation pattern recognition results, the cfDNA quantitative results of the first population samples and the second population samples are obtained respectively;

[0117] The characteristic interval screening module, based on the cfDNA quantitative results, screens out characteristic intervals that are different between the first population samples and the second population samples from the methylation intervals, and determines them as markers.

[0118] It should be noted that the marker screening system of the present application can execute the marker screening method of the present application. Those skilled in the art can understand that the specific functions and operations of each module in the screening system have been introduced in detail in the description of the screening method, and therefore, their repeated description will be omitted.

[0119] As described above, the screening system of the present application can be implemented in various terminal devices, such as a server for constructing a disease prediction model based on cfDNA methylation markers, and further, a server for constructing a disease prediction model based on cfDNA methylation markers. In one embodiment, the screening system described in the present application can be integrated into a terminal device as a software module and / or a hardware module. For example, the screening system can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; the screening system can also be one of the many hardware modules of the terminal device.

[0120] Alternatively, in another embodiment, the screening system and the terminal device may be separate devices, and the screening system may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0121] In the present application, an electronic device is also included, comprising:

[0122] Processor; and

[0123] A memory stores computer program instructions, which, when executed by a processor, enable the processor to execute the marker screening method of the present application.

[0124] In the present application, the processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0125] The memory may include one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may, for example, include random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor may run the program instructions to implement the screening method described in the present application and / or other desired functions. Various contents such as sequencing data, methylation markers, etc. may also be stored in the computer-readable storage medium.

[0126] In one embodiment, the electronic device may further include: an input device and an output device, which are interconnected via a bus system and / or other forms of connection mechanisms. The input device may include, for example, a keyboard, a mouse, etc. The output device may output various information to the outside, including a trained disease prediction model based on cfDNA methylation markers, etc. The output device may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0127] The present application also includes a marker for cancer detection, wherein the marker comprises one or more of the following:

[0128] The genomic region from 41284213 to 41284519 of chromosome 1, the genomic region from 156814769 to 156815069 of chromosome 1, the genomic region from 30453757 to 30454057 of chromosome 2, the genomic region from 13589071 to 13589195 of chromosome 3, the genomic region from 50243493 to 50243657 of chromosome 4 The genomic region from positions 151177304 to 151177622 of chromosome 7, the genomic region from positions 99691599 to 99691943 of chromosome 7, the genomic region from positions 101838935 to 101839237 of chromosome 7, the genomic region from positions 17767751 to 17767994 of chromosome 8, the genomic region from positions 112081154 to 112081515 of chromosome 9, The genomic region from 126772025 to 126772354 of chromosome 11, the genomic region from 3313213 to 3313341 of chromosome 10, the genomic region from 48413903 to 48414022 of chromosome 10, the genomic region from 69516983 to 69517300 of chromosome 11, the genomic region from 120435185 to 120435492 of chromosome 11, The genomic region from 90921341 to 90921645 of chromosome 14, the genomic region from 103011099 to 103011399 of chromosome 14, the genomic region from 27893261 to 27893592 of chromosome 17, the genomic region from 32610596 to 32610787 of chromosome 17, and the genomic region from 5059596 to 5059899 of chromosome 19.

[0129] The present application also includes the use of markers in preparing a kit for cancer detection, wherein the markers include one or two or more of the following:

[0130] The genomic region from 41284213 to 41284519 of chromosome 1, the genomic region from 156814769 to 156815069 of chromosome 1, the genomic region from 30453757 to 30454057 of chromosome 2, the genomic region from 13589071 to 13589195 of chromosome 3, the genomic region from 50243493 to 50243657 of chromosome 4 The genomic region from positions 151177304 to 151177622 of chromosome 7, the genomic region from positions 99691599 to 99691943 of chromosome 7, the genomic region from positions 101838935 to 101839237 of chromosome 7, the genomic region from positions 17767751 to 17767994 of chromosome 8, the genomic region from positions 112081154 to 112081515 of chromosome 9, The genomic region from 126772025 to 126772354 of chromosome 11, the genomic region from 3313213 to 3313341 of chromosome 10, the genomic region from 48413903 to 48414022 of chromosome 10, the genomic region from 69516983 to 69517300 of chromosome 11, the genomic region from 120435185 to 120435492 of chromosome 11, The genomic region from 90921341 to 90921645 of chromosome 14, the genomic region from 103011099 to 103011399 of chromosome 14, the genomic region from 27893261 to 27893592 of chromosome 17, the genomic region from 32610596 to 32610787 of chromosome 17, and the genomic region from 5059596 to 5059899 of chromosome 19.

[0131] The present application also includes a composition for cancer detection, wherein the composition includes:

[0132] A nucleic acid for detecting the methylation status of the marker proposed in this application,

[0133] The methylation status of the marker is characterized by the methylation of the target sequence of the marker.

[0134] The present application also includes a kit, which contains reagents for detecting the markers proposed in the present application.

[0135] The marker screening method based on cfDNA methylation pattern proposed in this application, the method for screening multiple binary markers based on cfDNA methylation pattern, the acquisition of markers and the combination and application of markers can be conveniently applied to early non-invasive screening of diseases, and can also ensure that the markers have high sensitivity and specificity in detection; the method of this application not only considers the distance between CpG sites, but also considers the directionality between CpG sites based on the methylation state of the interval where the CpG sites are located, and then considers the difference in methylation levels between sites, calculates the characteristic values ​​of the difference intervals corresponding to diseased samples and healthy samples, so as to identify the disease information contained in various types of cfDNA in plasma. Therefore, by using the cfDNA markers obtained by the detection method of this application for early screening and multiple screening of diseases, the deterioration of the disease can be effectively avoided, thereby improving the subsequent treatment effect and survival rate.

[0136] Example

[0137] The present invention provides a general and / or specific description of the materials and test methods used in the test. The reagents or instruments used without indicating the manufacturer are all conventional reagent products that can be obtained through commercial purchase.

[0138] Example 1 Selection of marker methylation intervals

[0139] (1) Select a reference genome, whose version is hg19 (GRCh37, downloaded from UCSC Genome Browser), and merge CpG sites based on the proximity relationship between CpG sites in the genome sequence: if the relative distance between two adjacent CpG sites does not exceed d1 = 100 bp, then merge the two CpG sites to obtain and extend the merge interval. Find the next CpG site adjacent to the current CpG site, and continue to merge CpG sites that satisfy d1 ≤ 100 bp until the total length of the current extended merge interval reaches d2 = 300 bp, then no further extension of the merge interval is made, and the end point of the merge interval is determined, thereby obtaining a series of genomic intervals containing several adjacent CpG sites with an average length of no more than 300 bp. The currently obtained genomic interval is called MethBin.

[0140] (2) The MethBin interval obtained in (1) was further screened: if the number of CpG sites in the MethBin interval was ≥ 3, the MethBin intervals containing less than 3 CpG sites were removed; if the MethBin interval was required to fall within the CGI interval (downloaded from UCSC Genome Browser), the MethBin intervals that were partially or completely outside the CGI interval were removed; if the MethBin interval was required to fall within the range of cancer-related genes, the MethBin intervals that were outside the range of cancer-related genes were removed. Through the above methods, the target methylation interval was finally obtained, with a total length of 5.3 Mb.

[0141] (3) Based on the methylation intervals obtained in (2), probes designed and synthesized by Twist were used to obtain a panel for specifically capturing and enriching cfDNA molecules in the target region, wherein the ratio of effectively designed methylation intervals was set to 92%.

[0142] Example 2 Methylation pattern recognition and ctDNA quantification in methylation intervals

[0143] (1) Using 88 plasma samples of liver cancer (HCC) and 181 plasma samples of healthy people collected by Boercheng, 10 ml of blood samples were drawn from each sample and cfDNA samples were extracted therefrom. Then, panel capture and methylation library construction sequencing were performed as in Example 1, and sequencing reads were obtained from the cfDNA samples, which are called sequence reads. On average, 243M reads were obtained for each sample through sequencing.

[0144] (2) Through a series of downstream basic analyses, the alignment results of reads and the information of methylation sites in the methylation intervals were obtained. Specifically, first, the fastp (v0.23.2) quality control software was used to check the sequencing quality and remove low-quality sequencing bases or reads; then the bismark (v0.23.0) alignment software was used to align the quality-controlled data to the reference genome hg19, and the sequencing duplications were removed to obtain the deduplicated alignment results, which included the detailed alignment results of the reads and the methylation information of each CpG site on the reads.

[0145] (3) Further analyze the alignment results obtained in the previous step, including the detailed alignment results of reads and the methylation information of each site on the reads, to obtain the reads that meet the screening conditions and the methylation status of each CpG site therein, called readMeth; extract the information in readMeth and obtain the methylation status information of each DNA fragment. Specifically, the form of the extraction result of one DNA fragment is as follows:

[0146] 1_10484_10572

[0147] HWY5HDSX21

[0148] 10483,10488,10492,10496,10524,10541,10562,10570Z,z,Z,Z,Z,Z,Z,Z

[0149] Each element from top to bottom corresponds to the HMB interval index, read index, CpG site coordinates and the methylation coding state of each CpG site in the DNA fragment. Specifically, in the methylation coding state element of each CpG site, 'Z' and 'z' correspond to the methylated state and unmethylated state of the site, respectively.

[0150] (4) Determine the methylation rate of each DNA fragment by the methylation coding state of each CpG site in the DNA fragment. For example, for the methylation interval in the above example, according to the methylation state information in readMeth extracted in the previous step, the number of CpG sites detected in the DNA fragment is 8, of which 7 sites are methylated and 1 site is not methylated. Then the methylation rate of the DNA fragment is α=7 / (7+1)=0.875, and the value range of α is between 0-1. By the above calculation method, the α of all DNA fragments is obtained.

[0151] (5) For two given groups of samples, a methylation pattern is set for each corresponding methylation interval between the samples, and the methylation pattern recognition results are obtained for each DNA fragment MethBin in the methylation interval. Specifically, an appropriate α is set for each methylation interval. cut and m d , so that α is satisfied cut and m d The DNA fragments were almost undetectable in negative samples of healthy human plasma (with a specificity of no less than 0.9), but had a higher detection rate in HCC.

[0152] Based on the above methylation pattern recognition results, the α cut and m d , set the relative quantitative abundance of ctDNA in the methylation interval to the ratio of all DNA fragments with a specific methylation pattern in the methylation interval. Specifically, for the methylation interval where the DNA fragment obtained in (3) is located, set its methylation pattern to α cut =0.25, m d = hypo, when the α of the DNA fragment is ≤α cutWhen α≤0.25, the DNA fragment satisfies the methylation pattern of the methylation interval in which it is located. A total of 50 DNA fragments are detected in this methylation interval for an HCC, of ​​which 3 DNA fragments satisfy α≤0.25. The relative quantitative abundance of ctDNA in the sample is calculated to be 3 / 50=0.06.

[0153] Example 3 Feature interval screening

[0154] (1) 88 HCC plasma samples and 181 healthy human plasma samples were divided into training set and test set in a ratio of 2:1. The requirement was that the number of liver cancer samples and healthy samples in the training set was close to ensure the sample balance in the original input of the training set. The training set contained 21 HCC plasma samples and 160 healthy human plasma samples, and the test set contained 10 HCC plasma samples and 78 healthy human plasma samples. In order to further explore whether the binary classification model constructed based on this method can be applied to the classification prediction of benign tumors, we added 14 additional benign plasma samples (Benign) to the test set.

[0155] (2) The above methylation pattern recognition and ctDNA quantification steps were completed in the training set to obtain the methylation pattern and relative quantitative abundance of ctDNA in each MethBin interval. In addition, the sensitivity values ​​s of all MethBin intervals of HCC plasma samples were obtained, and all MethBin intervals were classified according to their methylation pattern m. d The species are divided into two groups, including m d =hyper's high methylation direction interval and m d = hypomethylation direction interval, the two groups of MethBin intervals are arranged in descending order according to the value of s, and the 10 markers with the best sensitivity are selected for the two groups of MethBin intervals. Finally, 20 characteristic intervals are obtained as markers for the construction and evaluation of subsequent classification models. Among them, the positions of the 20 characteristic intervals in the chromosome are shown in Table 1:

[0156] Table 1 Feature interval position information obtained in Example 3

[0157]

[0158]

[0159] Example 4 Model construction, classification evaluation and marker determination

[0160] The relative quantitative abundance values ​​of ctDNA in the 20 feature intervals of the training set were used as the input of the model, and a binary classification prediction model was constructed using the random forest algorithm. The test set samples were predicted and the prediction performance of the model was evaluated. Figure 1-Figure 2As shown in the figure, the predicted values ​​of healthy people, HCC patients and benign samples (the value range is between 0 and 1, and the larger the value, the greater the possibility of it being a positive sample / cancer sample). In the case of the healthy negative / cancer positive sample grouping in the training set and the test set, that is, the HCC vs. Healthy grouping, there is a significant degree of differentiation, as shown in Figure 3 As shown in the figure, the AUC of the test set reached 1. In addition, the prediction model also has a high specificity for the detection of benign tumor samples. In the case of liver cancer / non-liver cancer (healthy+benign) sample grouping, that is, HCC vs. Healthy&Benign, and the benign / liver cancer sample grouping in non-liver cancer, that is, Benign vs. HCC, the AUCs reached 0.98 and 0.9 respectively.

[0161] Obviously, in the present application, the methylation pattern of the methylation interval is obtained, and the relative quantitative abundance of the interval ctDNA is determined based on the methylation rate of the DNA fragments in the methylation interval. The interval markers that can distinguish cancer patients from healthy controls are screened out using the sensitivity of the methylation interval. The classification model trained using the marker has both high sensitivity and specificity in cancer detection, and is suitable for non-invasive screening and diagnosis of tumors. Especially in the diagnosis of cancer and screening in the early stages of tumor development, the interval markers of the present application have obvious advantages.

[0162] This specific embodiment is merely an explanation of the present application and is not a limitation of the present application. After reading this specification, those skilled in the art may make modifications to the present embodiment without any creative contribution as needed, but such modifications are protected by the patent law as long as they are within the scope of the claims of the present application.

Claims

1. A marker screening method based on cfDNA methylation pattern, comprising: A collection step, obtaining corresponding methylation intervals of the first population sample and the second population sample, and determining methylation status information of the methylation intervals of the first population sample and the second population sample respectively; A pattern recognition quantitative step, based on the methylation status information, respectively obtaining methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample; as well as Based on the methylation pattern recognition results, respectively obtaining cfDNA quantitative results of the first population sample and the second population sample; The characteristic interval screening step is to screen out characteristic intervals that are different between the first population sample and the second population sample from the methylation intervals based on the cfDNA quantification results, and determine them as markers.

2. The method according to claim 1, wherein: The step of respectively obtaining the corresponding methylation intervals of the first population sample and the second population sample comprises: Merge CpG sites with relative distance ≤ d1 to obtain the merged interval; The merged intervals are further screened to obtain the methylated intervals.

3. The method according to claim 2, wherein: The merged intervals are further screened to obtain the methylation intervals including: The merged interval is further screened using one or a combination of the following methods: Method 1: specifying that the interval length of the merged interval is ≤ d2; Method 2: Select the merged interval containing the number of CpG sites ≥ i; Method 3: Selecting the merged interval located in the methylation island (CGI); and Method 4: Select the merged interval within the range of known disease-related genes.

4. The method according to claim 1, wherein: The step of respectively determining the methylation status information of the methylation intervals of the first population sample and the second population sample comprises: The methylation status of all CpG sites in the corresponding methylation intervals of the first population sample and the second population sample are obtained respectively.

5. The method according to claim 1, wherein: The obtaining, based on the methylation status information, methylation pattern recognition results of the first population sample and the second population sample respectively includes: Based on the methylation status information of the first population sample and the second population sample, respectively calculating the methylation rates α of the methylation intervals of the first population sample and the second population sample; and Based on the methylation rate α, the methylation threshold α of the methylation interval of the first population sample and the second population sample is determined respectively. cut and methylation direction m d .

6. The method according to claim 1, wherein: The obtaining of cfDNA quantitative results of the first population sample and the second population sample respectively based on the methylation pattern recognition result comprises: Based on the methylation pattern recognition results, the relative quantitative abundance of cfDNA corresponding to the methylation interval of the first population sample and the second population sample is calculated respectively.

7. The method according to claim 5 or 6, wherein: The step of screening out characteristic intervals having differences between the first population sample and the second population sample from the methylation intervals includes: The methylation threshold α based on the methylation interval cut , the methylation direction m d and the relative quantitative abundance of the cfDNA, to obtain a sensitivity value s of the methylation interval in the first population sample or the second population sample; For different methylation directions m d The methylation intervals are arranged in descending order according to the sensitivity value s; and Select the first m different methylation directions m respectively d The methylation interval is used as the characteristic interval.

8. The method according to claim 1, wherein: The methylation pattern recognition results determine the differential methylation intervals of high methylation and / or low methylation in the first population sample and the second population sample to improve the detection rate of disease-related cfDNA.

9. A marker screening system based on cfDNA methylation pattern, comprising: A collection module, which obtains corresponding methylation intervals of the first population sample and the second population sample, and determines methylation status information of the methylation intervals of the first population sample and the second population sample respectively; A pattern recognition quantitative module, based on the methylation state information, obtains methylation pattern recognition results of the methylation intervals of the first population sample and the second population sample respectively; as well as Based on the methylation pattern recognition results, respectively obtaining cfDNA quantitative results of the first population sample and the second population sample; The feature interval screening module screens out feature intervals that are different between the first population sample and the second population sample from the methylation intervals based on the cfDNA quantification results, and determines them as markers.

10. The system according to claim 9, which is configured to execute the method according to any one of claims 1 to 8.

Images