Cancer marker detection method and system based on BS-seq data methylation vector

Abstract

The invention provides a cancer marker detection method and system based on a BS-seq data methylation vector, and the method comprises the steps: S1, obtaining a plurality of test samples of a cancer group and a test group, and cutting the BS-seq data of each test sample through a plurality of windows to obtain a plurality of fragments; s2, for each fragment, representing a methylated CpG site in the fragment by 1, and representing a non-methylated CpG site by 0 to obtain a methylated vector; s3, clustering all methylation vectors in each window to form a plurality of methylation vector clusters; s4, aiming at each methylation vector cluster, judging whether all methylation vectors in the methylation vector cluster are from a cancer group or not: if so, taking each methylation vector as a biomarker for cancer detection; and if not, exiting. The method has the beneficial effects that methylation state information of adjacent CpG sites can be fully utilized, the sequencing depth requirement on a single test sample is reduced, and the tumor purity influence of a cancer tumor sample is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of marker detection, and in particular to a cancer marker detection method and system based on BS-seq data methylation vector. Background Art

[0002] Clinically, there has always been a huge demand for non-invasive detection, risk stratification and treatment progress monitoring of tumors. In this regard, circulating tumor cells, extracellular vesicles, RNA, proteins and metabolites have made significant progress in recent years as potential non-invasive biomarkers, but cell-free DNA (cfDNA) has attracted the most attention. In a typical biological plasma specimen, the total cfDNA population represents a very rich source of biological and pathological information and shows great potential as a multifunctional biomarker in oncology, non-invasive prenatal testing and transplantation monitoring.

[0003] At present, the methylation analysis indicators of BS-seq data mainly rely on the methylation rate of CpG sites. CpG sites are sequences of cytosine (C) and guanine (G) that appear continuously on DNA. DNA methylation detection usually refers to the detection of methylation of cytosine in CpGs. However, the methylation rate only considers the methylation status of a single CpG site and cannot utilize the methylation information of adjacent CpG sites. Indicators such as methylation entropy and methylation haplotype can consider the methylation status of adjacent CpG sites. Fragments are DNA fragments used for library construction in the library construction stage of BS-seq experiments. These scalar indicators treat all fragments in each sample in a unified manner, requiring each sample to have a high sequencing depth. In addition, the above-mentioned methylation scalar indicators are greatly affected by the tumor purity of cancer samples, which will affect the detection accuracy. Summary of the invention

[0004] The technical problem to be solved by the present invention is to make full use of the methylation status information of adjacent CpG sites, reduce the sequencing depth requirements for a single test sample, and reduce the impact of tumor purity on cancer tumor samples. In order to overcome the defects of the above-mentioned prior art (or related technology), the present invention provides a cancer marker detection method and system based on BS-seq data methylation vector.

[0005] The present invention provides a cancer marker detection method based on BS-seq data methylation vector, comprising the following steps:

[0006] Step S1, obtaining multiple test samples of a cancer group and a test group, and cutting the BS-seq data of each test sample using multiple windows containing 4 CpGs to obtain multiple fragments;

[0007] Step S2, for each of the fragments, the methylated CpG sites in the fragment are represented by 1, and the unmethylated CpG sites are represented by 0, to obtain a corresponding methylation vector;

[0008] Step S3, clustering all the methylation vectors in each of the windows to form multiple methylation vector clusters;

[0009] Step S4, for each of the methylation vector clusters, determining whether all of the methylation vectors in the methylation vector cluster are derived from the cancer group:

[0010] If so, each of the methylation vectors in the methylation vector cluster is used as a biomarker for cancer detection;

[0011] If not, exit.

[0012] Compared with the prior art, the cancer marker detection method based on BS-seq data methylation vector in this application has the following advantages:

[0013] In the present application, the methylated CpG sites in each fragment are represented by 1, and the unmethylated CpG sites are represented by 0, so as to make full use of the methylation status information of adjacent CpG sites. In the present application, the BS-seq data is cut by multiple windows containing 4 CpGs to obtain multiple fragments, and then a methylation vector is generated for each fragment. It is not necessary for each test sample to have a high sequencing depth, which can reduce the sequencing depth requirement for a single test sample. In the present application, the test samples of the cancer group and the test group are uniformly processed, and biomarkers for cancer detection are obtained through fragment processing, vector generation, clustering operation and other steps, thereby reducing the influence of tumor purity of cancer tumor samples.

[0014] In a possible embodiment, in step S1, for each subject, 10 ml of blood is drawn from the subject, serum is separated, cfDNA is extracted using a cfDNA purification kit, and a BS-seq conversion kit is used to perform BS-seq library construction to obtain the corresponding test sample.

[0015] In a possible implementation, in step S3, the clustering method used is:

[0016] All the methylation vectors in each of the windows are merged to form a methylation matrix, and similar methylation vectors in each of the windows are clustered into clusters to obtain a plurality of methylation vector clusters.

[0017] In a possible implementation manner, the step S3 further includes:

[0018] During the clustering process, the source of each methylation vector in each window is recorded.

[0019] In a possible implementation, in step S3, an MRESC clustering algorithm is used to cluster all the methylation vectors in each of the windows.

[0020] Another technical solution of the present invention is to provide a cancer marker detection system based on BS-seq data methylation vector, applying the above-mentioned cancer marker detection method, including:

[0021] A sample collection module is used to obtain multiple test samples of the cancer group and the test group, and cut the BS-seq data of each test sample with multiple windows containing 4 CpGs to obtain multiple fragments;

[0022] a fragment processing module, connected to the sample collection module, for each of the fragments, representing the methylated CpG sites in the fragment with 1 and the unmethylated CpG sites with 0, to obtain a corresponding methylation vector;

[0023] a vector processing module, connected to the fragment processing module, clustering all the methylation vectors in each of the windows to form a plurality of methylation vector clusters;

[0024] a marker analysis module, connected to the vector processing module, for taking, for each methylation vector cluster, each methylation vector in the methylation vector cluster as a biomarker for cancer detection when all the methylation vectors in the methylation vector cluster are derived from the cancer group; or

[0025] When any of the methylation vectors in the methylation vector cluster is not derived from the cancer group, exit.

[0026] In a possible implementation, the vector processing module includes a clustering unit, which is used to merge all the methylation vectors in each of the windows to form a methylation matrix, and cluster similar methylation vectors in each of the windows into clusters to obtain a plurality of methylation vector clusters.

[0027] In a possible implementation manner, the vector processing module further includes a recording unit connected to the clustering unit, and configured to record the source of each of the methylation vectors in each of the windows during the clustering process of the clustering unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of the method steps of the present invention;

[0029] Figure 2 It is the system structure principle diagram of the present invention;

[0030] Figure 3 It is a schematic diagram of the process flow of the actual operation of the present invention;

[0031] Explanation of the accompanying drawings: 1. Sample collection module; 2. Fragment processing module; 3. Vector processing module; 31. Clustering unit; 32. Recording unit; 4. Marker analysis module. DETAILED DESCRIPTION

[0032] First, those skilled in the art should understand that these implementations are only used to explain the technical principles of the embodiments of the present application, and are not intended to limit the protection scope of the embodiments of the present application. Those skilled in the art can make adjustments to them as needed to adapt to specific application scenarios.

[0033] The present application is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] See also Figure 1 The embodiment of the present application discloses a cancer marker detection method based on BS-seq data methylation vector, including step S1: sample acquisition and cutting, step S2: methylation vector generation, step S3: clustering processing; step S4: marker screening, wherein step S1 is specifically: obtaining multiple test samples of a cancer group and a test group, and cutting the BS-seq data of each test sample with multiple windows containing 4 CpGs to obtain multiple fragments; step S2 is specifically: for each fragment, the methylated CpG site in the fragment is represented by 1, and the non-methylated CpG site is represented by 0 to obtain the corresponding methylation vector; step S3 is specifically: clustering all methylation vectors in each window to form multiple methylation vector clusters; step S4 is specifically: for each methylation vector cluster, judging whether all methylation vectors in the methylation vector cluster are derived from the cancer group: if so, each methylation vector in the methylation vector cluster is used as a biomarker for cancer detection; if not, exit.

[0035] Continue to see Figure 1 In this application, a vector indicator for measuring methylation status is proposed for the first time: methylation vector, which has the following advantages compared with existing methylation indicators: 1. It can make full use of the methylation status information of adjacent CpG sites; 2. It has low sequencing depth requirements for a single test sample; 3. It is not affected by the tumor purity of cancer tumor samples.

[0036] Continue to see Figure 1 , this application mainly includes three stages in the actual operation process:

[0037] The first stage is vectorization: the BS-seq data of each test sample is cut into windows containing 4 CpGs. For the part of each fragment in the window, the methylated CpG site is represented by 1, and the unmethylated CpG site is represented by 0. In this way, the part of each fragment in the window can be represented by a vector, which we call a methylation vector. Multiple methylation vectors in a window can form a methylation matrix reference. Figure 3 The "vectorization" part in the box part is a 0-1 matrix;

[0038] The second stage, clustering: merge the methylation vectors in each window of all test samples in the cancer group and the test group. For example, the methylation vectors of a window in each test sample are: v1, v2, v3, ... vn, and merge these vectors into a matrix V merge = [v1, v2, v3, ... vn], cluster all methylation vectors in each window, and after clustering, similar methylation vectors in each window are clustered into clusters;

[0039] The third stage, separation: During the second stage of merge clustering, the sources of all methylation vectors in each window will be recorded, and all windows will be traversed. If all methylation vectors in a methylation vector cluster in the window come from the cancer group, then this methylation vector cluster can be considered a cancer-related cluster, and the methylation vectors in this methylation vector cluster can be used as biomarkers for cancer detection.

[0040] Continue to see Figure 1 When merging clusters in the third stage, you can choose existing clustering algorithms such as K-means, DBSCAN or HDBSCAN. In this application, the MRESC multiple repeats and equal spacing clustering algorithm designed specifically for methylation vector analysis is used.

[0041] See also Figure 2The embodiment of the present application discloses a cancer marker detection system based on BS-seq data methylation vector, including a sample collection module 1, a fragment processing module 2, a vector processing module 3, and a marker analysis module 4, wherein the sample collection module 1 is used to obtain multiple test samples of a cancer group and a test group, and the BS-seq data of each test sample is cut with multiple windows containing 4 CpGs to obtain multiple fragments; the fragment processing module 2 is used to, for each fragment, represent the methylated CpG site in the fragment with 1, and the non-methylated CpG site with 0, to obtain the corresponding methylation vector; the vector processing module 3 is used to cluster all methylation vectors in each window to form multiple methylation vector clusters; the marker analysis module 4 is used to, for each methylation vector cluster, when all methylation vectors in the methylation vector cluster are derived from the cancer group, use each methylation vector in the methylation vector cluster as a biomarker for cancer detection; or exit when there is any methylation vector in the methylation vector cluster that does not come from the cancer group.

[0042] Continue to see Figure 2 The vector processing module 3 includes a clustering unit 31, which is used to merge all methylation vectors in each window to form a methylation matrix, and cluster similar methylation vectors in each window into clusters to obtain multiple methylation vector clusters.

[0043] Continue to see Figure 2 The vector processing module 3 further includes a recording unit 32 connected to the clustering unit 31, and used to record the source of each methylation vector in each window during the clustering process of the clustering unit.

[0044] In the description of the present application, the description with reference to the terms "one embodiment", "some embodiments", "in the present embodiment", "specific example", or "some examples" etc. means that the specific features, mechanisms, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, mechanisms, materials or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0045] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A cancer marker detection method based on methylation vector of BS-seq data, characterized in that: The following steps are involved: Step S1, obtaining multiple test samples of a cancer group and a test group, and cutting the BS-seq data of each test sample using multiple windows containing 4 CpGs to obtain multiple fragments; Step S2, for each of the fragments, the methylated CpG sites in the fragment are represented by 1, and the unmethylated CpG sites are represented by 0, to obtain a corresponding methylation vector; Step S3, clustering all the methylation vectors in each of the windows to form multiple methylation vector clusters; Step S4, for each of the methylation vector clusters, determining whether all of the methylation vectors in the methylation vector cluster are derived from the cancer group: If so, each of the methylation vectors in the methylation vector cluster is used as a biomarker for cancer detection; If not, exit.

2. The method for detecting cancer markers according to claim 1, characterized in that: In step S1, for each subject, 10 ml of blood is drawn from the subject, serum is separated, cfDNA is extracted using a cfDNA purification kit, and a BS-seq conversion kit is used to perform BS-seq library construction to obtain the corresponding test sample.

3. The method for detecting cancer markers according to claim 1, characterized in that: In step S3, the clustering method used is: All the methylation vectors in each of the windows are merged to form a methylation matrix, and similar methylation vectors in each of the windows are clustered into clusters to obtain a plurality of methylation vector clusters.

4. The method for detecting cancer markers according to claim 1, characterized in that: The step S3 further includes: During the clustering process, the source of each methylation vector in each window is recorded.

5. The method for detecting cancer markers according to claim 1, characterized in that: In the step S3, the MRESC clustering algorithm is used to cluster all the methylation vectors in each of the windows.

6. A cancer marker detection system based on methylation vectors of BS-seq data, characterized in that: The method for detecting a cancer marker according to any one of claims 1 to 5 comprises: A sample collection module (1) is used to obtain multiple test samples of the cancer group and the test group, and cut the BS-seq data of each test sample using multiple windows containing 4 CpGs to obtain multiple fragments; a fragment processing module (2), connected to the sample collection module (1), for representing the methylated CpG sites in each fragment with 1 and the unmethylated CpG sites in the fragment with 0, to obtain a corresponding methylation vector; a vector processing module (3), connected to the fragment processing module (2), clustering all the methylation vectors in each of the windows to form a plurality of methylation vector clusters; a marker analysis module (4), connected to the vector processing module (3), for, for each of the methylation vector clusters, when all the methylation vectors in the methylation vector cluster are derived from the cancer group, using each of the methylation vectors in the methylation vector cluster as a biomarker for cancer detection; or When any of the methylation vectors in the methylation vector cluster is not derived from the cancer group, exit.

7. The cancer marker detection system according to claim 6, characterized in that: The vector processing module (3) comprises a clustering unit (31) for merging all the methylation vectors in each of the windows to form a methylation matrix, and clustering similar methylation vectors in each of the windows into clusters to obtain a plurality of methylation vector clusters.

8. The method for detecting cancer markers according to claim 7, characterized in that: The vector processing module (3) further comprises a recording unit (32), connected to the clustering unit (31), and used for recording the source of each methylation vector in each window during the clustering process of the clustering unit.

Images