System, device or medium for diagnosing or predicting breast cancer based on a combination of methylation markers

Abstract

The present application discloses a system, device or medium for diagnosing or predicting breast cancer based on a combination of methylation markers, which belongs to the field of tumor marker technology. Wherein, the methylation marker combination includes a methylation gene combination, and the methylation gene combination includes MLLT6, AR1, MED24, KCTD11, AADAT, PTPRN2, ACTR1A, CELSR1, ZC3H7A, TRIM8, IGLL5, CORO2B, THADA, CNNM1, SLC16A3, DEAF1 and LCP2. Using the methylation level data of the methylation marker combination of the present application, a machine learning model is constructed, which can be used to diagnose whether a subject has breast cancer or predict whether a subject has the risk of breast cancer, and has great clinical application value.

Description

Technical Field

[0001] The present application relates to the technical field of tumor markers, and in particular to a system, device or medium for diagnosing or predicting breast cancer based on a combination of methylation markers. Background Art

[0002] DNA methylation chips can comprehensively analyze the degree of gene methylation in breast cancer samples, revealing which genes show abnormal DNA methylation levels in breast cancer. These differentially expressed methylation levels may be potential biomarkers for diagnosis, prognosis or treatment.

[0003] With the rapid development of methylomics methods, non-invasive breast cancer detection methods based on DNA methylation chips are increasingly showing great application potential. However, there is still a lack of clinically useful breast cancer detection markers. Summary of the Invention

[0004] In order to solve at least one of the above technical problems, the technical solution adopted in this application is as follows.

[0005] In a first aspect, the present application provides the use of a methylation level detection reagent of a methylation marker combination in the preparation of a kit for diagnosing or predicting breast cancer, wherein the methylation marker combination includes a methylation gene combination, and the methylation gene combination includes MLLT6, AR1, MED24, KCTD11, AADAT, PTPRN2, ACTR1A, CELSR1, ZC3H7A, TRIM8, IGLL5, CORO2B, THADA, CNNM1, SLC16A3, DEAF1 and LCP2.

[0006] In this application, the diagnosis is an auxiliary diagnosis and needs to be made in combination with other clinical indicators. If the subject does not show other clinical features of breast cancer, the subject is considered to be at risk of breast cancer and requires medical intervention.

[0007] In some embodiments of the present application, the detection reagent is used to detect the methylation level of the methylation site of each gene in the methylation gene combination.

[0008] Furthermore, the methylation sites of the MLLT6 gene include cg01485645; the methylation sites of the AR1 gene include cg06982805; the methylation sites of the MED24 gene include cg14700282; the methylation sites of the KCTD11 gene include cg02676175; the methylation sites of the AADAT gene include cg00309402; the methylation sites of the PTPRN2 gene include cg14189678; the methylation sites of the ACTR1A gene include cg09800781; the methylation sites of the CELSR1 gene include cg04332442; the methylation sites of the ZC3H7A ... The methylation sites of the TRIM8 gene include cg06388937; the methylation sites of the IGLL5 gene include cg23477395; the methylation sites of the CORO2B gene include cg23597015; the methylation sites of the THADA gene include cg17277833; the methylation sites of the CNNM1 gene include cg25842285; the methylation sites of the SLC16A3 gene include cg07016605; the methylation sites of the DEAF1 gene include cg22132501; and the methylation sites of the LCP2 gene include cg01199321.

[0009] Thus, the methylation marker combination includes the following methylation sites: cg01485645, cg06982805, cg14700282, cg02676175, cg00309402, cg14189678, cg09800781, cg04332442, cg04010868, cg06388937, cg23477395, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501 and cg01199321.

[0010] Furthermore, the methylation marker combination also includes at least one of the following methylation sites: cg07677157, cg06123699, cg27469726, cg12277627, cg12804791 and cg17551400.

[0011] In some embodiments of the present application, the methylation markers include at least one of the following methylation sites: cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678, cg0980078 1, cg04332442, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321.

[0012] In some embodiments of the present application, the methylation markers include the following methylation sites: cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678, cg09800781, ...1485645, cg07677157, cg06982805, cg14700282, cg0 g04332442, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321.

[0013] In some embodiments of the present application, the detection reagent includes a methylation detection chip.

[0014] A second aspect of the present application provides a method for constructing a model for diagnosing or predicting breast cancer, comprising the following steps:

[0015] S1, obtaining methylation level data of any one of the methylation marker combinations described in the first aspect of the present application in biological samples of a population, wherein the population includes breast cancer patients and non-breast cancer subjects;

[0016] S2. Using the methylation level data obtained in step S1, a machine learning model is constructed and multi-fold cross-validation is performed.

[0017] In some embodiments of the present invention, the method further comprises:

[0018] S3: Verify the machine learning model obtained in step S2 in the validation set.

[0019] In some embodiments of the present application, the methylation level is obtained using one of methylation-sensitive bisulfite sequencing (Bisulfite Sequencing), high-resolution melting analysis (MS-HRM), methylation chip, reduced genome methylation sequencing (RRBS), whole genome bisulfite methylation sequencing (WGBS) and immunoprecipitation sequencing (MeDIP-seq).

[0020] Bisulfite sequencing is one of the most commonly used methods for DNA methylation analysis. It detects DNA methylation levels by converting unmethylated cytosine (C) to uracil (U), thereby converting cytosine bands into uracil bands, while methylated cytosine (C) remains unchanged. Bisulfite-treated DNA fragments are then selectively amplified by PCR, followed by DNA sequencing. Finally, the measured sequence is aligned with the original sequence to count the methylation sites and their number, and analyze the degree of methylation.

[0021] MS-HRM detects methylation based on differences in DNA melting curves. DNA treated with sodium bisulfite is amplified by PCR, and the amplified product melts during HRM analysis. Different methylation states result in different melting curves.

[0022] Methylation arrays, also known as methylation chips, can simultaneously detect the methylation status of thousands of CpG sites in a large number of samples. Fluorescently labeled DNA is hybridized with probes immobilized on the chip, and the methylation level is determined by detecting the hybridization signal.

[0023] RRBS uses bisulfite treatment and restriction enzyme digestion to enrich for highly methylated DNA fragments. This method can rapidly and efficiently detect highly methylated CpG sites in the genome.

[0024] WGBS is a high-throughput method that can measure the methylation levels of all individual cytosine bases across the genome. This method requires deep sequencing coverage to obtain accurate methylation data.

[0025] MeDIP-seq uses anti-5-methylcytosine antibody immunoprecipitation followed by high-throughput sequencing, enabling the methylation status of all CpG sites in the genome to be determined.

[0026] In some embodiments of the present application, the machine learning model is selected from any one of the following:

[0027] Logistic regression model, support vector machine model, decision tree model, random forest model, neural network model, XGBoost model, linear discriminant analysis model, GBDT model, ADABoost model, naive Bayes model, CatBoost model, LightGBM model, MLP model and ETC model.

[0028] A third aspect of the present application provides a system for diagnosing or predicting breast cancer, comprising the following modules:

[0029] a data input module for inputting methylation level data of a methylation marker combination obtained in a biological sample of a subject, wherein the methylation marker combination includes a methylation gene combination, and the methylation gene combination includes MLLT6, AR1, MED24, KCTD11, AADAT, PTPRN2, ACTR1A, CELSR1, ZC3H7A, TRIM8, IGLL5, CORO2B, THADA, CNNM1, SLC16A3, DEAF1, and LCP2;

[0030] a database storage module for storing methylation level data of the methylation marker combination in biological samples of a population, the population comprising breast cancer patients and non-breast cancer subjects;

[0031] A disease prediction module is connected to the data input module and the database storage module, respectively, and is used to construct a machine learning model using the methylation level data of the methylation marker combination in the biological samples of the population, and to diagnose whether the subject has breast cancer or predict whether the subject has a risk of developing breast cancer based on the methylation level data of the methylation marker combination in the biological samples of the subject obtained from the data input module.

[0032] In some embodiments of the present application, the detection reagent is used to detect the methylation level of the methylation site of each gene in the methylation gene combination.

[0033] Furthermore, the methylation sites of the MLLT6 gene include cg01485645; the methylation sites of the AR1 gene include cg06982805; the methylation sites of the MED24 gene include cg14700282; the methylation sites of the KCTD11 gene include cg02676175; the methylation sites of the AADAT gene include cg00309402; the methylation sites of the PTPRN2 gene include cg14189678; the methylation sites of the ACTR1A gene include cg09800781; the methylation sites of the CELSR1 gene include cg04332442; the methylation sites of the ZC3H7A ... The methylation sites of the TRIM8 gene include cg06388937; the methylation sites of the IGLL5 gene include cg23477395; the methylation sites of the CORO2B gene include cg23597015; the methylation sites of the THADA gene include cg17277833; the methylation sites of the CNNM1 gene include cg25842285; the methylation sites of the SLC16A3 gene include cg07016605; the methylation sites of the DEAF1 gene include cg22132501; and the methylation sites of the LCP2 gene include cg01199321.

[0034] Thus, the methylation marker combination includes the following methylation sites: cg01485645, cg06982805, cg14700282, cg02676175, cg00309402, cg14189678, cg09800781, cg04332442, cg04010868, cg06388937, cg23477395, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501 and cg01199321.

[0035] Furthermore, the methylation marker combination also includes at least one of the following methylation sites: cg07677157, cg06123699, cg27469726, cg12277627, cg12804791 and cg17551400.

[0036] In some embodiments of the present application, the methylation markers include at least one of the following methylation sites: cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678, cg0980078 1, cg04332442, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321.

[0037] In some embodiments of the present application, the methylation markers include the following methylation sites: cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678, cg09800781, ...1485645, cg07677157, cg06982805, cg14700282, cg0 g04332442, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321.

[0038] In a fourth aspect, the present application provides a computer device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any method described in the second aspect of the present application when executing the computer program.

[0039] In a fifth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of any method described in the second aspect of the present application are implemented.

[0040] Compared with the prior art, the invention of this application has the following beneficial effects:

[0041] Using the methylation level data of the methylation marker combination of the present application, a machine learning model is constructed, which can be used to diagnose whether a subject has breast cancer or predict whether a subject has the risk of breast cancer, and has great clinical application value.

[0042] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The above and other objects, features and advantages of the exemplary embodiments of the present application will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present application are shown in an illustrative and non-limiting manner, in which:

[0044] Figure 1 The figure shows the accuracy curve of training and cross-validation using the GBDT model in Example 2 of the present application;

[0045] Figure 2 The cross entropy loss curve for training and cross validation using the GBDT model in Example 2 of the present application is shown;

[0046] Figure 3 The ROC curve of the GBDT model in the independent validation set in Example 3 of the present application is shown;

[0047] Figure 4 The confusion matrix of the classification using the GBDT model in the independent validation set in Example 3 of the present application is shown. DETAILED DESCRIPTION

[0048] Unless otherwise indicated, implied from the context, or customary in the art, all parts and percentages in this application are based on weight, and the test and characterization methods used are current as of the filing date of this application. Where applicable, the contents of any patents, patent applications, or publications referred to in this application are incorporated herein by reference in their entirety, and their equivalent patent families are also incorporated by reference, particularly for definitions of relevant terms in the art disclosed in such documents. If the definition of a specific term disclosed in the prior art is inconsistent with any definition provided in this application, the definition of the term provided in this application shall prevail.

[0049] In order to make the technical problems, technical solutions and beneficial effects solved by the present application clearer and more understandable, the present application is further described in detail below in conjunction with the embodiments.

[0050] The following examples are provided herein to illustrate preferred embodiments of the present application. Those skilled in the art will appreciate that the techniques disclosed in the following examples represent techniques discovered by the inventors that can be used to implement the present application and, therefore, can be considered preferred embodiments of the present application. However, those skilled in the art will appreciate, based on this specification, that many modifications may be made to the specific embodiments disclosed herein while still achieving the same or similar results without departing from the spirit or scope of the present application.

[0051] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs, and the disclosure and materials cited therein are hereby incorporated by reference.

[0052] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many technical equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the claims.

[0053] The experimental methods in the following examples, unless otherwise specified, are all conventional methods. The instruments and equipment used in the following examples, unless otherwise specified, are all conventional laboratory instruments and equipment; the experimental materials used in the following examples, unless otherwise specified, are all purchased from conventional biochemical reagent stores.

[0054] Example 1 Marker Screening

[0055] 1. Inclusion and exclusion criteria of the study cohort

[0056] (1) Inclusion criteria for breast cancer patients:

[0057] Aged between 19 and 72 years old;

[0058] (2) Exclusion criteria for breast cancer patients:

[0059] BRCA test negative.

[0060] (3) Controls were subjects without breast cancer (negative mammogram).

[0061] Using the above criteria, peripheral blood samples were collected from 256 breast cancer patients and 258 non-breast cancer subjects at the Affiliated Cancer Hospital of Harbin Medical University.

[0062] 2. Sample Collection

[0063] Genomic DNA was extracted from whole blood using the QIAamp DNA Blood Kit (Qiagen). DNA concentration was determined using a quantitative fluorometric dsDNA system (Promega), and fluorescence readings at 504 nm Ex / 531 nm Em were measured using a 96-well plate reader (TECAN). DNA quality was assessed using a NanoDrop ND-1000 spectrophotometer (TermoSciff).

[0064] 3. Methylation Analysis

[0065] DNA methylation profiles of peripheral blood samples were analyzed using the Illumina Infnium Methylation EPIC array, a genome-wide methylation screening tool that targets over 850,000 CpG sites in biologically significant regions of the human methylome.

[0066] Methylation analysis was performed on at least 600 ng of genomic DNA obtained from each participant. The genomic DNA was first bisulfite-converted using the EZ DNA Methylation Kit (Zymo Research). The resulting bisulfite-converted DNA was amplified, hybridized to a methylation array, and scanned with an Illumina scanner according to standard procedures.

[0067] 4. Data Analysis

[0068] (1) Remove the control sample and the value of the Pval column in the methylation chip, and retain the methylation chip signal value;

[0069] (2) removing sites where the methylation chip signal value was NA in all samples;

[0070] (3) Normalize all signal values based on samples;

[0071] (4) Using the Boruta algorithm to screen methylation site markers:

[0072] The Boruta algorithm is a feature selection method based on random forests. It is mainly used to identify which features in a dataset are important and to exclude unimportant or redundant features.

[0073] Using the Boruta algorithm to select features is very useful for improving the performance and interpretability of machine learning models because it can reduce the number of input features, thereby avoiding overfitting and speeding up the training process. Boruta's n_estimators=100 was set, and other parameters were the default algorithm parameters. 23 characteristic methylation sites were screened out, as shown in Table 1.

[0074] Table 1 Information of 23 characteristic methylation sites

[0075]

[0076] The methylation sites in Table 1 can be used as markers for diagnosing or predicting breast cancer. By detecting the methylation levels of the methylation sites, it is possible to determine whether a subject has breast cancer or is at risk of developing breast cancer.

[0077] Example 2 Construction of a breast cancer diagnosis or prediction model

[0078] The inventors used different machine learning models: logistic regression model, support vector machine model, decision tree model, random forest model, XGBoost model, linear discriminant analysis model, GBDT model, ADABoost model, naive Bayes model, CatBoost model, LightGBM model, MLP model and ETC model to establish a breast cancer diagnosis model, and performed model evaluation through 5-fold cross-validation. The area under the curve (AUC) of the receiver operating characteristic curve (ROC) was used to evaluate the model classification efficiency. The results are shown in Table 2.

[0079] Table 2 Classification performance of different machine learning models

[0080]

[0081] As shown in Table 2, the GBDT model has the best classification performance. The LightGBM model, logistic regression model, and CatBoost model also achieved good classification performance.

[0082] The accuracy curve and cross entropy loss curve for training and verification using the GBDT model are as follows: Figure 1 and Figure 2 As shown. Figure 1 and Figure 2 It can be seen that as the number of training set samples increases, the accuracy of the validation set tends to stabilize and the cross entropy loss is reduced to the minimum.

[0083] In summary, based on the characteristic methylation sites obtained by screening, the model constructed using GBDT can accurately distinguish breast cancer patients and thus can be used for auxiliary diagnosis of breast cancer.

[0084] Example 3 Independent validation of the model

[0085] In order to verify the performance of the model constructed using GBDT in Example 2, the inventors further collected 56 independent breast cancer patients and 49 non-breast cancer subjects, obtained the methylation levels of the characteristic methylation sites screened in Example 1, and used the model for classification.

[0086] The ROC curve and the classification result confusion matrix are as follows: Figure 3 and Figure 4 shown.

[0087] comprehensive Figure 3 and Figure 4 The performance data of the model built using GBDT in the independent validation set are shown in Table 3. Figure 4 It can be seen that when the model built using GBDT is used for prediction, the number of true positives (TP) is 49 and the number of false positives (FN) is 0; the number of true negatives (TN) is 49 and the number of false negatives (FN) is 7.

[0088] Therefore, the accuracy, precision, and recall of the model built using GBDT in the independent validation set are calculated.

[0089] (1) Accuracy

[0090] Used to evaluate the correct prediction ratio in the results of model prediction.

[0091] Accuracy = (TP+TN) / (TP+FP+TN+FN) = 98 / 106 = 0.93

[0092] (2) Precision

[0093] Used to evaluate the proportion of samples predicted to be breast cancer by the model that are correctly predicted.

[0094] Precision = TP / (TP+FP) = 49 / 49 = 1

[0095] (3) Recall

[0096] Used to evaluate the proportion of samples predicted by the model to be breast cancer to the number of actual breast cancer cases in the sample.

[0097] Recall = TP / (TP+FN) = 49 / 56 = 0.88

[0098] (4) F1 score

[0099] The F1 score is the harmonic mean of precision and recall.

[0100]

[0101] Based on the above results, the performance of the model built using GBDT in the independent validation set is summarized in Table 3.

[0102] Table 3 Performance of the model built using GBDT in the independent validation set

[0103]

[0104] In summary, based on the characteristic methylation sites obtained by screening, the model constructed using GBDT in Example 2 still has very good discrimination performance in independent validation data.

[0105] All documents mentioned in this application are incorporated herein by reference, just as if each document were incorporated herein by reference individually. It should also be understood that after reading the above teachings of this application, those skilled in the art may make various changes or modifications to this application, and that such equivalents also fall within the scope of the claims appended hereto.

Claims

1. Use of a methylation level detection reagent of a methylation marker combination in the preparation of a kit for diagnosing or predicting breast cancer, characterized in that: The methylation marker combination includes methylation sites cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678, cg09800781, and cg043324 42, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321.

2. A method for constructing a model for diagnosing or predicting breast cancer, characterized in that: The following steps are involved: S1, obtaining methylation level data of the methylation marker combination according to claim 1 in biological samples of a population, wherein the population includes breast cancer patients and non-breast cancer subjects; S2. Using the methylation level data obtained in step S1, a machine learning model is constructed and multi-fold cross-validation is performed.

3. The method according to claim 2, characterized in that The machine learning model is selected from any one of the following: Logistic regression model, support vector machine model, decision tree model, random forest model, neural network model, XGBoost model, linear discriminant analysis model, GBDT model, ADABoost model, naive Bayes model, CatBoost model, LightGBM model, MLP model and ETC model.

4. A system for diagnosing or predicting breast cancer, characterized in that: Includes the following modules: A data input module is used to input the methylation level data of the methylation marker combination obtained in the subject's biological sample, wherein the methylation marker combination includes methylation sites cg01485645, cg07677157, cg06982805, cg14700282, cg06123699, cg27469726, cg02676175, cg00309402, cg14189678 , cg09800781, cg04332442, cg12277627, cg04010868, cg06388937, cg12804791, cg23477395, cg17551400, cg23597015, cg17277833, cg25842285, cg07016605, cg22132501, and cg01199321; a database storage module for storing methylation level data of the methylation marker combination in biological samples of a population, the population comprising breast cancer patients and non-breast cancer subjects; A disease prediction module is connected to the data input module and the database storage module, respectively, and is used to construct a machine learning model using the methylation level data of the methylation marker combination in the biological samples of the population, and to diagnose whether the subject has breast cancer or predict whether the subject has a risk of developing breast cancer based on the methylation level data of the methylation marker combination in the biological samples of the subject obtained from the data input module.

5. A computer device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the method according to any one of claims 2 to 3 when executing the computer program.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 2 to 3 are implemented.

Images