Method for screening non-small cell lung cancer markers based on enhancer methylation differences, markers thereof, and applications

Abstract

The present invention discloses a method for screening non-small cell lung cancer markers based on enhancer methylation differences, as well as the markers and applications thereof. Abnormal DNA methylation in the enhancer regions of the human genome can lead to abnormal gene expression regulation in non-small cell lung cancer. The method for screening non-small cell lung cancer markers described in the present invention is based on this principle and further combines paired-end tag sequencing analysis of chromatin interaction technology (ChIA-PET) to screen out DNA methylation markers in the genomic enhancer regions that cause abnormal gene expression regulation in non-small cell lung cancer. This screening method has the advantages of accuracy, rapidity, and high throughput. The DNA methylation markers described in the present invention are cg00787780, cg16434331, cg21862081, and cg24327132. The markers have good diagnostic efficacy for non-small cell lung cancer, showing high accuracy, high sensitivity, and strong specificity, and can better help improve the detection rate of non-small cell lung cancer.

Description

Technical Field

[0001] The present invention relates to the field of biomedicine, and specifically relates to a method for screening non-small cell lung cancer markers based on enhancer methylation differences, as well as the markers and applications thereof. Background Art

[0002] Lung cancer is one of the most common cancers worldwide and is also a major cause of cancer death. Non-small cell lung cancer (NSCLC), as the main histological type, accounts for approximately 85% of newly diagnosed lung cancers. Since most patients are diagnosed with advanced metastases upon detection, the prognosis of lung cancer patients is usually poor. The 5-year survival rate of patients diagnosed with metastatic lung cancer is less than 10%. In contrast, relying on surgery, the prognosis of patients with early-stage diagnosed lung cancer is much better, with a 5-year survival rate as high as 80%. Screening for lung cancer using low-dose helical computed tomography (CT) has been proven to improve the early diagnosis rate of lung cancer, thus helping to reduce the mortality rate.

[0003] Although screening based on low-dose CT is very sensitive, its low specificity can result in many false positives, which may lead to further follow-up or invasive procedures. Therefore, more accurate novel biomarkers are needed for clinical lung cancer diagnosis.

[0004] Aberrant DNA methylation is a common event in various cancers, indicating that DNA methylation can be used as a biomarker for cancer diagnosis. Compared with other biomarkers such as proteins and gene mutations, the changes in DNA methylation are very stable and occur at the early stage of cancer.

[0005] The mainstream lung cancer DNA methylation markers mainly focus on the abnormal DNA methylation regions in the promoters or within the genes of tumor suppressor genes or oncogenes. The latest research shows that abnormal DNA methylation in the distal regulatory regions of genes, especially in enhancer regions, can also cause the occurrence and development of cancer.

[0006] Currently, there is no reported method for screening non-small cell lung cancer markers based on enhancer methylation differences. Summary of the Invention

[0007] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a method for screening non-small cell lung cancer markers based on enhancer methylation differences, as well as the markers and applications thereof.

[0008] To achieve the above purpose, the present invention designs a method for screening non-small cell lung cancer markers based on enhancer methylation differences, including the following steps:

[0009] S1. Obtain the whole-genome DNA methylation data of non-small cell lung cancer samples and the whole-genome DNA methylation data of normal lung tissue samples respectively using WGBS sequencing technology;

[0010] S2. Use the BatMeth2 software to perform DNA methylation difference analysis on the whole-genome DNA methylation data of non-small cell lung cancer samples and the whole-genome DNA methylation data of normal lung tissue samples to obtain the genomic location information of differentially methylated regions (DMRs).

[0011] S3. Respectively obtain the genomic location information of enhancer regions in the lung tissue genome or the genomic location information of enhancer regions in non-small cell lung cancer tissue-related cell lines (such as A549) through the EnhancerAtlas 2.0 database. Combine the genomic location information of the differentially methylated regions (DMRs) obtained in step S2, and retain the DMRs that have an intersection with the genomic location of the enhancer region, that is, obtain the enhancer region differentially methylated regions (eDMRs).

[0012] S4. Use the chromatin interaction analysis with paired-end tag sequencing (ChIA-PET) technology to obtain the RNAPOL2 ChIA-PET data of lung tissue or the RNAPOL2 ChIA-PET data of non-small cell lung cancer-related cell lines (such as A549). Use the ChIA-PET Tool V3 software to obtain the whole-genome three-dimensional chromatin interaction information of RNAPOL2 in lung tissue or the whole-genome three-dimensional chromatin interaction information of RNAPOL2 in non-small cell lung cancer-related cell lines (such as A549). Combine the enhancer region differentially methylated regions (eDMRs) obtained in step S3 to obtain the whole-genome enhancer region differentially methylated regions (eDMR)-gene promoter chromatin interaction information with one end located in the enhancer region differentially methylated region (eDMR) and the other end located in the gene promoter.

[0013] S5. Use the whole-genome enhancer region differentially methylated region (eDMR)-gene promoter interaction pair information obtained in step S4 to screen for eDMRs that have chromatin interaction with the gene promoter.

[0014] S6. Respectively obtain the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples through Illumina Infinium HumanMethylation 450K BeadChip (450K methylation chip detection technology).

[0015] S7. Based on the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6, use statistical methods to obtain the differentially methylated sites (DMC1) in non-small cell lung cancer samples and the differentially methylated sites (DMC2) between normal lung tissue samples; based on the eDMRs screened in step S5, screen out the differentially methylated sites located within the eDMRs in the genome as candidate markers;

[0016] S8. Using the Lasso method, with the methylation levels of the candidate marker differentially methylated sites obtained in S7 as independent variables and the sample types in the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6 as dependent variables, establish a linear regression model on the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6, and retain the differentially methylated sites with non-zero coefficients in the model as the final markers;

[0017] S9. Use the 450K methylation chip detection technology to obtain the DNA methylation data sets of non-small cell lung cancer samples and normal lung tissue samples (data sets other than those in S6) respectively, or download the DNA methylation data sets of non-small cell lung cancer samples and normal lung tissue samples (data sets other than those in S6) from publicly available databases (such as the National Genomics Data Center NGDC in China); use the obtained data sets to construct a logistic regression model,

[0018] Use the logistic regression model to evaluate the efficacy of the markers obtained in step S8 for the diagnosis of non-small cell lung cancer.

[0019] Further, in the said step S2, the screening criteria for the genomic location information of the differentially methylated region (DMR) are as follows:

[0020] The differential methylation degree of the differentially methylated region (DMR): greater than or equal to 0.25 and less than or equal to 1 or greater than or equal to -1 and less than or equal to -0.25, and the multiple testing FDR is less than 0.05 and greater than 0.

[0021] Still further, in the said step S8, the sample types are set as:

[0022] Normal lung tissue samples are encoded as 0, and non-small cell lung cancer samples are encoded as 1.

[0023] Still further, in the said step S8, the final markers are four, namely cg00787780, cg16434331, cg21862081, and cg24327132.

[0024] Furthermore, in the step S9, the equation of the logistic regression model is as follows:

[0025] Y = 26.671 – 10.644X1 – 4.376X2 – 15.556X3 – 11.790X4;

[0026] where X1 is cg00787780, X2 is cg16434331, X3 is cg21862081, and X4 is cg24327132;

[0027] Y represents the probability that the sample has non-small cell lung cancer, and 0 ≤ Y ≤ 1.

[0028] Furthermore, in the step S9, the classification performance criteria of the logistic regression model are as follows:

[0029] When 0.5 < classification accuracy rate ≤ 1 and 0.5 < AUC ≤ 1, it indicates that the biomarker is effective (the closer the classification accuracy rate and AUC are to 1, the better the classification performance of the biomarker);

[0030] When the classification accuracy rate ≤ 0.5 or AUC ≤ 0.5, it indicates that the biomarker is ineffective.

[0031] The present invention also provides a DNA methylation biomarker screened by the above method, and the biomarker is cg00787780, cg16434331, cg21862081, cg24327132.

[0032] The present invention also provides an application of a kit for detecting the above DNA methylation biomarker in a sample in the preparation of products for diagnosing and predicting non-small cell lung cancer.

[0033] Explanation of important terms of the present invention:

[0034] 1. WGBS refers to Whole Genome Bisulfite Sequencing, which can accurately detect the methylation level of all single cytosine bases (C bases) at the whole genome level and is the gold standard for DNA methylation research.

[0035] 2. Illumina Infinium HumanMethylation 450K BeadChip refers to the 450K methylation chip technology, which can detect approximately 450,000 DNA methylation sites in the human whole genome and is a relatively common and cost-effective DNA methylation detection technology.

[0036] 3. DNA methylation refers to the methylation state of the 5th carbon atom of cytosine in CpG dinucleotides. As a relatively stable modification state, under the action of DNA methyltransferase, it can be inherited to the newly generated daughter DNA during DNA replication, and it is an important epigenetic mechanism. Methylation in 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 hypermethylation of tumor suppressor genes and DNA repair genes, hypomethylation of repetitive sequence DNA, and loss of imprinting of certain genes, which is related to the occurrence of various tumors.

[0037] 4. ChIA-PET refers to the technology of chromatin interaction analysis with paired-end-tag sequencing, which is a technology that integrates chromatin immunoprecipitation (ChIP), chromatin proximity ligation, paired-end tags, and high-throughput sequencing technology to study long-range chromatin interactions across the genome.

[0038] 5. A promoter is a segment of DNA sequence that RNA polymerase recognizes, binds to, and starts transcription. It contains conserved sequences required for specific binding of RNA polymerase and transcription initiation, and most are located upstream of the gene transcription start point. The promoter itself is not transcribed.

[0039] 6. An enhancer is a small segment of DNA that can bind to proteins. After binding to proteins, the transcriptional activity of the gene will be enhanced. The enhancer may be located upstream or downstream of the gene. And it does not necessarily linearly approach the gene to be affected on the genome, because of the winding structure of chromatin, which allows positions far apart in the sequence to have the opportunity to contact each other in space.

[0040] 7. RNAPOL2 refers to RNA polymerase II (RNA polymerase II, also known as RNAPⅡ or PolⅡ), which is a multi-protein complex. It is one of the three RNA polymerases found in the eukaryotic nucleus. It catalyzes the transcription of DNA to synthesize the precursors of mRNA, most snRNA, and microRNA. RNA polymerase II is a 550 kDa complex of 12 subunits and is the most studied type of RNA polymerase. It requires multiple transcription factors to bind to the upstream gene promoter and start transcription.

[0041] 8. Specificity refers to the ratio of patient samples without specific clinical diseases being detected as negative.

[0042] 9. Sensitivity refers to the ratio of patient samples with a definite clinical disease being detected as positive.

[0043] 10. AUC (Area Under Curve) is defined as the area enclosed by the ROC and the coordinate axes, and the value of this area ranges from 0 to 1. The closer the AUC is to 1, the higher the authenticity of the detection method.

[0044] 11. ROC (receiver operating characteristic curve) refers to the receiver operating characteristic curve, which is a comprehensive index reflecting continuous variables of sensitivity and specificity.

[0045] 12. FDR refers to the Bonferroni correction: If N independent hypotheses are simultaneously tested on the same data set, then the statistical significance level for each hypothesis should be 1 / N of the statistical significance level when only one hypothesis is tested.

[0046] Advantages of the present invention:

[0047] 1. The method for screening non-small cell lung cancer markers of the present invention is a method for screening DNA methylation markers based on abnormal DNA methylation in enhancer regions that have chromatin three-dimensional interactions with gene promoter regions. This method has the advantages of precision, rapidity, and high throughput. Through this method, the screening range of DNA methylation markers for non-small cell lung cancer diagnosis in the current field can be expanded. And since the mechanism of cancer caused by abnormal DNA methylation in enhancer regions and promoter regions or within-gene regions is different, this method is also expected to complement conventional DNA methylation marker screening methods to better improve the detection rate of non-small cell lung cancer.

[0048] 2. DNA methylation markers have extremely high sensitivity and specificity. Compared with other biomarkers such as proteins and gene mutations, DNA methylation changes are very stable and appear in the early stages of cancer, which is helpful for the early screening of non-small cell lung cancer. In addition, the method for screening non-small cell lung cancer markers based on abnormal enhancer DNA methylation can obtain more highly reliable cancer diagnosis markers, which is helpful for promoting the clinical application of DNA methylation markers, ultimately achieving accurate, efficient, economical, and non-invasive early cancer screening, and improving people's quality of life. Description of the Drawings

[0049] Figure 1 It is a flow chart of the method for screening non-small cell lung cancer DNA methylation markers based on enhancer methylation differences.

[0050] Figure 2Heatmap of the methylation levels of eDMRs in 17 samples. Each row represents an eDMR, and each column represents a sample. NC represents normal lung tissue samples, and TC represents non-small cell lung cancer samples. The colors in the heatmap represent the zscore values of the methylation levels.

[0051] Figure 3 Distribution map of the methylation levels of the biomarker in normal lung tissue and non-small cell lung cancer samples at each stage. The significance test method is the Wilcoxon sum rank test.

[0052] Figure 4 ROC curve graph of the training set of normal lung tissue and non-small cell lung cancer samples. AUC represents the area under the ROC curve. When the classification threshold is 0.743, the specificity is 0.972, and the sensitivity is 0.988. Detailed implementation manners

[0053] The present invention will be further described in detail below in conjunction with specific embodiments for the understanding of those skilled in the art.

[0054] Embodiment 1

[0055] A method for screening non-small cell lung cancer biomarkers based on enhancer methylation differences, comprising the following steps:

[0056] S1. Use the WGBS sequencing technology to obtain the whole-genome DNA methylation data of 8 non-small cell lung cancer samples and the whole-genome DNA methylation data of 9 normal lung tissue samples respectively. The specific information is shown in Table 1 below.

[0057] Table 1. Statistics of WGBS methylation information

[0058]

[0059] S2. Use the BatMeth2 software to perform DNA methylation difference analysis on the whole-genome DNA methylation data of non-small cell lung cancer samples and the whole-genome DNA methylation data of normal lung tissue samples to obtain the genomic location information of differentially methylated regions (DMRs); among them, the screening criteria for the genomic location information of differentially methylated regions (DMRs) are as follows:

[0060] The differential methylation degree of the differentially methylated region (DMR): greater than or equal to 0.25 and less than or equal to 1 or greater than or equal to -1 and less than or equal to -0.25, and the multiple test FDR is less than 0.05 and greater than 0.

[0061] S3. Obtain the genomic location information of the genomic enhancer regions of the non-small cell lung cancer tissue-related cell line A549 through the EnhancerAtlas 2.0 database. Combine the genomic location information of the differentially methylated regions (DMRs) obtained in step S2, and retain the DMRs that have an intersection with the genomic locations of the enhancer regions, that is, obtain the enhancer region differentially methylated regions (eDMRs). A total of 2595 enhancer region differentially methylated regions (eDMRs) are obtained here, and their methylation information in each sample is shown in Figure 2 , and the colors in the figure represent the standardized values of the eDMR methylation levels.

[0062] S4. Use the paired-end tag chromatin interaction analysis by paired-end tag sequencing (ChIA-PET) technology to obtain the RNAPOL2 ChIA-PET data of the non-small cell lung cancer-related cell line A549, and use the ChIA-PET Tool V3 software to obtain the three-dimensional chromatin interaction information of the whole genome of A549.

[0063] S5. Based on the previous 2595 eDMR regions, obtain the eDMR region-gene promoter interaction pair information with one end located at the gene promoter and the other end located at the eDMR region, so as to obtain the eDMRs that have three-dimensional chromatin interactions with the gene promoter.

[0064] S6. Obtain the 450K methylation chip dataset of normal lung tissues and non-small cell lung cancer samples from the TCGA database (number of non-small cell lung cancer samples: 807, number of normal samples: 71).

[0065] S7. Use statistical methods to obtain the information of differentially methylated sites (DMCs) in the 450K methylation chip dataset. Combine the eDMRs that have three-dimensional chromatin interactions with the gene promoter obtained in S5, and retain the DMCs whose genomic locations are in these eDMR regions as candidate DNA methylation markers. We obtained 14 candidate markers, and the specific results are shown in Table 2. The 14 CpG sites in Table 2 are the candidate DNA methylation markers, and they are all located in the genomic enhancer regions.

[0066] Table 2. Candidate markers of enhancer DNA methylation

[0067]

[0068] * The genomic locations are referenced to the human reference genome version hg38.

[0069] S8. Using the 14 CpG sites in Table 2, a Lasso regression model was established for feature screening of 878 non-small cell lung cancer samples (n = 807) and normal lung tissue samples (n = 71). CpG sites with a coefficient of 0 in the model were deleted, and finally 4 CpG sites were obtained, which were the final enhancer DNA methylation marker sites for non-small cell lung cancer, as shown in Table 3.

[0070] Table 3. Four selected enhancer DNA methylation marker sites for non-small cell lung cancer

[0071]

[0072] To verify the effectiveness of the DNA methylation markers, we examined the DNA methylation levels of these markers in normal lung tissue samples and non-small cell lung cancer samples at various pathological stages. These 4 CpG sites all showed abnormal DNA methylation and also showed significant differences at the Stage I. The specific results are shown in Figure 3 , and the Wilcoxon sum rank test was used for statistical analysis in the figure, with the sample size n = 878.

[0073] S9. Evaluate the efficacy of the enhancer DNA methylation markers for non-small cell lung cancer. Using the methylation information of these 4 CpG sites, a logistic regression model was established for 878 samples in TCGA. The equation of the model is:

[0074] Y = 26.671 – 10.644X1 – 4.376X2 – 15.556X3 – 11.790X4, where X1 is cg00787780, X2 is cg16434331, X3 is cg21862081, and X4 is cg24327132;

[0075] Y represents the probability of the sample having non-small cell lung cancer, and 0 ≤ Y ≤ 1.

[0076] Taking P = 0.743 as the screening threshold for Y, if Y is greater than or equal to P, the sample is determined to have non-small cell lung cancer; otherwise, the sample is determined to be a normal sample. The sensitivity in the training set (878 samples in TCGA) is 0.972, and the specificity is 0.988. The ROC curve of the marker in the training set is shown in Figure 4, where the AUC value is 0.996. Thus, it can be seen that the DNA methylation markers in these 4 enhancer regions have good classification performance. On the other hand, we evaluated the performance of the classification model established with these 4 methylation markers in the data of two other test sets, and also compared the classification performance of 3 other reported DNA methylation markers in promoter regions. The results showed that the DNA methylation markers in enhancer regions have more stable and better classification performance. The specific results are shown in Table 4 and Table 5.

[0077] Table 4. Validation of the performance of DNA methylation markers in enhancer regions on two independent data sets

[0078]

[0079] Note: Data Set data set; Tumor number of cancer samples; Normal number of normal lung tissue samples; TP true positive; TN true negative; FP false positive; FN false negative

[0080] Table 5. Comparison of the classification performance with previously reported DNA methylation markers for lung cancer

[0081]

[0082]

[0083] Note: AUC is the area under the ROC curve; Accuracy is the classification accuracy. The corresponding literature for the previously reported DNA methylation markers for lung cancer is:

[0084] 1. Li M, Zhang C, Zhou L, Li S, Cao YJ, Wang L, Xiang R, Shi Y & Piao Y (2020) Identification and validation of novel DNA methylation markers for early diagnosis of lung adenocarcinoma. Mol Oncol 14, 2744 - 2758.

[0085] 2. Diaz-Lagares A, Mendez-Gonzalez J, Hervas D, Saigi M, Pajares MJ, Garcia D, Crujerias AB, Pio R, Montuenga LM & Zulueta J (2016) A Novel Epigenetic Signature for Early Diagnosis in Lung Cancer Epigenetic Signature for Lung Cancer Diagnosis. Clin Cancer Res 22, 3361 - 3371.

[0086] 3. Dong S, Li W, Wang L, Hu J, Song Y, Zhang B, Ren X, Ji S, Li J & Xu P (2019) Histone-related genes are hypermethylated in lung cancer and hypermethylated HIST1H4F could serve as a pan-cancer biomarker. Cancer Res 79, 6101 - 6112.

[0087] Example 2 Kit for Detecting the Above DNA Methylation Biomarkers in a Test Sample

[0088] This kit detects the methylation levels of these four biomarkers in the test biopsy sample or blood, substitutes them into the above model equation, and determines whether the test subject has non-small cell lung cancer according to the above threshold. On the other hand, these biomarkers can also be used to construct a new and more accurate model equation in a larger-scale non-small cell lung cancer clinical cohort for the early diagnosis and screening of non-small cell lung cancer.

[0089] Other parts not described in detail are prior art. Although the above embodiments have described the present invention in detail, they are only a part of the embodiments of the present invention, not all of them. People can also obtain other embodiments based on this embodiment without creative efforts, and these embodiments all fall within the protection scope of the present invention.

Claims

1. A method for screening non-small cell lung cancer markers based on enhancer methylation differences, characterized in that: It includes the following steps: S1. Use the WGBS sequencing technology to obtain the whole-genome DNA methylation data of non-small cell lung cancer samples and the whole-genome DNA methylation data of normal lung tissue samples respectively; S2. Use the BatMeth2 software to perform DNA methylation difference analysis on the whole-genome DNA methylation data of non-small cell lung cancer samples and the whole-genome DNA methylation data of normal lung tissue samples, and obtain the genomic location information of differentially methylated regions; S3. Respectively obtain the genomic location information of enhancer regions in the lung tissue genome or the genomic location information of enhancer regions in non-small cell lung cancer tissue-related cell lines through the EnhancerAtlas 2.0 database. Combine the genomic location information of the differentially methylated regions DMR obtained in step S2, and retain the DMR that intersects with the genomic location of the enhancer region, that is, obtain the enhancer region differentially methylated region eDMR; S4. Use the paired-end tag sequencing technology to obtain the RNAPOL2 ChIA-PET data of lung tissue or the RNAPOL2 ChIA-PET data of non-small cell lung cancer-related cell lines. Use the ChIA-PET Tool V3 software to obtain the RNAPOL2 whole-genome three-dimensional chromatin interaction information of lung tissue or the RNAPOL2 whole-genome three-dimensional chromatin interaction information of non-small cell lung cancer-related cell lines. Combine the enhancer region differentially methylated region obtained in step S3 to obtain the whole-genome enhancer region differentially methylated region-gene promoter chromatin interaction information with one end located in the enhancer region differentially methylated region and the other end located in the gene promoter; S5. Use the whole-genome enhancer region differentially methylated region-gene promoter interaction pair information obtained in step S4 to screen out the eDMR with chromatin interaction with the gene promoter; S6. Respectively obtain the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples through the Illumina Infinium HumanMethylation 450K BeadChiP; S7. Based on the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6, use statistical methods to obtain the differentially methylated sites between non-small cell lung cancer samples and the differentially methylated sites between normal lung tissue samples; Based on the eDMR screened in step S5, screen out the differentially methylated sites whose genomic locations are within the eDMR as candidate markers; S8. Using the Lasso method, with the methylation levels of the candidate biomarker differential DNA methylation sites obtained in S7 as independent variables and the sample types in the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6 as dependent variables, a linear regression model is established on the DNA methylation data of non-small cell lung cancer samples and the DNA methylation data of normal lung tissue samples collected in S6, and the differential DNA methylation sites with non-zero coefficients in the model are retained as the final biomarkers; S9. Using the 450K methylation chip detection technology to obtain the DNA methylation data sets of non-small cell lung cancer samples and normal lung tissue samples respectively, or downloading the DNA methylation data sets of non-small cell lung cancer samples and normal lung tissue samples from publicly available databases respectively; constructing a logistic regression model using the obtained data sets, and using the logistic regression model to evaluate the efficacy of the biomarkers obtained in step S8 for the diagnosis of non-small cell lung cancer, where the DNA methylation data set is an independent data set rather than the data set in S6.

2. The method for screening non-small cell lung cancer markers based on enhancer methylation differences according to claim 1, characterized in that: In step S2, the screening criteria for the genomic location information of the differential DNA methylation regions are as follows: The differential methylation level of the differential DNA methylation region: greater than or equal to 0.25 and less than or equal to 1 or greater than or equal to -1 and less than or equal to -0.25, and the multiple test FDR is less than 0.05 and greater than 0.

3. The method for screening non-small cell lung cancer markers based on enhancer methylation differences according to claim 1, characterized in that: In step S8, the sample types are set as: The normal lung tissue samples are encoded as 0, and the non-small cell lung cancer samples are encoded as 1.

4. The method for screening non-small cell lung cancer markers based on enhancer methylation differences according to claim 1, wherein: In step S9, the efficacy criteria are as follows: When 0.5 < classification accuracy rate ≤ 1 and 0.5 < AUC ≤ 1, it indicates that the biomarker is effective; When the classification accuracy rate ≤ 0.5 or AUC ≤ 0.5, it indicates that the biomarker is ineffective.

Images