Screening method and kit of methylation markers for identifying and diagnosing benign and malignant pulmonary nodules and application of methylation markers for identifying and diagnosing benign and malignant pulmonary nodules

Abstract

The invention discloses a screening method of methylation markers for identifying and diagnosing benign and malignant pulmonary nodules, a kit and application. The method comprises the following steps: carrying out methylation difference analysis and reverse screening on a locally constructed pulmonary nodule methylation data set and a lung adenocarcinoma DNA methylation data set in TCGA, and carrying out analysis and filtration on consistency of candidate markers in tissues and paired plasma, so as to obtain 201 candidate methylation markers through screening; a diagnostic model is further constructed based on a machine learning method, and it is verified that when the combination of 30 methylation markers is used for identifying and diagnosing the benign and malignant pulmonary nodules, the sensitivity is high, the specificity is good, and the methylation markers can be used for preparing products for detecting the benign and malignant pulmonary nodules.

Description

[0001] The application is a divisional application, the original application number of which is "202310062741.1", the original filing date of which is January 19, 2023, and the invention name of which is "A set of methylation markers for differential diagnosis of lung nodule benignity and malignancy, and screening method and application thereof". TECHNICAL FIELD

[0002] The application belongs to the technical field of analysis, and particularly relates to a screening method, kit and application of a methylation marker for differential diagnosis of lung nodule benignity and malignancy. BACKGROUND

[0003] Lung cancer is one of the malignant tumors with the highest incidence and mortality in China. In the past three decades, the mortality rate of lung cancer has increased by about 5 times, and a major reason is that the early features of lung cancer are not obvious, and 75% of cancer patients are diagnosed in the middle and late stages. The early detection rate of lung cancer is less than 25%, but the 5-year survival rate of early lung cancer can be more than 90%.

[0004] At present, low-dose helical computed tomography (LDCT) screening for early lung cancer has been recognized, but with the popularization of LDCT screening, 50% of people receiving LDCT screening will detect lung nodules. Lung nodules refer to local, round, and density-increased lung shadows with a diameter of less than or equal to 30 mm. Among them, 95% of lung nodules are caused by benign lesions, including granulomas, lymph nodes, chronic inflammation, hamartoma, etc., and malignant lesions mainly include adenocarcinoma, squamous cell carcinoma, etc. In order to achieve accurate diagnosis and treatment of lung nodules, major clinical medical centers at home and abroad have developed a series of diagnosis and treatment paths and risk prediction models, but there is still an over-diagnosis rate of 18-25%, especially for patients with imaging as solid and sub-millimeter size, the proportion of postoperative pathological confirmation of benign diseases is as high as 30%, which not only brings serious psychological burden to patients, but also causes waste of national medical and health resources. Therefore, the differential diagnosis of lung nodule benignity and malignancy has become a clinical diagnosis and treatment pain point and research hotspot, and is also a major demand for healthy China construction and economic development.

[0005] At present, there are various markers and technical means for the benign and malignant differential diagnosis of lung nodules. The serological tumor markers commonly used in clinical practice, such as CEA, SCC, Cfra21-1, ProGRP and NSE, have certain value for the auxiliary diagnosis and differential diagnosis of tumors, but the sensitivity and specificity of the lung nodule benign and malignant differential diagnosis by the markers alone are very low, the joint detection cannot meet the clinical needs, and the detection rate of clinical stage I lung cancer is not more than 20%; sputum exfoliative cytology is convenient and economical, non-invasive, and has high patient acceptance, but the sensitivity is very low, and can only play a prompting role in the diagnosis of lung cancer; circulating tumor cells (CTC) are related to the stage of lung cancer, and the sensitivity of early lung cancer diagnosis reaches 67.2%, but the in vitro CTC detection technology is easily restricted by sample size, and less studied in early diagnosis and screening.

[0006] More and more new evidences show that in the occurrence and development of tumors, methylation abnormalities occurring in the epigenetic level are more common than somatic mutations. DNA methylation of different genes is highly related to tumor types, has high tissue specificity in the same individual, and the methylation characteristic spectrum of the same tissue of different individuals also has high consistency. Since the epigenetic modification represented by DNA methylation often occurs in the early stage of cancer, and compared with somatic mutations and copy number variations, it has high tissue specificity, and is more suitable for the research of lung nodule differential diagnosis markers in principle.

[0007] Circulating tumor DNA (ctDNA) as a kind of free DNA (cfDNA) molecule in liquid biopsy carries tumor-specific genetic and epigenetic changes. Compared with tissue biopsy, ctDNA has the advantages of real-time, convenience and non-invasiveness, making it a new type of tumor marker that continues to attract attention. More and more studies have shown that characteristic methylation "fingerprint" patterns can be used for early diagnosis and staging of cancer, efficacy evaluation, recurrence monitoring and prognosis judgment. However, at present, there are few lung nodule-related markers based on ctDNA methylation, and the diagnostic performance is limited.

[0008] Therefore, it is of great significance to develop a lung nodule benign and malignant differential diagnosis marker based on ctDNA methylation with low cost, non-invasiveness and high specificity and high sensitivity suitable for clinical promotion, for the scientific management of lung nodule population and the effective control of the incidence of lung cancer. SUMMARY

[0009] In order to solve the problems in the prior art, the application discloses a screening method, kit and application of a methylation marker for lung nodule benign and malignant differential diagnosis.

[0010] The first aspect of this invention provides the use of a reagent for detecting the methylation level of a methylation biomarker in the preparation of products for detecting benign and malignant lung nodules and / or lung cancer, wherein the methylation biomarker comprises the following 30 methylation regions:

[0011] chr7: 143059947-143060107; chr5: 178487412-178487572; chr11: 15136199-15136359; chr17: 43972911-4397 3071; chr12: 64062951-64063111; chr3: 87039622-87039782; chr6: 27835283-27835443; chr5: 146257858-1462 58018; chr16: 28075032-28075192; chr17: 47307472-47307632; chr19: 37407294-37407454; chr1: 179545118- 179545278; chr13: 20806316-20806476; chr7: 132261297-132261457; chr4: 17783234-17783394; chr1: 6777355 8-67773718; chr10: 105037463-105037623; chr19: 7735166-7735326; chr6: 26199948-26200108; chr6: 262733 57-26273517; chr12: 4381931-4382091; chr7: 30722046-30722206; chr5: 175792494-175792654; chr11: 695903 60-69590520; chr8: 143858458-143858618; chr14: 59104962-59105122; chr4: 57976315-57976475; chr4: 39529282-39529442; chr7: 55259381-55259541; chr10: 90343107-90343267; wherein, the methylation regions of the methylation markers are determined based on human genome hg19 alignment.

[0012] In another aspect, the present invention provides a kit for detecting benign or malignant pulmonary nodules and / or lung cancer, the kit comprising reagents for detecting the methylation levels of the first 30 methylation markers.

[0013] In another aspect, the present invention provides a method for screening methylation markers for the differential diagnosis of benign and malignant pulmonary nodules, comprising the following steps:

[0014] S1. By performing RRBS analysis on lung nodules and adjacent tissues, a local lung nodule methylation dataset was constructed. Based on the Illumina 450K LUAD DNAMethylation dataset (lung adenocarcinoma Illumina 450k methylation chip) in TCGA, methylation difference analysis was performed on the two datasets to screen out tissue-specific methylation regions of malignant lung nodules, which were denoted as the first screening DMR.

[0015] S2, reverse-screen the first screened DMR obtained in step S1, and remove methylated regions that meet the following criteria to obtain the second screened DMR:

[0016] (1) Methylated regions derived from blood cells and other organs;

[0017] (2) The target detection area is less than 4 CpG;

[0018] (3) Methylation regions in blood cells with an average methylation value greater than 0.03;

[0019] S3, based on the second screening DMR obtained in step S2, a set of candidate methylation markers is obtained through analysis and filtering of tissue and paired plasma consistent with each other;

[0020] S4. The candidate methylation biomarkers obtained in step S3 are subjected to machine learning in the plasma training and test sets to screen for methylation regions with significant differences, thereby obtaining methylation biomarkers for the differential diagnosis of benign and malignant pulmonary nodules.

[0021] Preferably, in step S1, the first screening DMR meets the following criteria in the methylation difference analysis of the two datasets:

[0022] (a) The mean methylation value in the control samples was less than 0.02;

[0023] (b) The ratio of the mean methylation value of the positive sample to that of the control sample is greater than 3.0;

[0024] (c) The average methylation value in positive samples is greater than 0.05;

[0025] (d) DMR region greater than 20bp;

[0026] The positive sample is malignant nodule tissue; the control samples are tissue adjacent to malignant nodules, benign nodule tissue, and tissue adjacent to benign nodules.

[0027] Compared with the prior art, the beneficial effects of the present invention are:

[0028] This invention discloses a screening method, kit, and application for methylation biomarkers used in the differential diagnosis of benign and malignant pulmonary nodules. The screening method yields 201 candidate methylation biomarkers. Furthermore, a diagnostic model is constructed based on machine learning, which verifies that the combination of 30 methylation biomarkers has high sensitivity and specificity in the differential diagnosis of benign and malignant pulmonary nodules and can be used to prepare products for the detection of benign and malignant pulmonary nodules and / or lung cancer. Attached Figure Description

[0029] Figure 1 This is a heatmap of P201 in different paired samples from 30 lung cancer patients in Example 1 of the present invention.

[0030] Figure 2 The ROC curve of the diagnostic model in Example 2 is shown below.

[0031] A represents the ROC curves of the diagnostic model constructed using the P30 methylation biomarker set on the training and validation sets;

[0032] B represents the ROC curves of the diagnostic model constructed using the P45 methylation biomarker set on the training and validation sets;

[0033] C represents the ROC curves of the diagnostic model constructed using the P42 methylation biomarker set on the training and validation sets. Detailed Implementation

[0034] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0035] Experimental methods in the following embodiments of the present invention, unless otherwise specified, are generally performed under conventional conditions, such as those described in Sambrook et al., Molecular Cloning: A Laboratory Manual (New York: Cold Spring Harbor Laboratory Press, 1989), or as recommended by the manufacturer. All commonly used chemical reagents used in the embodiments are commercially available products. Unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention.

[0036] The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps is not limited to the steps or modules listed, but may optionally include steps not listed, or may optionally include other steps inherent to such process, method, product, or device.

[0037] In this invention, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0038] The patients with malignant pulmonary nodules mentioned in this invention are lung cancer patients. The two terms have the same meaning and can be used interchangeably.

[0039] As mentioned above, in view of the shortcomings of the prior art, the applicant of this invention first performed RRBS analysis on 50 pairs of clinical pulmonary nodules and adjacent tissues to construct a local pulmonary nodule methylation dataset. Based on the Illumina 450K LUADDNAMethylation dataset in TCGA, methylation difference analysis was performed on the two datasets, and 376 tissue-specific methylation regions of malignant pulmonary nodules were screened out, which were denoted as the first screening DMRs. The first screening DMRs were further reverse-screened to obtain 201 second screening DMRs. The second screening DMRs were analyzed and filtered for consistency between tissue and paired plasma to obtain a candidate methylation biomarker set. The candidate methylation biomarker set was subjected to machine learning on the plasma training set and validation set to screen out methylation regions with significant differences, resulting in 78 methylation biomarkers for the differential diagnosis of benign and malignant pulmonary nodules.

[0040] The technical solution of the present invention will be described in detail below with reference to specific embodiments and accompanying drawings.

[0041] The materials and experimental methods involved in this application are as follows:

[0042] 1. Research Subjects

[0043] This study, conducted from February 2016 to August 2021, enrolled 347 participants from Shanghai Chest Hospital. Participant information is shown in Table 1. Among them, 296 patients presented with benign or malignant pulmonary nodules on CT scans (including 106 patients with benign pulmonary diseases and 190 lung cancer patients) and 51 healthy controls were included. Benign pulmonary diseases included pneumonia, chronic obstructive pulmonary disease, and tuberculosis. Lung cancer patients were confirmed by histopathology and / or cytopathology, and staging was based on the 8th edition of the TNM staging system. Healthy controls were outpatients undergoing physical examinations. Patients lacking histopathological diagnosis, acute illness history, or other malignant tumors were excluded. All participants signed informed consent forms. This study has completed clinical trial registration (ChiCTR2000036938). All participants in this study were randomly assigned to the training and validation sets. The enrolled lung cancer patients were primarily early-stage lung cancer patients (Tis, Stage I, and Stage II), accounting for 72.6% (138 / 190) of the participants.

[0044] Table 1. Relevant information of 296 patients with pulmonary nodules

[0045]

[0046] 2. Methylation sequencing and analysis

[0047] 2.1 Extraction of cfDNA from plasma samples

[0048] Participants underwent peripheral blood collection at a rate of 10 mL using EDTA anticoagulant tubes. After centrifugation to separate plasma (approximately 4 mL), cell-free DNA samples were extracted using a cell-free nucleic acid extraction reagent (APG-61001-050, Shanghai Yiming Biotechnology Co., Ltd.). The samples were then concentrated to a volume of 20 μL using a lyophilizer. Quantification was performed using Qubit, with concentrations ranging from 1 to 5 ng. 15 μL of this quantified sample was used for subsequent library construction.

[0049] 2.2 Construction of methylated libraries

[0050] The main reagents used in the methylation library preparation process of this application are shown in Table 2. Among them, Methylation-Lightning... TM Kit was purchased from the USA. The company can efficiently convert DNA to bisulfite. The linear amplification primer premix contains 201 custom-mixed primers for the specific linear amplification of 201 candidate methylation markers (the screening process for these 201 candidate methylation markers will be described later); OPERA The Linear Amplification Kit is specifically designed for linear amplification of DNA samples after bisulfite conversion; OPERA Universal library construction reagents (for This is a universal library preparation kit for matching single primers for amplification; Database creation uses single-chain tag connectors (Set 1 or 2, for...) (This refers to a labeled adapter kit used in the ligation reaction during library construction.) Pre-expansion labeled primers for library construction (Set 1 or 2, for...) () is a tag primer kit used in the pre-expansion reaction during the library construction process; Library quantification reagents (for () is a kit for quantifying molecules in a library that has been constructed.

[0051] Table 2. Main reagents used for methylation library construction.

[0052]

[0053]

[0054] The specific process is as follows:

[0055] 2.2.1 Bisulfite transformation of sample DNA

[0056] The commercially available EZ DNAMethylation-Lightning conversion kit was used. TM The Kit (ZYMO) is used to transform cfDNA samples. Please refer to the product instructions for specific operating procedures.

[0057] 2.2.2 Single-stranded linear amplification

[0058] The preparation method and reaction procedure for the single-stranded linear amplification system are as follows: OPERA Follow the instructions for the linear amplification kit, adding bisulfite conversion products, customized linear amplification primer premix, and linear amplification reagents.

[0059] For linear amplification program setup and installation methods, please refer to OPERA. Linear amplification reagents (for) Follow the instructions in the kit.

[0060] Purification of linear amplification products using magnetic beads; specific purification methods are described in the OPERA documentation. Universal library construction reagents (for Follow the kit instructions and use 20 μL of purified elution product for subsequent experiments.

[0061] 2.2.3 Single-chain linkage reaction

[0062] The purified product obtained in step 2.2.2 was ligated with the tagged adapter. The preparation of the ligation reaction system and the reaction procedure were performed according to OPERA. Universal library construction reagents (for ) Reagent kit instructions and Database creation uses a single-chain tag connector Set 1 or 2 (for The process was carried out according to the instructions to obtain the joining product.

[0063] 2.2.4 Pre-expansion reaction

[0064] The ligation product obtained in step 2.2.3 was subjected to a pre-expansion reaction. The preparation of the pre-expansion reaction system and the reaction procedure were performed according to OPERA. Universal library construction reagents (for ) Reagent kit instructions and Pre-expansion labeled primers Set 1 or 2 (for) are used for library construction. Follow the kit instructions to obtain pre-expansion products, which are library molecules containing tag primer sequences.

[0065] 2.2.5 Expanding the warehouse

[0066] The pre-expansion product obtained in step 2.2.5 was subjected to an expansion reaction. The preparation of the expansion reaction system and the reaction procedure were carried out in accordance with OPERA. Universal library construction reagents (for The amplified library products were obtained by following the kit instructions. PCR quality control and quantification methods for the library products were performed according to the OPERA library quantification kit (for...). Follow the instructions. After quantification, mix the libraries as needed and then perform sequencing.

[0067] 2.3 Bioinformatics Analysis of Library Sequencing and Implantation Data

[0068] The library was sequenced at 150 bp paired ends using Illumina's NovaSeq 6000 platform. After quality control using FastP software, the data was analyzed using SYMPHO BcDNA Methylation Analysis (PR) software (APG_81002, version v0.2) developed by Shanghai Yiming Biotechnology. For FastQ data, primer information was extracted using the cut adapter, followed by bismark alignment to obtain BAM files. After deduplication using UMIcollapse, all primer results were summarized to calculate the deduplicated CpG depth and haplomethylation level (MHC). Finally, the normalized number of haplomethylated molecules (nMHC) in the sample was calculated for subsequent analysis.

[0069] Regarding the aforementioned methylation sequencing and analysis process, the applicant wishes to clarify that, as is known in the art, when transforming cfDNA in peripheral blood samples, the purpose of the transformation can be achieved using bisulfite, bisulfite, bisulfite, or other similar reagents. This involves converting unmethylated cytosine in cell-free DNA from peripheral blood into unmethylated thymine, thus obtaining a transformed sample. Therefore, the use of any of these reagents for transformation is within the scope of this invention. Furthermore, the reagents used can be commercially available products or prepared in-house.

[0070] When constructing a library, sequencing, and obtaining methylation results for each methylation region of peripheral blood cfDNA in the test sample, techniques commonly used in the field can also be employed, such as targeted methylation sequencing based on hybridization capture, methylation sequencing based on multiplex PCR, or methylation detection based on quantitative real-time PCR, and are not limited to the methods described in this application.

[0071] Example 1: Screening of candidate methylation biomarkers for benign and malignant lung nodules

[0072] 1. Initial screening of specific methylation markers for benign and malignant lung nodules

[0073] A lung nodule methylation dataset was constructed by performing genome-wide degenerate methylation differential analysis (RRBS) on 50 pairs of clinical pulmonary nodule lesions and adjacent tissue samples (40 pairs malignant and 10 pairs benign). Simultaneously, based on the Illumina 450K LUAD DNAMethylation dataset from TCGA, the two datasets were analyzed according to the following criteria: (a) the mean methylation value of the candidate biomarker in control samples (including adjacent tissues of malignant lesions, benign lesions, and adjacent tissues of benign lesions) was less than 0.02; (b) the ratio of the mean methylation value of the candidate biomarker in positive samples (malignant lesions) to that in control samples was greater than 3.0; (c) the mean methylation value of the candidate biomarker in positive samples (malignant lesions) was greater than 0.05; and (d) the DMR region of the candidate biomarker was greater than 20 bp. 376 methylation regions (DMRs) showing significant differences between malignant nodule tissue (lung cancer) and adjacent / benign nodule tissue were identified as the first-stage DMRs.

[0074] It should be noted that the DMR average methylation numerical analysis adopts the standard method in the field of methylation analysis.

[0075] 2. Reverse screening of methylation markers

[0076] Based on the tissue-specific methylation database in published literature (Moss, J., et al. Nat Commun, 2018), methylation regions derived from blood cells and other organs were reverse-selected and removed from the aforementioned 376 DMRs. DMRs with a target detection region of less than 4 CpG were also removed. Then, using whole blood samples from healthy controls, single-primer panels were designed for the obtained DMRs, and library construction and sequencing were performed according to the methylation sequencing method described above. DMRs with an average methylation value greater than 0.03 in blood cells were removed. Finally, 201 candidate methylation markers for malignant lung nodules with significant differences were obtained (hereinafter referred to as P201), i.e., the second-stage screening DMRs.

[0077] 3. Consistency verification of P201 in tissue and plasma samples from patients with malignant pulmonary nodules.

[0078] Thirty patients with malignant pulmonary nodules (i.e., lung cancer patients) were randomly selected for the study. Paired tumor tissue DNA, adjacent normal tissue DNA, paired plasma cfDNA, and blood cell DNA were extracted and methylated for library construction and sequencing according to the aforementioned methylation sequencing method. The results are as follows: Figure 1As shown, the methylation level of P201 in lung cancer patient tissue and plasma samples was significantly higher than that in adjacent normal tissue and blood cells, and P201 showed a high degree of consistency between tumor tissue and plasma samples. After this step, P201 becomes the candidate methylation biomarker set.

[0079] Example 2: Screening for methylation biomarkers in benign and malignant lung nodules using different machine learning methods

[0080] All subjects were randomly divided into training and validation sets. Using the method described above, cfDNA was extracted from plasma samples of all subjects to construct a single-primer amplification methylation library targeting P201. Three different screening strategies were employed to obtain significantly different methylation characteristics of lung cancer compared to benign and healthy controls in the training set. Three different models for differentiating benign and malignant lung nodules were then constructed, and the performance of these models was evaluated in the validation set. Details are as follows:

[0081] 1. Determination, modeling, and performance of the P30 methylation marker set

[0082] Eighty percent of the samples from all subjects were used as the training set, and the remaining 20% ​​as the validation set. LASSO analysis (α = 0.03) was used to screen the following 30 key methylation markers (denoted as P30) in the training set:

[0083] chr7: 143059947-143060107; chr5: 178487412-178487572; chr11: 15136199-15136359; chr17: 43972911- 43973071; chr12: 64062951-64063111; chr3: 87039622-87039782; chr6: 27835283-27835443; chr5: 14625 7858-146258018; chr16: 28075032-28075192; chr17: 47307472-47307632; chr19: 37407294-37407454; ch r1: 179545118-179545278; chr13: 20806316-20806476; chr7: 132261297-132261457; chr4: 17783234-1778 3394; chr1: 67773558-67773718; chr10: 105037463-105037623; chr19: 7735166-7735326; chr6: 26199948 -26200108; chr6: 26273357-26273517; chr12: 4381931-4382091; chr7: 30722046-30722206; chr5: 1757924 94-175792654; chr11: 69590360-69590520; chr8: 143858458-143858618; chr14: 59104962-59105122; chr 4: 57976315-57976475; chr4: 39529282-39529442; chr7: 55259381-55259541; chr10: 90343107-90343267.

[0084] Furthermore, a Gaussian process was used to construct a P30-based differential diagnosis model for benign and malignant pulmonary nodules, and the performance of this model in differentiating benign and malignant pulmonary nodules was evaluated on the training set and validation set.

[0085] The results are as follows Figure 2 As shown in Figure A, the model built using Gaussian processes on P30 exhibits excellent performance in differentiating between benign and malignant pulmonary nodules. As shown in Table 3, in the training set, the AUC reached 0.943 (95% CI 0.918–0.967), with a sensitivity of 86.9% and a specificity of 83.9%; in the validation set, the AUC reached 0.910 (95% CI 0.845–0.974), with a sensitivity of 79.5% and a specificity of 87.1%.

[0086] 2. Determination, modeling, and performance of the P45 methylation marker set

[0087] Eighty percent of the samples from all subjects were used as the training set, and the remaining 20% ​​as the validation set. LASSO analysis (α = 0.02) was used to screen the following 45 key methylation markers (denoted as P45) in the training set:

[0088] chr7:143059947-143060107;chr5:178487412-178487572;chr11:15136199-15136359;chr17:43972911-43973071;chr8:104383554-104383714;chr12:64062951-64063111;chr3:87039622-87039782;chr6:27835283-27835443;chr17:77020105-77020265;chr4:54966998-54967158;chr7:98467849-98468009;chr16:28075032-28075192;chr4:48485804-48485964;chr19:37407294-37407454;chr1:179545118-179545278;chr18:43652068-43652228;chr13:20806316-20806476;chr8:116660588-116660748;chr7:132261297-132261457;chr19:53636048-53636208;chr5:33936140-33936300;chr4:17783234-17783394;chr1:67773558-67773718;chr10:105037463-105037623;chr12:52400797-52400957;chr5:374080-374240;chr19:7735166-7735326;chr6:26199948-26200108;chr6:26273357-26273517;chr12:4381931-4382091;chr10:135050089-135050249;chr7:30722046-30722206;chr5:175792494-175792654;chr11:69590360-69590520;chr17:47074677-47074837;chr8:143858458-143858618;chr14:59104962-59105122;chr4:57976315-57976475;chr8:53852244-53852404;chr3:50242683-50242843;chr4:39529282-39529442;chr7:55259381-55259541;chr5: 112073416-112073576; chr10: 90343107-90343267; chr2: 43451673-43451833. ;

[0089] Furthermore, a P45-based differential diagnosis model for benign and malignant pulmonary nodules was constructed using support vector machines, and the performance of this model in differentiating benign and malignant pulmonary nodules was evaluated on the training set and validation set.

[0090] The results are as follows Figure 2 As shown in Figure B, the model built using support vector machines (SVM) on P45 exhibits excellent performance in differentiating between benign and malignant pulmonary nodules. As shown in Table 3, in the training set, the AUC reached 0.969 (95% CI 0.948–0.990), with a sensitivity of 89.5% and a specificity of 93.5%; in the validation set, the AUC reached 0.913 (95% CI 0.848–0.978), with a sensitivity of 69.2% and a specificity of 96.8%.

[0091] 3. Determination, modeling, and performance of the P42 methylation marker set

[0092] 45% of the samples from all subjects were used as the training set, and the remaining 55% as the validation set. The Xgboost method was then used to screen the training set for the following 42 key methylation markers (denoted as P42):

[0093] chr18:49867019-49867179;chr11:15136199-15136359;chr14:74892573-74892733;chr22:44420459-44420619;chr20:62283562-62283722;chr4:54966998-54967158;chr12:54427101-54427261;chr2:182321922-182322082;chr13:28674645-28674805;chr17:45810438-45810598;chr6:29760212-29760372;chr1:179545118-179545278;chr5:128797252-128797412;chr13:20806316-20806476;chr8:116660588-116660748;chr5:153784739-153784899;chr19:56904958-56905118;chr1:54204130-54204290;chr10:105037463-105037623;chr19:2290434-2290594;chr12:52400797-52400957;chr5:374080-374240;chr19:7735166-7735326;chr7:134143579-134143739;chr6:26199948-26200108;chr6:26273357-26273517;chr14:52781261-52781421;chr6:26204595-26204755;chr19:58951468-58951628;chr5:157098320-157098480;chr14:77228082-77228242;chr6:27100719-27100879;chr6:26189078-26189238;chr4:110224117-110224277;chr15:83316243-83316403;chr6:29716362-29716522;chr14:52734595-52734755;chr11:124735024-124735184;chr20:45338378-45338538;chr3:50242683-50242843;chr7:151107115-151107275;chr1:29586353-29586513。

[0094] Furthermore, logistic regression was used to construct a P42-based differential diagnostic model for benign and malignant pulmonary nodules, and the performance of this model in differentiating benign and malignant pulmonary nodules was evaluated on the training set and validation set.

[0095] The results are as follows Figure 2 As shown in Figure C, the model established by logistic regression in P42 exhibits excellent performance in differentiating between benign and malignant pulmonary nodules. As shown in Table 3, in the training set, the AUC reached 0.936 (95% CI 0.897–0.976), with a sensitivity of 80% and a specificity of 98%; in the validation set, the AUC reached 0.888 (95% CI 0.838–0.987), with a sensitivity of 73% and a specificity of 99%.

[0096] Table 3 shows the performance of the methylation marker models established on P30, P45, and P42 on the training and validation sets.

[0097]

[0098] A total of 78 methylation biomarkers with statistically significant differences were obtained by screening using different strategies, among which 39 methylation biomarkers showed good independent discrimination performance. See Table 4 for details.

[0099] Table 4. 78 methylation biomarkers, their specific locations, and independent diagnostic performance.

[0100]

[0101]

[0102] In summary, this invention discloses a set of methylation biomarkers for differentiating benign from malignant pulmonary nodules, along with their screening methods and applications. The screening method yielded 78 methylation biomarkers for differentiating benign from malignant pulmonary nodules and / or lung cancer. Furthermore, the diagnostic performance of the methylation biomarkers provided by this invention was evaluated through model construction. The methylation biomarkers or combinations thereof provided by this invention exhibit high sensitivity and specificity in differentiating benign from malignant pulmonary nodules and / or lung cancer, making them suitable for widespread application in the differential diagnosis of benign from malignant pulmonary nodules.

[0103] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. Use of a reagent for detecting the methylation level of a methylation marker in the manufacture of a product for detecting lung nodule malignancy and / or lung cancer, characterized in that, The methylation markers include the following 30 methylation regions: chr7: 143059947-143060107; chr5: 178487412-178487572; chr11: 15136199-15136359; chr17: 43972911-43973071; chr12: 64062951-64063111; chr3: 87039622-87039782; chr6: 27835283-27835443; chr5: 146257858-146258018; chr16: 28075032-28075192; chr17: 47307472-47307632; chr19: 37407294-37407454; chr1: 179545118-179545278; chr13: 20806316-20806476; chr7: 132261297-132261457; chr4: 17783234-17783394; chr1: 67773558-67773718; chr10: 105037463-105037623; chr19: 7735166-7735326; chr6: 26199948-26200108; chr6: 26273357-26273517; chr12: 4381931-4382091; chr7: 30722046-30722206; chr5: 175792494-175792654; chr11: 69590360-69590520; chr8: 143858458-143858618; chr14: 59104962-59105122; chr4: 57976315-57976475; chr4: 39529282-39529442; chr7: 55259381-55259541; chr10: 90343107-90343267; wherein the methylation regions of the methylation markers are determined based on human genome hg19 alignment.

2. A kit for detecting lung nodule malignancy and / or lung cancer, characterized in that, The kit includes reagents for detecting the methylation level of the methylation marker, and the methylation marker includes the following 30 methylation regions: chr7: 143059947-143060107; chr5: 178487412-178487572; chr11: 15136199-15136359; chr17: 43972911-43973071; chr12: 64062951-64063111; chr3: 87039622-87039782; chr6: 27835283-27835443; chr5: 146257858-146258018; chr16: 28075032-28075192; chr17: 47307472-47307632; chr19: 37407294-37407454; chr1: 179545118-179545278; chr13: 20806316-20806476; chr7: 132261297-132261457; chr4: 17783234-17783394; chr1: 67773558-67773718; chr10: 105037463-105037623; chr19: 7735166-7735326; chr6: 26199948-26200108; chr6: 26273357-26273517; chr12: 4381931-4382091; chr7: 30722046-30722206; chr5: 175792494-175792654; chr11: 69590360-69590520; chr8: 143858458-143858618; chr14: 59104962-59105122; chr4: 57976315-57976475; chr4: 39529282-39529442; chr7: 55259381-55259541; chr10: 90343107-90343267; wherein the methylation region of the methylation marker is determined based on human genome hg19 alignment.

3. A screening method of methylation markers for differential diagnosis of lung nodule benignity and malignancy, characterized in that, comprising the following steps: S1, constructing a local lung nodule methylation dataset by performing RRBS analysis on lung nodules and perinodular tissues, and performing methylation difference analysis on the two datasets according to the lung adenocarcinoma illumina 450k methylation chip dataset in TCGA, and screening to obtain a methylation region specific to malignant lung nodule tissue, denoted as the first screening DMR; S2, performing reverse screening on the first screening DMR obtained in step S1 to remove methylation regions meeting the following criteria to obtain the second screening DMR: (1) methylation regions from blood cells and other organs; (2) the target detection region is less than 4 CpGs; (3) methylation regions with an average methylation value greater than 0.03 in blood cells; S3, based on the second screening DMR obtained in step S2, obtaining a candidate methylation marker set by analyzing and filtering the consistency between tissues and paired plasma; S4, performing machine learning on the candidate methylation marker set obtained in step S3 in the plasma training set and test set, screening methylation regions with significant differences, and obtaining methylation markers for differential diagnosis of lung nodules.

4. The screening method according to claim 3, wherein In step S1, the first screening DMR meets the following criteria in the methylation difference analysis of the two datasets: (a) the average methylation value of the control sample is less than 0.02; (b) the ratio of the average methylation value of the positive sample to that of its control sample is greater than 3.0; (c) the average methylation value of the positive sample is greater than 0.05; (d) the DMR region is greater than 20 bp; wherein the positive sample is a malignant nodule tissue; and the control sample is a tissue adjacent to the malignant nodule, a benign nodule tissue and a tissue adjacent to the benign nodule.

Images