System for identifying benign thyroid tumor and malignant thyroid tumor based on combined gene

Abstract

The invention discloses a system for identifying benign and malignant thyroid tumors based on combined genes. The invention provides a computer device, which comprises a memory, a processor and a computer program stored on the memory, and the processor executes the computer program to realize the following steps: collecting methylation level data of DNA methylation marker combinations (selected from LZTFL1, BCAR3 and LPR4) of two types of samples A and B, establishing a model, and determining a threshold value; inputting methylation level data of a DNA methylation marker combination of a to-be-detected person, substituting the methylation level data into the model to obtain a detection index, and comparing the detection index with a threshold to determine whether a to-be-detected sample is a type A or a type B; the A type sample and the B type sample are thyroid benign tumor and malignant tumor respectively. Compared with the low methylation phenomenon of LZTFL1, BCAR3 and LPR4 genes in benign tumor and thyroid malignant tumor tissues, the invention has important clinical application value for improving the early diagnosis and treatment effect of thyroid cancer.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS:

[0002] This application claims priority to the Chinese patent application (application number 202311178697.7) filed on September 13, 2023, and the entire contents of that patent application are incorporated herein by reference. Technical Field

[0003] The present invention relates to the field of medical informatics, and in particular to a system for distinguishing benign and malignant thyroid tumors based on combined genes. Background Art

[0004] Thyroid cancer is the most common malignant tumor of the endocrine system, including papillary thyroid cancer, follicular thyroid cancer, anaplastic thyroid cancer and medullary cancer. Papillary thyroid cancer (PTC) is the most common, accounting for more than 90% of all thyroid malignancies [Xing, Mingzhao; Haugen, BryanR; Schlumberger, Martin (2013). Progress in molecular-based management of differentiated thyroid cancer. The Lancet, 381 (9871), 1058–1069.]. According to statistics, the prevalence of thyroid nodules in adults is about 5-10%, and the prevalence is most serious in people over 60 years old, up to 50-70% [Guth S, Theune U, Aberle J, et al. Very high prevalence of thyroid nodules detected by high frequency (13MHz) ultrasound examination. Eur J Clin Invest 2009; 39: 699-706.]. Imaging examination is a common method for thyroid diagnosis. This method mostly relies on the doctor's experience and judgment, and there are certain errors in the results. In addition, imaging has certain radiation damage to the human body. Fine needle aspiration biopsy is also a common clinical diagnosis technology for thyroid cancer. This method evaluates the benign or malignant nature of the nodule based on the cytological morphology of the puncture. Since the cytological features of benign and malignant thyroid tumors often overlap, about 10-30% of fine needle aspiration diagnoses are unclear cytological results [Cibas ES, Ali SZ. The 2017 Bethesda System for Reporting Thyroid Cytopathology. Thyroid. 2017; 27(11): 1341-6.]. Uncertain puncture results lead to overtreatment or missed diagnosis in approximately 60% of patients [Stewart R, Leang YJ, Bhatt CR, Grodski S, Serpell J, Lee JC. Quantifying the differences in surgical management of patients with definitive and indeterminate thyroid nodule cytology. Eur J Surg Oncol. 2020; 46(2): 252-7.].This not only increases the economic and physical and mental burden on patients, but also occupies a large amount of public health resources, resulting in huge financial costs for the healthcare system. Therefore, differentiating between benign and malignant thyroid tumors is helpful for clinicians to adopt more precise treatment plans, which has important clinical and public health significance.

[0005] Epigenetics is a heritable gene expression regulation method that does not involve changes in DNA sequence and can be passed on to the next generation [Nicoglou A, Merlin F. Epigenetics: A way to bridge the gap between biological fields. Stud Hist Philos Biol Biomed Sci. 2017; 66: 73-82]. DNA methylation is one of the important ways of epigenetic regulation, which refers to the covalent bond of a methyl group to the 5' carbon position of cytosine in genomic CpG dinucleotides under the action of DNA methyltransferase [Bird A. Perceptions of epigenetics. Nature. 2007; 447: 396-398]. A large number of studies have shown that DNA methylation can cause changes in chromatin structure, DNA conformation, DNA stability and the way DNA interacts with proteins, thereby controlling gene expression [Moore LD, LeT, Fan G. DNA methylation and its basic function. Neuropsychopharmacology. 2013; 38: 23-38].

[0006] DNA methylation markers are currently the best in vitro early diagnosis molecular markers for tumors. At present, the sensitivity and specificity of clinical diagnostic markers for thyroid cancer are very limited, especially the lack of early diagnosis markers. Therefore, more sensitive and specific early molecular markers are urgently needed to be discovered. Summary of the invention

[0007] The purpose of the present invention is to provide a system for distinguishing benign and malignant thyroid tumors based on combined genes.

[0008] In a first aspect, the present invention claims a computer device.

[0009] The computer device claimed in the present invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the following steps:

[0010] Collecting methylation level data of each gene in a DNA methylation marker combination of n1 type A samples and n2 type B samples; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene, and the LPR4 gene or any two of them;

[0011] The methylation level data of each gene in the DNA methylation marker combination of the n1 type A samples and the n2 type B samples are used to establish a mathematical model through a binary classification logistic regression method according to the classification method of type A and type B, and determine the threshold value of classification judgment;

[0012] Inputting the methylation level data of each gene in the DNA methylation marker combination of the subject to be tested;

[0013] Substituting the methylation level data of each gene in the DNA methylation marker combination of the subject into the mathematical model to calculate the detection index;

[0014] Comparing the detection index with the threshold to obtain a comparison result;

[0015] Outputting a conclusion of whether the type of the sample to be tested is type A or type B according to the comparison result;

[0016] The type A sample and the type B sample are any of the following:

[0017] (F1) Benign thyroid tumors and malignant thyroid tumors;

[0018] (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors;

[0019] (E3) Benign thyroid tumors and different stages of malignant thyroid tumors;

[0020] (F4) Different subtypes of thyroid malignancies;

[0021] (F5) Thyroid malignancies at different stages.

[0022] In a second aspect, the present invention claims a computer program product.

[0023] The computer program product claimed in the present invention comprises a computer program, which implements the steps described in the first aspect above when executed by a processor.

[0024] The computer program product should be understood as a software product that implements its solution mainly through a computer program.

[0025] In a third aspect, the present invention claims a computer-readable storage medium.

[0026] The computer-readable storage medium claimed for protection by the present invention stores a computer program thereon, and when the computer program is executed by a processor, the steps described in the first aspect above are implemented.

[0027] Among them, the computer-readable storage medium refers to a carrier for storing data, which may be a tape, disk, floppy disk, CD, magneto-optical disk, ROM, PROM, VCD, DVD, hard disk, flash memory, U disk, CF card, SD card, MMC card, SM card, memory stick (MemoryStick) or xD card, etc.

[0028] In a fourth aspect, the present invention claims a system.

[0029] The system claimed in the present invention may include:

[0030] (G1) a reagent and / or instrument (such as a time-of-flight mass spectrometer) for detecting the methylation level of each gene in a DNA methylation marker combination; the DNA methylation marker combination is composed of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of the genes therein;

[0031] (G2) a device comprising unit X and unit Y;

[0032] The unit X is used to establish a mathematical model, including a data acquisition module, a data analysis and processing module and a model output module;

[0033] The data collection module is configured to collect (G1) the methylation level data of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples detected;

[0034] The data analysis and processing module is configured to receive the methylation level data of each gene in the DNA methylation marker combination of the n1 type A samples and the n2 type B samples from the data acquisition module, establish a mathematical model through a binary classification logistic regression method according to the classification method of type A and type B, and determine the threshold value of classification judgment;

[0035] Wherein, n1 and n2 can both be positive integers greater than 50 (eg, greater than 100).

[0036] In a specific implementation of the present invention, the threshold is set to 0.5. Values ​​greater than 0.5 are classified into one category, values ​​less than 0.5 are classified into another category, and values ​​equal to 0.5 are classified into an uncertain gray area. Type A and Type B are two corresponding categories. The grouping of the two categories, which group is type A and which group is type B, should be determined according to a specific mathematical model without any agreement.

[0037] In practical applications, the threshold value can also be determined according to the maximum Youden index (specifically, it can be the value corresponding to the maximum Youden index). Those greater than the threshold value are classified into one category, those less than the threshold value are classified into another category, and those equal to the threshold value are regarded as an uncertain gray area. Type A and Type B are two corresponding categories. The grouping of the two categories, which group is type A and which group is type B, should be determined according to a specific mathematical model without any agreement.

[0038] The model output module is configured to receive the mathematical model established by the data analysis and processing module and output it;

[0039] The unit Y is used to determine the type of sample to be tested, and includes a data input module, a data calculation module, a data comparison module and a conclusion output module;

[0040] The data input module is configured to input (G1) data of the methylation level of each gene in the DNA methylation marker combination of the subject detected;

[0041] The data operation module is configured to receive the data of the methylation level of each gene in the DNA methylation marker combination of the subject to be tested from the data input module, and substitute the data of the methylation level of each gene in the DNA methylation marker combination of the subject to be tested into the mathematical model established by the data analysis and processing module in the unit X to calculate the detection index;

[0042] The data comparison module is configured to receive the detection index calculated by the data operation module and compare the detection index with the threshold value determined in the data analysis and processing module in the unit X;

[0043] The conclusion output module is configured to receive the comparison result from the data comparison module, and output a conclusion of whether the type of the sample to be tested is type A or type B according to the comparison result;

[0044] The type A sample and the type B sample are any of the following:

[0045] (F1) Benign thyroid tumors and malignant thyroid tumors;

[0046] (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors;

[0047] (F3) Benign thyroid tumors and different stages of malignant thyroid tumors;

[0048] (F4) Different subtypes of thyroid malignancies;

[0049] (F5) Thyroid malignancies at different stages.

[0050] Furthermore, the different subtypes described in (F2) and (F4) may be pathological subtypes, such as histological subtypes. The same shall apply below.

[0051] Furthermore, the different stages described in (F3) and (F5) may be clinical stages. The same shall apply below.

[0052] In a specific embodiment of the present invention, the benign thyroid tumor and different subtypes of malignant thyroid tumors in (F2) can be any of the following: benign thyroid tumor and papillary thyroid cancer, benign thyroid tumor and follicular thyroid cancer, benign thyroid tumor and medullary thyroid cancer, benign thyroid tumor and anaplastic thyroid cancer. The same applies below.

[0053] In a specific embodiment of the present invention, the benign thyroid tumor and the thyroid malignant tumors of different stages in (F3) can be any of the following: benign thyroid tumor and stage I thyroid malignant tumor, benign thyroid tumor and stage II thyroid malignant tumor, benign thyroid tumor and stage III thyroid malignant tumor, benign thyroid tumor and stage IV thyroid malignant tumor. The same below.

[0054] In a specific embodiment of the present invention, the different subtypes of thyroid malignant tumors described in (F4) can be any of the following: papillary thyroid carcinoma and follicular thyroid carcinoma, papillary thyroid carcinoma and medullary thyroid carcinoma, papillary thyroid carcinoma and anaplastic thyroid carcinoma, follicular thyroid carcinoma and medullary thyroid carcinoma, follicular thyroid carcinoma and anaplastic thyroid carcinoma, medullary thyroid carcinoma and anaplastic thyroid carcinoma. The same applies below.

[0055] In a specific embodiment of the present invention, the different stages of thyroid malignant tumors in (F5) can be any of the following: stage I thyroid malignant tumors and stage II thyroid malignant tumors, stage I thyroid malignant tumors and stage III thyroid malignant tumors, stage I thyroid malignant tumors and stage IV thyroid malignant tumors, stage II thyroid malignant tumors and stage III thyroid malignant tumors, stage II thyroid malignant tumors and stage IV thyroid malignant tumors, stage III thyroid malignant tumors and stage IV thyroid malignant tumors. The same below.

[0056] In a fifth aspect, the present invention claims the use of a substance for detecting the methylation level of each gene in a DNA methylation marker combination in the preparation of a product; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of the genes therein; the use of the product is at least one of the following:

[0057] (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors;

[0058] (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors;

[0059] (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages;

[0060] (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies;

[0061] (5) Differentiate or assist in differentiating the different stages of thyroid malignant tumors.

[0062] Furthermore, the different subtypes described in (2) and (4) may be pathological subtypes, such as histological subtypes. The same shall apply below.

[0063] Furthermore, the different stages described in (3) and (5) may be clinical stages. The same shall apply hereinafter.

[0064] In a specific embodiment of the present invention, the differentiation or auxiliary differentiation of benign thyroid tumors and different subtypes of malignant thyroid tumors in (2) can be any of the following: differentiation or auxiliary differentiation of benign thyroid tumors and papillary thyroid carcinoma, differentiation or auxiliary differentiation of benign thyroid tumors and follicular thyroid carcinoma, differentiation or auxiliary differentiation of benign thyroid tumors and medullary thyroid carcinoma, differentiation or auxiliary differentiation of benign thyroid tumors and anaplastic thyroid carcinoma. The same below.

[0065] In a specific embodiment of the present invention, the differentiation or auxiliary differentiation of benign thyroid tumors and thyroid malignant tumors of different stages in (3) can be any of the following: differentiation or auxiliary differentiation of benign thyroid tumors and stage I thyroid malignant tumors, differentiation or auxiliary differentiation of benign thyroid tumors and stage II thyroid malignant tumors, differentiation or auxiliary differentiation of benign thyroid tumors and stage III thyroid malignant tumors, differentiation or auxiliary differentiation of benign thyroid tumors and stage IV thyroid malignant tumors. The same below.

[0066] In a specific embodiment of the present invention, the differentiation or auxiliary differentiation of different subtypes of thyroid malignant tumors in (4) can be any of the following: differentiation or auxiliary differentiation of papillary thyroid carcinoma and follicular thyroid carcinoma, differentiation or auxiliary differentiation of papillary thyroid carcinoma and medullary thyroid carcinoma, differentiation or auxiliary differentiation of papillary thyroid carcinoma and anaplastic thyroid carcinoma, differentiation or auxiliary differentiation of follicular thyroid carcinoma and medullary thyroid carcinoma, differentiation or auxiliary differentiation of follicular thyroid carcinoma and anaplastic thyroid carcinoma, differentiation or auxiliary differentiation of medullary thyroid carcinoma and anaplastic thyroid carcinoma. The same below.

[0067] In a specific embodiment of the present invention, the differentiation or auxiliary differentiation of different stages of thyroid malignant tumors in (5) can be any of the following: differentiation or auxiliary differentiation of stage I thyroid malignant tumors and stage II thyroid malignant tumors, differentiation or auxiliary differentiation of stage I thyroid malignant tumors and stage III thyroid malignant tumors, differentiation or auxiliary differentiation of stage I thyroid malignant tumors and stage IV thyroid malignant tumors, differentiation or auxiliary differentiation of stage II thyroid malignant tumors and stage III thyroid malignant tumors, differentiation or auxiliary differentiation of stage II thyroid malignant tumors and stage IV thyroid malignant tumors, differentiation or auxiliary differentiation of stage III thyroid malignant tumors and stage IV thyroid malignant tumors. The same below.

[0068] In a sixth aspect, the present invention claims the use of a substance for detecting the methylation level of each gene in a DNA methylation marker combination and a medium storing a mathematical model and / or a method for using the mathematical model in preparing a product; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of the genes therein; the use of the product is at least one of the following:

[0069] (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors;

[0070] (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors;

[0071] (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages;

[0072] (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies;

[0073] (5) Differentiate or assist in differentiating the different stages of thyroid malignancies;

[0074] The mathematical model is obtained according to a method comprising the following steps:

[0075] (D1) detecting the methylation level of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples respectively;

[0076] (D2) taking the methylation level data of each gene in the DNA methylation marker combination of all samples obtained in step (D1), establishing a mathematical model by binary logistic regression method according to the classification method of type A and type B, and determining the threshold value of classification judgment;

[0077] The method for using the mathematical model comprises the following steps:

[0078] (E1) detecting the methylation level of each gene in the DNA methylation marker combination of the sample to be tested;

[0079] (E2) substituting the methylation level data of each gene in the DNA methylation marker combination of the sample to be tested obtained in step (E1) into the mathematical model to obtain a detection index; then comparing the detection index with a threshold value, and determining whether the type of the sample to be tested is type A or type B according to the comparison result;

[0080] The type A sample and the type B sample are any of the following:

[0081] (F1) Benign thyroid tumors and malignant thyroid tumors;

[0082] (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors;

[0083] (E3) Benign thyroid tumors and different stages of malignant thyroid tumors;

[0084] (F4) Different subtypes of thyroid malignancies;

[0085] (F5) Thyroid malignancies at different stages.

[0086] In a seventh aspect, the present invention claims protection for the use of the medium storing the mathematical model and / or the method for using the mathematical model as described in the sixth aspect above in preparing a product; the use of the product is at least one of the following:

[0087] (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors;

[0088] (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors;

[0089] (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages;

[0090] (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies;

[0091] (5) Differentiate or assist in differentiating the different stages of thyroid malignant tumors.

[0092] In the above-mentioned related aspects, the methylation level of each gene in the DNA methylation marker combination is composed of the following three items (A1)-(A3) or any two of them:

[0093] (A1) methylation levels of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene;

[0094] (A2) methylation levels of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene;

[0095] (A3) Methylation levels of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene.

[0096] Furthermore, the methylation level of each gene in the DNA methylation marker combination may be the methylation level of any of the following CpG sites:

[0097] (B1) all CpG sites on the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene;

[0098] (B2) all CpG sites on the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene;

[0099] (B3) all CpG sites on the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene;

[0100] (B4) All CpG sites on the DNA fragment shown by SEQ ID No. 1 in the LZTFL1 gene, all CpG sites on the DNA fragment shown by SEQ ID No. 2 in the BCAR3 gene, and all CpG sites on the DNA fragment shown by SEQ ID No. 3 in the LPR4 gene.

[0101] In a specific embodiment of the present invention, some adjacent methylation sites are indistinguishable in the peak graph when DNA methylation analysis is performed using time-of-flight mass spectrometry because several CpG sites are located on one methylation fragment (the indistinguishable sites are recorded in Table 5). Therefore, when performing methylation level analysis and constructing and using relevant mathematical models, they are treated as one methylation site.

[0102] In the above-mentioned related aspects, the substance / reagent for detecting the methylation level of each gene in the DNA methylation marker combination comprises (or is) the following three items (C1)-(C3) or any two thereof:

[0103] (C1) Primer pairs used to amplify the full length or partial fragment of the LZTFL1 gene;

[0104] (C2) primer pairs for amplifying the full length or partial fragment of the BCAR3 gene;

[0105] (C3) Primer pairs used to amplify the full-length or partial fragment of the LPR4 gene.

[0106] Furthermore, the partial fragment in (C1) may be a DNA fragment shown in SEQ ID No. 1 or a DNA fragment contained therein or a DNA fragment having more than 80% identity therewith.

[0107] Furthermore, the partial fragment in (C2) may be a DNA fragment shown in SEQ ID No. 2 or a DNA fragment contained therein or a DNA fragment having more than 80% identity therewith.

[0108] Furthermore, the partial fragment in (C3) may be a DNA fragment shown in SEQ ID No. 3 or a DNA fragment contained therein or a DNA fragment having more than 80% identity therewith.

[0109] Furthermore, the primer pair in (C1) is primer pair A, the primer pair in (C2) is primer pair B, and the primer pair in (C3) is primer pair C.

[0110] The primer pair A is a primer pair consisting of primer A1 and primer A2; the primer A1 is a single-stranded DNA represented by nucleotides 11 to 35 of SEQ ID No.4; the primer A2 is a single-stranded DNA represented by nucleotides 32 to 56 of SEQ ID No.5 or SEQ ID No.5;

[0111] The primer pair B is a primer pair consisting of primer B1 and primer B2; the primer B1 is a single-stranded DNA represented by nucleotides 11 to 35 of SEQ ID No.6 or SEQ ID No.6; the primer B2 is a single-stranded DNA represented by nucleotides 32 to 56 of SEQ ID No.7 or SEQ ID No.7;

[0112] The primer pair C is a primer pair consisting of primer C1 and primer C2; the primer C1 is a single-stranded DNA represented by SEQ ID No.8 or nucleotides 11-35 of SEQ ID No.8; the primer C2 is a single-stranded DNA represented by SEQ ID No.9 or nucleotides 32-56 of SEQ ID No.9.

[0113] Any of the above mathematical models may vary in practical applications depending on the detection method and fitting method of DNA methylation. It should be determined based on the specific mathematical model without the need for agreement.

[0114] In an embodiment of the present invention, the model is specifically lg(y / (1-y))=b0+b1x1+b2x2+b3x3+…+bnXn, wherein y is the dependent variable, i.e., the detection index obtained by substituting the methylation value of one or more methylation sites of the test sample into the model, b0 is a constant, x1-xn is an independent variable, i.e., the methylation value of one or more methylation sites of the test sample (each value is a value between 0-1), and b1-bn is the weight assigned to the methylation value of each site by the model.

[0115] A specific model established in an embodiment of the present invention is a model for distinguishing or assisting in distinguishing benign thyroid tumors from malignant thyroid tumors, and the model is specifically: lg(y / (1-y))=2.710+0.539*LZTFL1_A_1-4.362*LZTFL1_A_2+3.115*LZTFL1_A_3+0.901*LZTFL1_A_4–5.183*BCAR3_B_1-0.718*BCAR3_B_2.3.4.5+2.609*BCAR3_B_6+1.402*LPR4_C_1-0.113*LPR4_C_2+4.091*LPR4_C_3-0.692*LPR4_C_4+1.341*LPR4_C_5+0.805*LPR4_C_6.The LZTFL1_A_1 is the methylation level of the CpG site shown at positions 46-47 from the 5' end of the DNA fragment shown in SEQ ID No.1; the LZTFL1_A_2 is the methylation level of the CpG site shown at positions 75-76 from the 5' end of the DNA fragment shown in SEQ ID No.1; the LZTFL1_A_3 is the methylation level of the CpG site shown at positions 128-129 from the 5' end of the DNA fragment shown in SEQ ID No.1; the LZTFL1_A_4 is the methylation level of the CpG site shown at positions 156-157 from the 5' end of the DNA fragment shown in SEQ ID No.1; the BCAR3_B_1 is the methylation level of the CpG site shown at positions 58-59 from the 5' end of the DNA fragment shown in SEQ ID No.2; the BCAR3_B_2.3.4.5 are methylation levels of the CpG sites shown in SEQ ID The methylation levels of the CpG sites shown at positions 82-83, 84-85, 86-87 and 88-89 of the DNA fragment shown in SEQ ID No.2 from the 5' end; the BCAR3_B_6 is the methylation level of the CpG site shown at positions 113-114 of the DNA fragment shown in SEQ ID No.2 from the 5' end; the LPR4_C_1 is the methylation level of the CpG site shown at positions 35-36 of the DNA fragment shown in SEQ ID No.3 from the 5' end; the LPR4_C_2 is the methylation level of the CpG site shown at positions 60-61 of the DNA fragment shown in SEQ ID No.3 from the 5' end; the LPR4_C_3 is the methylation level of the CpG site shown at positions 88-89 of the DNA fragment shown in SEQ ID No.3 from the 5' end; the LPR4_C_4 is the methylation level of the CpG site shown at positions 88-89 of the DNA fragment shown in SEQ ID The methylation level of the CpG site shown at positions 100-101 from the 5' end of the DNA fragment shown in No.3; the LPR4_C_5 is the methylation level of the CpG site shown at positions 126-127 from the 5' end of the DNA fragment shown in SEQ ID No.3; the LPR4_C_6 is the methylation level of the CpG site shown at positions 135-136 from the 5' end of the DNA fragment shown in SEQ ID No.3. The threshold of the model is 0.5. Patients with a detection index greater than 0.5 calculated by the model are candidates for thyroid malignant tumors, and patients with a detection index less than 0.5 are candidates for thyroid benign tumors.

[0116] In the above aspects, the detecting the methylation level of each gene in the DNA methylation marker combination is detecting the methylation level of each gene in the DNA methylation marker combination in a tumor tissue sample.

[0117] In the present invention, the methylation level of each gene in the DNA methylation marker combination in thyroid malignant tumor tissue is significantly lower than the methylation level at the corresponding site in thyroid benign tumor.

[0118] In the present invention, the methylation level of each gene in the DNA methylation marker combination in thyroid cancers with different clinical characteristics, such as papillary carcinoma, follicular carcinoma, medullary carcinoma and undifferentiated carcinoma tumor tissues, all showed a downward trend.

[0119] In the present invention, the methylation levels of the methylation sites of the DNA methylation marker combination in the tissues show a decreasing trend as the stage of thyroid malignant tumor increases.

[0120] For any of the above LZTFL1 genes, please refer to the Genbank accession number: NM_020347.4 (GI: 1519243235), transcript variant 1; NM_001276378.2 (GI: 1889541251), transcript variant 2; NM_001276379.2 (GI: 1889706081), transcript variant 3; NM_001386451.1 (GI: 1894925280), transcript variant 4; NM_001386452.1 (GI: 1894925276), transcript variant 5.

[0121] For any of the above BCAR3 genes, please refer to Genbank accession numbers: NM_001261408.2 (GI: 1675178356), transcript variant 1; NM_003567.4 (GI: 1519311780), transcript variant 2; NM_001261409.1 (GI: 387157908), transcript variant 3; NM_001261410.2 (GI: 1675175859), transcript variant 4; NM_001308251.1 (GI: 815891261), transcript variant 5; NM_001412043.1 (GI: 2302515426), transcript variant 7; N M_001412044.1 (NM_001412044.1 GI: 2303311949), transcript variant 8; NM_001412045.1 (GI: 2303318054), transcript variant 9; NM_001412046.1 (GI: 2303321770), transcript variant 10; NM_001412048.1 (GI: 2303321774), transcript variant 12; NM_001412049.1 (GI: 2303321776), transcript variant 13; NM_001412050.1 (GI: 2303321778), transcript variant 14; 051.1 (GI: 2303321780), transcript variant 15; NM_001412052.1 (GI: 2303321783), transcript variant 16; NM_001412054.1 (GI: 2302515512), transcript variant 17; NM_001412055.1 (GI: 2302515475) transcript variant 18; NM_001412056.1 (GI: 2302515533), transcript variant 19; NM_001412057.1 (GI: 2302515479) transcript variant 20; NM_001412058.1 (GI: 2302515417, Transcript variant 21; NM_001412059.1 (GI: 2302515497), transcript variant 22; NM_001412060.1 (GI: 2302515409), transcript variant 23; NM_001412061.1 (GI: 2302515435), transcript variant 24; NM_001412062.1 (GI: 2302515466), transcript variant 25; NM_001412064.1 (GI: 2302515451), transcript variant 27; NM_001412065.1 (GI: 2302515508), transcript variant 28; NM_001412066.1 (GI: 2302515449), transcript variant 29; NM_001412067.1 (GI: 2302515528), transcript variant 30; NM_001412072.1 (GI: 2302515488), transcript variant 31; NM_001412073.1 (GI: 2302515437), transcript variant 32; NM_001412074.1 (GI: 2302515463), transcript variant 33; 1412074.1 (GI: 2302515492), transcript variant 33; NM_001412075.1 (GI: 2302515490), transcript variant 34; NM_001412076.1 (GI: 2302515414) transcript variant 35; NM_001412077.1 (GI: 2302515526) transcript variant 36. .

[0122] For details of any of the above LRP4 genes, please refer to Genbank accession number: NC_000011.10.

[0123] The present invention proves that the combined gene methylation of LZTFL1, BCAR3 and LPR4 in biopsy samples can be used as a potential marker combination for differential diagnosis of benign thyroid tumors and malignant thyroid tumors, and thyroid malignant tumors of different subtypes or stages. The present invention has important scientific significance and clinical application value for distinguishing benign thyroid tumors and malignant thyroid tumors, thyroid malignant tumors of different subtypes or stages, and guiding the formulation of reasonable clinical treatment plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0124] Figure 1 A computer flow chart for assisting in distinguishing type A samples from type B samples according to the present invention.

[0125] Figure 2 Schematic diagram of the mathematical model.

[0126] Figure 3 Give an example to illustrate the mathematical model of benign and malignant thyroid tumors. DETAILED DESCRIPTION

[0127] The present invention is further described in detail below in conjunction with specific embodiments, and the examples provided are only for illustrating the present invention, rather than for limiting the scope of the present invention. The examples provided below can be used as a guide for further improvements by those of ordinary skill in the art, and do not constitute a limitation of the present invention in any way.

[0128] The experimental methods in the following examples, unless otherwise specified, are all conventional methods, and are performed according to the techniques or conditions described in the literature in the field or according to the product instructions. The materials, reagents, etc. used in the following examples, unless otherwise specified, can all be obtained from commercial channels.

[0129] Figure 1 A computer flow chart for assisting in distinguishing type A samples from type B samples according to the present invention.

[0130] In step S1, the methylation level data of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples are collected;

[0131] In step S2, the methylation level data of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples in S1 are used to establish a mathematical model through binary classification logistic regression according to the classification method of type A and type B, and determine the threshold value of classification judgment;

[0132] In step S3, the methylation level data of each gene in the DNA methylation marker combination of the subject is input;

[0133] In step S4, the methylation level data of each gene in the DNA methylation marker combination of the subject in S3 is substituted into the mathematical model in S2 to calculate the detection index;

[0134] In step S5, the detection index in S4 is compared with the threshold in S2 to obtain a comparison result;

[0135] In step S6, a conclusion is output as to whether the type of the sample to be tested is type A or type B according to the comparison result in S5.

[0136] Example 1. Primer design for detecting methylation sites of LZTFL1, BCAR3 and LPR4 genes

[0137] In this test, CpG sites on the LZTFL1_A fragment, BCAR3_B fragment and LPR4_C fragment were selected for correlation analysis between methylation levels and thyroid malignancies.

[0138] The LZTFL1_A fragment (SEQ ID No. 1) is located in the hg19 reference genome chr3:45957192-45957378, positive strand.

[0139] The BCAR3_B fragment (SEQ ID No. 2) is located in the hg19 reference genome chr1:94240928-94241067, negative sense strand.

[0140] The LPR4_C fragment (SEQ ID No. 3) is located in the hg19 reference genome chr11:46940579-46940753, positive strand.

[0141] The site information in the LZTFL1_A fragment is shown in Table 1.

[0142] The site information in the BCAR3_B fragment is shown in Table 2.

[0143] The site information in the LPR4_C fragment is shown in Table 3.

[0144] Table 1. CpG site information in LZTFL1_A fragment

[0145] CpG sites The location of the CpG site in the sequence LZTFL1_A_1 SEQ ID No.1 from 5' end 46-47 LZTFL1_A_2 SEQ ID No.1 from 5' end 75-76 LZTFL1_A_3 SEQ ID No.1 from 128 to 129 at the 5' end LZTFL1_A_4 SEQ ID No.1 from 156 to 157 at the 5' end

[0146] Table 2. CpG site information in BCAR3_B fragment

[0147] CpG sites The location of the CpG site in the sequence BCAR3_B_1 SEQ ID No.2 from 5' end 58-59 BCAR3_B_2 SEQ ID No.2 from 5' end 82-83 BCAR3_B_3 SEQ ID No.2 from 5' end 84-85 BCAR3_B_4 SEQ ID No.2 from 5' end 86-87 BCAR3_B_5 SEQ ID No.2 from 5' end 88-89 BCAR3_B_6 SEQ ID No.2 from 113 to 114 at the 5' end

[0148] Table 3. CpG site information in LPR4_C fragment

[0149] CpG sites The location of the CpG site in the sequence LPR4_C_1 SEQ ID No.3 from 5' end 35-36 LPR4_C_2 SEQ ID No.3 from 6' end 60-61 LPR4_C_3 SEQ ID No.3 from 88 to 89 at the 7' end LPR4_C_4 SEQ ID No.3 from 8' end 100-101 LPR4_C_5 SEQ ID No.3 from 9' end 126-127 LPR4_C_6 SEQ ID No.3 from 10' end 135-136

[0150] Specific PCR primers were designed for the three fragments (LZTFL1_A fragment, BCAR3_B fragment and LPR4_C fragment), as shown in Table 4. SEQ ID No.4, SEQ ID No.6 and SEQ ID No.8 are forward primers; SEQ ID No.5, SEQ ID No.7 and SEQ ID No.9 are reverse primers. From the 5' end, positions 1 to 10 are non-specific tags, and positions 11 to 35 are specific primer sequences in SEQ ID No.4, SEQ ID No.6 and SEQ ID No.8; from the 5' end, positions 1 to 31 are non-specific tags, and positions 32 to 56 are specific primer sequences in SEQ ID No.5, SEQ ID No.7 and SEQ ID No.9. SNP and CpG sites are not included in the primer sequence.

[0151] Table 4. Methylation detection primer sequences of the DNA methylation marker combination of the present invention

[0152]

[0153] Example 2: Detection of combined methylation of LZTFL1, BCAR3 and LPR4 genes and analysis of results (training set)

[0154] 1. Research Sample

[0155] With the informed consent of the patients, a total of 380 cases of benign thyroid tumor tissues and 598 cases of malignant thyroid tumor tissues were collected. The staging of thyroid cancer was based on the eighth edition of the staging system of the American Joint Committee on Cancer (AJCC). Benign thyroid tumors included 186 adenomas and 194 adenoma. According to the pathological type, thyroid malignant tumors include four major categories: papillary thyroid carcinoma, follicular thyroid carcinoma, medullary carcinoma, and anaplastic carcinoma. The 598 patients with thyroid malignant tumors collected this time included 380 cases of papillary thyroid carcinoma, 138 cases of follicular thyroid carcinoma, 44 cases of medullary thyroid carcinoma, and 36 cases of anaplastic thyroid carcinoma. According to the pathological staging, there were 470 stage I patients, 68 stage II patients, 24 stage III patients, and 36 stage IV patients among the 598 patients with thyroid malignant tumors.

[0156] 2. Methylation Detection

[0157] 1. Extract total DNA from tumor tissue.

[0158] 2. The DNA extracted above was subjected to sulfite conversion to obtain sulfite-converted DNA, and the specific operation was performed according to the instructions of Qiagen's DNA methylation kit. Here, the principle is that "bisulfite can convert all unmethylated cytosines into uracil, while methylated cytosines remain unchanged".

[0159] 3. Using the bisulfite-treated DNA in step 2 as a template, the three pairs of specific primers in Table 4 were used to perform PCR amplification in a reaction system according to conventional PCR reaction requirements using DNA polymerase. All primers used a conventional standard PCR reaction system and were amplified according to the procedure shown in the following Table 5.

[0160] Table 5. Amplification procedures

[0161]

[0162] 4. Take the amplified product of step 3 and perform DNA methylation analysis by time-of-flight mass spectrometry. The specific method is as follows:

[0163] (1) Add 2 μl of shrimp alkaline phosphate (SAP) solution (0.3 ml SAP [0.5 U] + 1.7 ml H2O) to 5 μl of PCR product and incubate in a PCR instrument according to the following procedure (37°C, 20 min → 85°C, 5 min → 4°C, 5 min);

[0164] (2) Take out 2 μl of the SAP-treated product obtained in step (1), add it to 5 μl of T-Cleavage reaction system according to the instructions, and then incubate at 37°C for 3 h;

[0165] (3) Take the product of step (2), add 19 μl of deionized water, and then incubate with 6 μg of Resin on a rotary shaker for 1 h for deionization;

[0166] (4) Centrifuge at 2000 rpm for 5 min at room temperature, and load a small amount of supernatant onto 384 SpectroCHIP using a Nanodispenser robot;

[0167] (5) Time-of-flight mass spectrometry analysis; the data were collected using SpectroACQUIRE v3.3.1.3 software and visualized using MassArray EpiTyper v1.2 software.

[0168] The reagents used in the above-mentioned time-of-flight mass spectrometry detection are all from a kit (T-Cleavage MassCLEAVEReagent Auto Kit, catalog number: 10129A); the detection instrument used in the above-mentioned time-of-flight mass spectrometry detection is a MassARRAY○R Analyzer Chip Prep Module 384, model: 41243; and the above-mentioned data analysis software is the software provided by the detection instrument.

[0169] 5. Analyze the data obtained in step 4.

[0170] Data statistical analysis was performed using SPSS Statistics 23.0.

[0171] Nonparametric tests were used for comparative analysis between the two groups.

[0172] The discriminatory effect of the combination of multiple CpG sites on different sample groups was achieved by using the statistical methods of logistic regression and receiver operating curve.

[0173] All statistical tests were two-sided, and p values ​​< 0.05 were considered statistically significant.

[0174] Through mass spectrometry experiments, a total of 13 distinguishable peaks were obtained. Using SpectroACQUIRE v3.3.1.3 software, the methylation level of each sample at each CpG site can be automatically calculated by peak area according to the formula "methylation level = peak area of ​​methylated fragment / (peak area of ​​non-methylated fragment + peak area of ​​methylated fragment)".

[0175] 3. Statistical Analysis

[0176] The median was used to represent the methylation level, and the nonparametric test was used to compare the differences in methylation levels between two or more groups. The receiver operating characteristic curve (ROC curve) was used to evaluate the differential diagnostic value of a single CpG site and a combination of multiple CpG sites. The difference was considered statistically significant when the bilateral P value was less than 0.05. All data were statistically analyzed using SPSS 25.0.

[0177] IV. Results Analysis

[0178] 1. Methylation levels of LZTFL1, BCAR3 and LPR4 genes in benign and malignant thyroid tumors

[0179] The methylation levels of all CpG sites in LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No. 1 to SEQ ID No. 3) were analyzed using tissue samples from 380 cases of benign thyroid tumors and 598 cases of malignant thyroid tumors. The specific results of the methylation levels of the three fragments of LZTFL1, BCAR3 and LPR4 combined genes in cases of benign thyroid tumors, malignant thyroid tumors, papillary thyroid carcinoma, follicular thyroid carcinoma, medullary thyroid carcinoma, anaplastic thyroid carcinoma, stage I thyroid malignant tumors, stage II thyroid malignant tumors, stage III thyroid malignant tumors and stage IV thyroid malignant tumors are shown in Table 6. The results showed that the median methylation level of the combined gene LZTFL1, BCAR3 and LPR4 in benign thyroid tumors was 0.61 (IQR = 0.36-0.87), and the median methylation level of the combined gene LZTFL1, BCAR3 and LPR4 in malignant thyroid tumors was 0.56 (IQR = 0.35-0.77). The median methylation levels of the combined gene LZTFL1, BCAR3 and LPR4 in different subtypes were as follows: 0.59 (IQR = 0.36-0.81) in papillary thyroid carcinoma, 0.57 (IQR = 0.36-0.81) in follicular thyroid carcinoma, 0.58 (IQR = 0.36-0.82) in thyroid carcinoma, 0.59 ... The median methylation levels of LZTFL1, BCAR3 and LPR4 combination genes in different stages were as follows: stage I was 0.58 (IQR=0.36-0.81), stage II was 0.56 (IQR=0.34-0.79), stage III was 0.53 (IQR=0.32-0.78), and stage IV was 0.51 (IQR=0.31-0.75). By comparing and analyzing the methylation levels among the several, it was found that the methylation levels of all CpG sites in the LZTFL1, BCAR3 and LPR4 combination genes in benign thyroid tumors were significantly higher than those in the LZTFL1, BCAR3 and LPR4 combination genes in malignant thyroid tumors (Table 6). In addition, the methylation levels of the methylation sites on the DNA fragments shown in SEQ ID No.1, SEQ ID No.2 and SEQ ID No.3 in the LZTFL1, BCAR3 and LPR4 combination genes in several types of thyroid cancer subtypes listed in the present invention, such as papillary thyroid carcinoma, follicular thyroid carcinoma, medullary thyroid carcinoma and anaplastic thyroid carcinoma tissues, were successively decreasing (Table 6), and the difference in methylation levels between benign thyroid tumors was becoming more and more obvious.In addition, with the increase of the stage of thyroid malignant tumors, the methylation levels of the methylation sites on the DNA fragments shown by SEQ ID No. 1, SEQ ID No. 2 and SEQ ID No. 3 in the LZTFL1, BCAR3 and LPR4 combination genes in the tissues also showed a decreasing trend (Table 6), and the difference in methylation levels between benign thyroid tumors was becoming more and more obvious.

[0180] Table 6. Methylation levels of genes in benign and malignant thyroid tumors and their subtypes and stages

[0181]

[0182] Note: The CpG sites in the table refer to distinguishable CpG sites.

[0183] 2. Establishment of a mathematical model for LZTFL1, BCAR3 and LPR4 combined gene methylation to assist cancer diagnosis

[0184] The mathematical model established by the present invention can be used to achieve the following purposes:

[0185] (1) Differentiate between patients with malignant thyroid tumors and benign thyroid tumors;

[0186] (2) Differentiate between benign thyroid tumors and different subtypes of malignant thyroid tumors;

[0187] (3) Differentiate between benign thyroid tumors and different stages of malignant thyroid tumors;

[0188] (4) Differentiate between different subtypes of thyroid malignancies;

[0189] (5) Differentiate between different stages of thyroid malignant tumors.

[0190] The mathematical model is established as follows:

[0191] (A) Data source: The methylation levels of target CpG sites (a combination of one or more of Tables 1 to 3) of tissue samples from 380 cases of benign thyroid tumors and 598 cases of malignant thyroid tumors (380 cases of papillary thyroid carcinoma, 138 cases of follicular thyroid carcinoma, 44 cases of medullary thyroid carcinoma, and 36 cases of anaplastic thyroid carcinoma) listed in step 1 (detection method is the same as step 2).

[0192] (B) Model building

[0193] According to the needs, select any two different types of patient data as the training set (for example: patients with benign thyroid tumors and malignant thyroid tumors, patients with benign thyroid tumors and papillary thyroid cancer, patients with benign thyroid tumors and follicular thyroid cancer, patients with benign thyroid tumors and medullary thyroid cancer, patients with benign thyroid tumors and anaplastic thyroid cancer, patients with papillary thyroid cancer and follicular thyroid cancer, patients with papillary thyroid cancer and medullary thyroid cancer, patients with papillary thyroid cancer and anaplastic thyroid cancer, patients with follicular thyroid cancer and medullary thyroid cancer, patients with follicular thyroid cancer and anaplastic thyroid cancer, patients with medullary thyroid cancer and anaplastic thyroid cancer, patients with benign thyroid tumors and stage I thyroid malignant tumors, patients with benign thyroid The data for establishing the model were selected from patients with stage Ⅱ thyroid malignancies, patients with benign thyroid tumors and stage Ⅲ thyroid malignancies, patients with benign thyroid tumors and stage Ⅳ thyroid malignancies, patients with stage Ⅰ thyroid malignancies and stage Ⅱ thyroid malignancies, patients with stage Ⅰ thyroid malignancies and stage Ⅲ thyroid malignancies, patients with stage Ⅰ thyroid malignancies and stage Ⅳ thyroid malignancies, patients with stage Ⅱ thyroid malignancies and stage Ⅲ thyroid malignancies, patients with stage Ⅱ thyroid malignancies and stage Ⅳ thyroid malignancies, patients with stage Ⅲ thyroid malignancies and stage Ⅳ thyroid malignancies) and statistical software such as SAS, R, and SPSS were used to establish the mathematical model through the formula using the statistical method of binary logistic regression. The value corresponding to the maximum Youden index calculated by the mathematical model formula was the threshold value or 0.5 was directly set as the threshold value. After the test sample was tested and substituted into the model calculation, the detection index obtained was greater than the threshold value and was classified into one category (category B), less than the threshold value and was classified into another category (category A), and equal to the threshold value as an uncertain gray area. When predicting a new sample to be tested to determine which category it belongs to, firstly, the methylation level of one or more CpG sites on the LZTFL1, BCAR3 and LPR4 genes of the sample to be tested is detected by a DNA methylation determination method, and then the data of these methylation levels are substituted into the above-mentioned mathematical model to calculate the detection index corresponding to the sample to be tested, and then the detection index corresponding to the sample to be tested and the threshold are compared, and the category of the sample to be tested is determined according to the comparison result.

[0194] For example: Figure 2As shown, the data of the methylation level of all or part of the CpG site combination in the LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No.1 to SEQID No.3) in the training set are used by statistical software such as SAS, R, SPSS, etc., using the formula of binary logistic regression to establish a mathematical model for distinguishing between Class A and Class B. The mathematical model is a binary logistic regression model, specifically: ln(y / (1-y))=b0+b1x1+b2x2+b3x3+…+bnXn, wherein y is the dependent variable, that is, the detection index obtained after the methylation level of one or more methylation sites of the sample to be tested is substituted into the model, b0 is a constant, x1-xn is an independent variable, that is, the methylation level of one or more methylation sites of the test sample (each value is a value between 0-1), and b1-bn is the weight assigned to the methylation level of each site by the model. In specific applications, a mathematical model is first established based on the methylation levels (x1-xn) of one or more DNA methylation sites of samples that have been tested in the training set and their known classifications (class A or class B, assigning y values ​​of 0 and 1 respectively), thereby determining the constant b0 of the mathematical model and the weights b1-bn of each methylation site, and the numerical value corresponding to the maximum Youden index calculated by the mathematical model is used as the threshold or 0.5 is directly set as the threshold for division. The detection index, i.e., y value, obtained after the test sample is tested and substituted into the model calculation is classified as class B if it is greater than the threshold, classified as class A if it is less than the threshold, and classified as an uncertain gray area if it is equal to the threshold. Among them, class A and class B are two corresponding classifications (the grouping of the two classifications, which group is class A and which group is class B, should be determined according to the specific mathematical model, and no agreement is made here). When predicting the subject's sample to determine which class it belongs to, first collect the subject's biopsy sample (i.e., tumor tissue), and then extract DNA from it. After the extracted DNA was converted by bisulfite, the methylation level of all or part of the CpG site combination in the subject's LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No.1 to SEQ ID No.3) was detected using a DNA methylation determination method, and then the methylation data obtained by the detection was substituted into the above-mentioned mathematical model.If the value calculated by substituting the methylation level of all or part of the CpG site combination in the LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No.1 to SEQ ID No.3) of the subject into the above mathematical model is greater than the threshold value, the subject is determined to belong to the same category (category B) as the subject with a detection index greater than the threshold value in the training set; if the methylation level data of all or part of the CpG site combination in the LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No.1 to SEQ ID No.3) of the subject is substituted into the above mathematical model, the value calculated by the detection index is less than the threshold value, the subject belongs to the same category (category A) as the subject with a detection index less than the threshold value in the training set; if the value calculated by substituting the methylation level data of all or part of the CpG site combination in the LZTFL1, BCAR3 and LPR4 gene fragments (SEQ ID No.1 to SEQ ID No.3) of the subject into the above mathematical model is equal to the threshold value, it cannot be determined whether the subject is category A or category B.

[0195] For example: Figure 3As shown, the following 13 distinguishable CpG sites (LZTFL1_A_1, LZTFL1_A_2, LZTFL1_A_3, LZTFL1_A_4, BCAR3_B_1, BCAR3_B_2.3.4.5, BCAR3_B_6, LPR4_C_1, LPR4_C_2, LPR4_C_3, LPR4_C_4, LPR4_C_5, LPR4_C_6) of the combined gene of LZTFL1_A, BCAR3_B and LPR4_C are illustrated. Application of methylation and mathematical modeling in distinguishing benign thyroid tumor and malignant thyroid tumor tissues: The data of methylation levels of the above 13 distinguishable CpG sites of the LZTFL1_A, BCAR3_B and LPR4_C combination genes that have been detected in the training set of patients with benign thyroid tumors and malignant thyroid tumors (here: 380 patients with benign thyroid tumors and 598 patients with malignant thyroid tumors) were used to establish a mathematical model for distinguishing patients with malignant thyroid tumors using the formula of binary logistic regression using SPSS software or R software. The mathematical model is a two-class logistic regression model, which determines the constant b0 of the mathematical model and the weights b1-bn of each methylation site, which are specifically: lg(y / (1-y))=2.710+0.539*LZTFL1_A_1-4.362*LZTFL1_A_2+3.115*LZTFL1_A_3+0.901*LZTFL1_A_4–5.183*BCAR3_B_1-0.718*BCAR3_B_2.3.4.5+2. 609*BCAR3_B_6+1.402*LPR4_C_1-0.113*LPR4_C_2+4.091*LPR4_C_3-0.692*LPR4_C_4+1.341*LPR4_C_5+0.805*LPR4_C_6, where y is the dependent variable, that is, the detection index is obtained by substituting the methylation levels of the above 13 distinguishable CpG sites of the LZTFL1_A, BCAR3_B and LPR4_C combination genes of the test sample into the model and then converting them into the test index. By setting 0.5 as the threshold, the methylation levels of the above 13 distinguishable CpG sites of the LZTFL1_A, BCAR3_B and LPR4_C combination genes of the tested samples were substituted into the model for calculation, and the obtained detection index, i.e., the y value was less than the threshold and classified as a patient with benign thyroid tumor, greater than the threshold and classified as a patient with malignant thyroid tumor, and equal to the threshold, it was uncertain whether it was a patient with benign thyroid tumor or a patient with malignant thyroid tumor. The area under the curve (AUC) of this model was calculated to be 0.95 (Table 7).The specific subject judgment method is shown below. Biopsy samples were collected from two subjects (A and B) to extract DNA. After the extracted DNA was converted by bisulfite, the methylation levels of the 13 distinguishable CpG sites (LZTFL1_A_1, LZTFL1_A_2, LZTFL1_A_3, LZTFL1_A_4, BCAR3_B_1, BCAR3_B_2.3.4.5, BCAR3_B_6, LPR4_C_1, LPR4_C_2, LPR4_C_3, LPR4_C_4, LPR4_C_5, LPR4_C_6) of the subjects were detected by DNA methylation determination method. Then the methylation level data information obtained by the detection was substituted into the above mathematical model. When the methylation level data of the above 13 distinguishable CpG sites of the LZTFL1_A, BCAR3_B and LPR4_C combination genes of Subject A are substituted into the above mathematical model, the calculated value is 0.84, which is greater than 0.5, and Subject A is determined to be a patient with malignant thyroid tumor (consistent with the clinical diagnosis); when the methylation level data of the above 13 distinguishable CpG sites of the LZTFL1_A, BCAR3_B and LPR4_C combination genes of Subject B are substituted into the above mathematical model, the calculated value is 0.37, which is less than 0.5, and Subject B is determined to be a patient with benign thyroid tumor (consistent with the clinical diagnosis).

[0196] (C) Model effect evaluation

[0197] According to the above methods, detection methods for detecting patients with benign thyroid tumors and malignant thyroid tumors, patients with benign thyroid tumors and papillary thyroid cancer, patients with benign thyroid tumors and follicular thyroid cancer, patients with benign thyroid tumors and medullary thyroid cancer, patients with benign thyroid tumors and anaplastic thyroid cancer, patients with papillary thyroid cancer and follicular thyroid cancer, patients with papillary thyroid cancer and medullary thyroid cancer, patients with papillary thyroid cancer and anaplastic thyroid cancer, patients with follicular thyroid cancer and medullary thyroid cancer, patients with follicular thyroid cancer and anaplastic thyroid cancer, patients with medullary thyroid cancer and anaplastic thyroid cancer, patients with benign thyroid tumors and stage I malignant thyroid tumors were established respectively. , patients with benign thyroid tumors and stage II thyroid malignant tumors, patients with benign thyroid tumors and stage III thyroid malignant tumors, patients with benign thyroid tumors and stage IV thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage II thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage IV thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage IV thyroid malignant tumors, and the effectiveness of the mathematical model was evaluated by the receiver operating characteristic curve (ROC curve). The larger the area under the curve (AUC) obtained by the ROC curve, the better the discrimination of the model and the more effective the molecular marker. The evaluation results after the mathematical model was constructed using different CpG sites are shown in Tables 7 to 10.

[0198] The model constructed by the combination of LZTFL1_A, BCAR3_B and LPR4_C methylation markers or any two of them has predictive performance for benign thyroid tumors and thyroid cancer and their different subtypes. As shown in Table 7, the three genes have the best predictive performance when combined together, and the AUC value for distinguishing benign thyroid tumors from malignant thyroid tumors can reach 0.95.

[0199] The model constructed by the combination of LZTFL1_A, BCAR3_B and LPR4_C methylation markers or the combination of any two of them has predictive performance for different stages of benign thyroid tumors and thyroid cancer. As shown in Table 8, the three genes combined together have the best predictive efficiency, with the highest AUC value reaching 0.98 (benign thyroid tumor vs stage IV thyroid cancer).

[0200] The model constructed by the combination of LZTFL1_A, BCAR3_B and LPR4_C methylation markers or any two of them has predictive performance for different subtypes of thyroid cancer. As shown in Table 9, the three genes have the best predictive performance when combined together, with the highest AUC value reaching 0.95 (papillary thyroid carcinoma vs. undifferentiated carcinoma).

[0201] The combination of LZTFL1_A, BCAR3_B and LPR4_C methylation markers or the model constructed by combining any two of them has predictive performance for different stages of thyroid cancer. As shown in Table 10, the three genes combined have the best predictive efficiency, with the highest AUC value reaching 0.95 (stage I vs stage IV).

[0202] In summary, combining the three methylation markers of LZTFL1_A, BCAR3_B and LPR4_C for early screening of thyroid cancer can effectively distinguish patients with benign thyroid tumors and papillary thyroid cancer, patients with benign thyroid tumors and follicular thyroid cancer, patients with benign thyroid tumors and medullary thyroid cancer, patients with benign thyroid tumors and anaplastic thyroid cancer, patients with papillary thyroid cancer and follicular thyroid cancer, patients with papillary thyroid cancer and medullary thyroid cancer, patients with papillary thyroid cancer and anaplastic thyroid cancer, patients with follicular thyroid cancer and medullary thyroid cancer, patients with follicular thyroid cancer and anaplastic thyroid cancer, patients with medullary thyroid cancer and anaplastic thyroid cancer. Patients with anaplastic thyroid carcinoma, patients with benign thyroid tumors and stage I thyroid malignant tumors, patients with benign thyroid tumors and stage II thyroid malignant tumors, patients with benign thyroid tumors and stage III thyroid malignant tumors, patients with benign thyroid tumors and stage IV thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage II thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage IV thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage IV thyroid malignant tumors, patients with stage III thyroid malignant tumors and stage IV thyroid malignant tumors.

[0203] Table 7. CpG sites of LZTFL1_A, BCAR3_B and LPR4_C combined genes and their combinations for distinguishing benign thyroid tumors from different subtypes of malignant thyroid tumors

[0204]

[0205] Note: The sites mentioned in the table are distinguishable sites

[0206] Table 8. CpG sites of LZTFL1_A, BCAR3_B and LPR4_C combined genes and their combinations for distinguishing benign thyroid tumors from thyroid malignant tumors of different stages

[0207]

[0208]

[0209] Note: The sites mentioned in the table are distinguishable sites

[0210] Table 9. CpG sites of LZTFL1_A, BCAR3_B and LPR4_C combined genes and their combinations are used to distinguish patients with different subtypes of thyroid malignancies

[0211]

[0212] Note: The sites mentioned in the table are distinguishable sites

[0213] Table 10. CpG sites of LZTFL1_A, BCAR3_B and LPR4_C combined genes and their combinations for distinguishing patients with different stages of thyroid malignancies

[0214]

[0215] Note: The sites mentioned in the table are distinguishable sites

[0216] Example 3. Experimental results of LZTFL1, BCAR3 and LRP4 gene methylation markers in the validation set

[0217] 1. Research Sample

[0218] The present invention selects another group of subjects as a validation set (the samples are completely different from the training set in Example 2), and the specific sample collection conditions are as follows:

[0219] 190 cases of benign thyroid tumor tissues and 300 cases of malignant thyroid tumor tissues. The staging of thyroid cancer is based on the eighth edition of the staging system of the American Joint Committee on Cancer (AJCC). According to the pathological type, thyroid malignant tumors include four major categories: papillary thyroid cancer, follicular thyroid cancer, medullary carcinoma, and anaplastic cancer. The 300 patients with thyroid malignancies collected this time included 188 cases of papillary thyroid cancer, 70 cases of follicular thyroid cancer, 22 cases of medullary carcinoma, and 20 cases of anaplastic cancer. According to the pathological staging, there were 200 stage I patients, 40 stage II patients, 20 stage III patients, and 40 stage IV patients among the 300 patients with thyroid malignancies.

[0220] 2. Methylation Detection

[0221] The methylation detection and data analysis methods are the same as those in Example 2.

[0222] 3. Results Analysis

[0223] As shown in Table 11, the specific results of the methylation levels of the three fragments of the LZTFL1, BCAR3 and LRP4 combination genes in cases of benign thyroid tumors, malignant thyroid tumors, papillary thyroid carcinoma, follicular thyroid carcinoma, medullary carcinoma, undifferentiated carcinoma, stage I thyroid malignant tumors, stage II thyroid malignant tumors, stage III thyroid malignant tumors and stage IV thyroid malignant tumors are shown in Table 11. The results showed that the median methylation level of the LZTFL1 combination gene in benign thyroid tumors was 0.61 (IQR = 0.35-0.81), the median methylation level of the LZTFL1, BCAR3 and LRP4 genes in malignant thyroid tumors was 0.54 (IQR = 0.33-0.77), and the median methylation level of the LZTFL1 gene in different subtypes was as follows: papillary carcinoma 0.58 (IQR = 0.34-0.81), follicular carcinoma 0.56 (IQR = 0.3 The median methylation levels of LZTFL1 gene in different stages were as follows: stage I was 0.56 (IQR = 0.35-0.80), stage II was 0.54 (IQR = 0.32-0.78), stage III was 0.51 (IQR = 0.31-0.77), and stage IV was 0.50 (IQR = 0.31-0.75). By comparing the methylation levels among the three groups, it was found that the methylation levels of all CpG sites in LZTFL1, BCAR3 and LRP4 genes in benign thyroid tumors were significantly higher than those in malignant thyroid tumors. In addition, the methylation levels of the methylation sites on the DNA fragments shown in SEQ ID No. 1, SEQ ID No. 2 and SEQ ID No. 3 in the LZTFL1, BCAR3 and LRP4 genes of several thyroid cancer subtypes listed in the present invention, such as papillary carcinoma, follicular carcinoma, medullary carcinoma and undifferentiated carcinoma tissues, are decreasing in sequence, and the difference in methylation levels between benign thyroid tumors is becoming more and more obvious. In addition, with the increase in the stage of thyroid malignant tumors, the methylation levels of the methylation sites on the DNA fragments shown in SEQ ID No. 1, SEQ ID No. 2 and SEQ ID No. 3 in the LZTFL1, BCAR3 and LRP4 genes in the tissues also show a decreasing trend, and the difference in methylation levels between benign thyroid tumors is becoming more and more obvious.

[0224] Table 11. Methylation levels of genes in benign and malignant thyroid tumors and their subtypes and stages

[0225]

[0226] Note: The CpG sites in the table refer to distinguishable CpG sites.

[0227] 6. Model Validation

[0228] The methylation levels of the three fragments of the LZTFL1, BCAR3 and LRP4 gene combination of different types of samples in the validation set were substituted into the various mathematical models obtained in the training set of Example 2 for verification.

[0229] As shown in Tables 12 to 15, consistent with the results of the training set, the prediction efficiency of LZTFL1_A, BCAR3_B and LRP4_C methylation markers is the best when used together, and the AUC value for distinguishing benign thyroid tumors from malignant thyroid tumors can reach 0.93 (Table 12). When LZTFL1_A, BCAR3_B and LRP4_C methylation markers are combined together, they have the best predictive efficacy for different stages of thyroid benign tumors and thyroid cancer, with an AUC value of up to 0.96 (benign thyroid tumors vs stage IV thyroid cancer, Table 13); when LZTFL1_A, BCAR3_B and LRP4_C methylation markers are combined together, they have the best predictive efficacy for different subtypes of thyroid cancer, with an AUC value of up to 0.92 (papillary thyroid carcinoma vs undifferentiated carcinoma, Table 14); when LZTFL1_A, BCAR3_B and LRP4_C methylation markers are combined together, they have the best predictive efficacy for different stages of thyroid cancer, with an AUC value of up to 0.93 (stage I vs stage IV, Table 15).

[0230] Table 12. CpG sites of LZTFL1_A, BCAR3_B and LRP4_C combined genes and their combinations for distinguishing benign thyroid tumors from different subtypes of malignant thyroid tumors

[0231]

[0232] Note: The loci described in the table are distinguishable loci.

[0233] Table 13. CpG sites of LZTFL1_A, BCAR3_B and LRP4_C combined genes and their combinations for distinguishing benign thyroid tumors from thyroid malignant tumors of different stages

[0234]

[0235] Note: The loci described in the table are distinguishable loci.

[0236] Table 14. CpG sites of LZTFL1_A, BCAR3_B and LRP4_C combined genes and their combinations are used to distinguish patients with different subtypes of thyroid malignancies

[0237]

[0238]

[0239] Note: The loci described in the table are distinguishable loci.

[0240] Table 15. CpG sites of LZTFL1_A, BCAR3_B and LRP4_C combined genes and their combinations are used to distinguish patients with different stages of thyroid malignancies

[0241]

[0242] Note: The sites mentioned in the table are distinguishable sites

[0243] In summary, combining the three methylation markers of LZTFL1_A4, BCAR3_B and LRP4_C for early screening of thyroid cancer can effectively distinguish patients with benign thyroid tumors and papillary thyroid cancer, patients with benign thyroid tumors and follicular thyroid cancer, patients with benign thyroid tumors and medullary thyroid cancer, patients with benign thyroid tumors and anaplastic thyroid cancer, patients with papillary thyroid cancer and follicular thyroid cancer, patients with papillary thyroid cancer and medullary thyroid cancer, patients with papillary thyroid cancer and anaplastic thyroid cancer, patients with follicular thyroid cancer and medullary thyroid cancer, patients with medullary thyroid cancer and anaplastic thyroid cancer. patients with benign thyroid tumors and stage I thyroid malignant tumors, patients with benign thyroid tumors and stage II thyroid malignant tumors, patients with benign thyroid tumors and stage III thyroid malignant tumors, patients with benign thyroid tumors and stage IV thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage II thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage I thyroid malignant tumors and stage IV thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage III thyroid malignant tumors, patients with stage II thyroid malignant tumors and stage IV thyroid malignant tumors, and patients with stage III thyroid malignant tumors and stage IV thyroid malignant tumors.

[0244] The present invention has been described in detail above. For those skilled in the art, without departing from the purpose and scope of the present invention, and without the need to carry out unnecessary experimental conditions, the present invention can be implemented in a wide range under equivalent parameters, concentrations and conditions. Although the present invention provides specific embodiments, it should be understood that the present invention can be further improved. In a word, according to the principles of the present invention, the application is intended to include any changes, uses or improvements to the present invention, including departure from the disclosed scope in the application, and changes made with conventional techniques known in the art.

Claims

1. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the following steps: Collecting methylation level data of each gene in a DNA methylation marker combination of n1 type A samples and n2 type B samples; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene, and the LPR4 gene or any two of them; The methylation level data of each gene in the DNA methylation marker combination of the n1 type A samples and the n2 type B samples are used to establish a mathematical model through a binary classification logistic regression method according to the classification method of type A and type B, and determine the threshold value of classification judgment; Inputting the methylation level data of each gene in the DNA methylation marker combination of the subject to be tested; Substituting the methylation level data of each gene in the DNA methylation marker combination of the subject into the mathematical model to calculate the detection index; Comparing the detection index with the threshold to obtain a comparison result; Outputting a conclusion of whether the type of the sample to be tested is type A or type B according to the comparison result; The type A sample and the type B sample are any of the following: (F1) Benign thyroid tumors and malignant thyroid tumors; (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors; (E3) Benign thyroid tumors and different stages of malignant thyroid tumors; (F4) Different subtypes of thyroid malignancies; (F5) Thyroid malignancies at different stages.

2. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, the steps recited in claim 1 are implemented.

3. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps recited in claim 1 are implemented.

4. A system comprising: (G1) Reagents and / or instruments for detecting the methylation level of each gene in the DNA methylation marker combination; The DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of the genes therein; (G2) a device comprising unit X and unit Y; The unit X is used to establish a mathematical model, including a data acquisition module, a data analysis and processing module and a model output module; The data collection module is configured to collect (G1) the methylation level data of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples detected; The data analysis and processing module is configured to receive the methylation level data of each gene in the DNA methylation marker combination of the n1 type A samples and the n2 type B samples from the data acquisition module, establish a mathematical model through a binary classification logistic regression method according to the classification method of type A and type B, and determine the threshold value of classification judgment; The model output module is configured to receive the mathematical model established by the data analysis and processing module and output it; The unit Y is used to determine the type of sample to be tested, and includes a data input module, a data calculation module, a data comparison module and a conclusion output module; The data input module is configured to input (G1) data of the methylation level of each gene in the DNA methylation marker combination of the subject detected; The data operation module is configured to receive the data of the methylation level of each gene in the DNA methylation marker combination of the subject to be tested from the data input module, and substitute the data of the methylation level of each gene in the DNA methylation marker combination of the subject to be tested into the mathematical model established by the data analysis and processing module in the unit X to calculate the detection index; The data comparison module is configured to receive the detection index calculated by the data operation module and compare the detection index with the threshold value determined in the data analysis and processing module in the unit X; The conclusion output module is configured to receive the comparison result from the data comparison module, and output a conclusion of whether the type of the sample to be tested is type A or type B according to the comparison result; The type A sample and the type B sample are any of the following: (F1) Benign thyroid tumors and malignant thyroid tumors; (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors; (F3) Benign thyroid tumors and different stages of malignant thyroid tumors; (F4) Different subtypes of thyroid malignancies; (F5) Thyroid malignancies at different stages.

5. Use of a substance for detecting the methylation level of each gene in a DNA methylation marker combination in the preparation of a product; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of the genes; the use of the product is at least one of the following: (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors; (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors; (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages; (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies; (5) Differentiate or assist in differentiating the different stages of thyroid malignant tumors.

6. Use of a substance for detecting the methylation level of each gene in a DNA methylation marker combination and a medium storing a mathematical model and / or a method for using the mathematical model in preparing a product; the DNA methylation marker combination consists of the LZTFL1 gene, the BCAR3 gene and the LPR4 gene or any two of them; the use of the product is at least one of the following: (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors; (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors; (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages; (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies; (5) Differentiate or assist in differentiating the different stages of thyroid malignancies; The mathematical model is obtained according to a method comprising the following steps: (D1) detecting the methylation level of each gene in the DNA methylation marker combination of n1 type A samples and n2 type B samples respectively; (D2) taking the methylation level data of each gene in the DNA methylation marker combination of all samples obtained in step (D1), establishing a mathematical model by binary logistic regression method according to the classification method of type A and type B, and determining the threshold value of classification judgment; The method for using the mathematical model comprises the following steps: (E1) detecting the methylation level of each gene in the DNA methylation marker combination of the sample to be tested; (E2) substituting the methylation level data of each gene in the DNA methylation marker combination of the sample to be tested obtained in step (E1) into the mathematical model to obtain a detection index; then comparing the detection index with a threshold value, and determining whether the type of the sample to be tested is type A or type B according to the comparison result; The type A sample and the type B sample are any of the following: (F1) Benign thyroid tumors and malignant thyroid tumors; (F2) Benign thyroid tumors and different subtypes of malignant thyroid tumors; (E3) Benign thyroid tumors and different stages of malignant thyroid tumors; (F4) Different subtypes of thyroid malignancies; (F5) Thyroid malignancies at different stages.

7. Use of the medium storing the mathematical model and / or the method for using the mathematical model as claimed in claim 6 in preparing a product; the product is used for at least one of the following purposes: (1) Differentiate or assist in differentiating between benign and malignant thyroid tumors; (2) Differentiate or assist in differentiating benign thyroid tumors from different subtypes of malignant thyroid tumors; (3) Differentiate or assist in differentiating benign thyroid tumors from thyroid malignant tumors of different stages; (4) Differentiate or assist in differentiating different subtypes of thyroid malignancies; (5) Differentiate or assist in differentiating the different stages of thyroid malignant tumors.

8. The computer device or computer program product or computer readable storage medium or system or application according to any one of claims 1 to 7, characterized in that: The methylation level of each gene in the DNA methylation marker combination is composed of the following three items (A1)-(A3) or any two of them: (A1) methylation levels of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene; (A2) methylation levels of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene; (A3) methylation level of all or part of the CpG sites in the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene; and / or The methylation level of each gene in the DNA methylation marker combination is the methylation level of any of the following CpG sites: (B1) all CpG sites on the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene; (B2) all CpG sites on the DNA fragment shown in SEQ ID No. 1 in the LZTFL1 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene; (B3) all CpG sites on the DNA fragment shown in SEQ ID No. 2 in the BCAR3 gene and all CpG sites on the DNA fragment shown in SEQ ID No. 3 in the LPR4 gene; (B4) All CpG sites on the DNA fragment shown by SEQ ID No. 1 in the LZTFL1 gene, all CpG sites on the DNA fragment shown by SEQ ID No. 2 in the BCAR3 gene, and all CpG sites on the DNA fragment shown by SEQ ID No. 3 in the LPR4 gene.

9. The computer device or computer program product or computer readable storage medium or system or application according to any one of claims 1 to 8, characterized in that: The substance for detecting the methylation level of each gene in the DNA methylation marker combination and the reagent for detecting the methylation level of each gene in the DNA methylation marker combination both comprise the following three items (C1)-(C3) or any two thereof: (C1) Primer pairs used to amplify the full length or partial fragment of the LZTFL1 gene; (C2) primer pairs for amplifying the full length or partial fragment of the BCAR3 gene; (C3) a primer pair for amplifying the full length or partial fragment of the LPR4 gene; Furthermore, the partial fragment in (C1) is the DNA fragment shown in SEQ ID No. 1; and / or Furthermore, the partial fragment in (C2) is the DNA fragment shown in SEQ ID No. 2; and / or Furthermore, the partial fragment in (C3) is the DNA fragment shown in SEQ ID No. 3; Furthermore, the primer pair in (C1) is primer pair A, and / or the primer pair in (C2) is primer pair B, and / or the primer pair in (C3) is primer pair C; The primer pair A is a primer pair consisting of primer A1 and primer A2; the primer A1 is a single-stranded DNA represented by nucleotides 11 to 35 of SEQ ID No.4; the primer A2 is a single-stranded DNA represented by nucleotides 32 to 56 of SEQ ID No.5 or SEQ ID No.5; The primer pair B is a primer pair consisting of primer B1 and primer B2; the primer B1 is a single-stranded DNA represented by nucleotides 11 to 35 of SEQ ID No.6 or SEQ ID No.6; the primer B2 is a single-stranded DNA represented by nucleotides 32 to 56 of SEQ ID No.7 or SEQ ID No.7; The primer pair C is a primer pair consisting of primer C1 and primer C2; the primer C1 is a single-stranded DNA represented by SEQ ID No.8 or nucleotides 11-35 of SEQ ID No.8; the primer C2 is a single-stranded DNA represented by SEQ ID No.9 or nucleotides 32-56 of SEQ ID No.

9.

10. The computer device or computer program product or computer readable storage medium or system or application according to any one of claims 1 to 9, characterized in that: The detecting of the methylation level of each gene in the DNA methylation marker combination is detecting the methylation level of each gene in the DNA methylation marker combination in a tumor tissue sample.

Images