Characteristic gene set for detecting generic cancer BRAF mutation activity and application

Abstract

The invention discloses a characteristic gene set for detecting the BRAF mutation activity of panthenic cancer and application, and particularly, a group of characteristic gene sets capable of specifically reflecting the BRAF mutation activity is firstly constructed, and a customized fluorescent quantitative PCR kit is developed on the basis of the characteristic gene sets and is specially used for evaluating the transcriptional activity state of a BRAF pathway in a tumor tissue sample. The characteristic gene set comprises the following genes: ADAM9, CCND1, CLDN1, CTDSPL, DUSP6, EPHA2, ERRFI1, ETV4, ETV5, FZD7, GPRC5A, IGFBP3, ITGA3, MTUS1, PHLDA2, PLEKH1, S100A16, S100A6, SEMA3A, SLC1A1, SPRY4, TFAP2C, TNFRSF12A, TNFRSF21, and TPD52L1. The characteristic gene set can be used for detecting the disease resistance of a human body.

Description

Technical Field

[0001] The present invention belongs to the field of bioinformatics and provides a characteristic gene set for detecting pan-cancer BRAF mutation activity and its application. Background Art

[0002] BRAF is an important oncogene with significant activating mutations (such as V600E) in various tumor types and is widely used as a biomarker for targeted therapy. Traditional BRAF mainstream practices for detecting gene mutations in patents mainly focus on detecting BRAF classical mutation sites (such as V600E, V600K) and mostly use the method of allele-specific amplification (AS-PCR) combined with qPCR, the aim of which is to distinguish mutant and wild-type sequences.

[0003] However, clinical studies have shown that BRAF the mutation status cannot fully predict the responses of all patients to BRAF inhibitors or chemotherapy, showing significant heterogeneity and prognostic divergence. In recent years, more and more studies have focused on the transcriptional activity of the BRAF signaling pathway rather than relying solely on gene mutations as an evaluation indicator. The BRAF pathway can be activated through multiple mechanisms, including enhanced upstream signals, co-mutations, epigenetic regulation, etc. Single mutation detection methods are difficult to comprehensively reflect the functional status of BRAF signals. Therefore, it is of great significance to develop a multi-gene detection method for evaluating the activation degree of the BRAF pathway from the expression level.

[0004] The existing technologies have the following disadvantages: (1) The defect that traditional mutation detection cannot evaluate the functional activation status; (2) In actual clinical practice, some BRAF wild-type patients with mutations also have pathway activation and are sensitive to drugs; (3) Cannot identify variations in genes such as RAS, MEK, PTEN that co-affect BRAF signals; (4) Can only judge the presence or absence of mutations and cannot predict risks such as patient survival and drug resistance; (5) Some BRAF mutations (such as D594G) are "inhibitory" mutations and do not lead to signal activation. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a characteristic gene set for detecting pan-cancer BRAF mutation activity and its application.

[0006] To achieve the above purpose, the present invention provides the following technical solutions: 1. The present invention provides a method for detecting pan-cancerBRAF The characteristic gene set of mutation activity includes the genes shown in Table 1 ; Table 1 gene NCBI ID number ADAM9 8754 CCND1 595 CLDN1 9076 CTDSPL 10217 DUSP6 1848 EPHA2 1969 ERRFI1 54206 ETV4 2118 ETV5 2119 FZD7 8324 GPRC5A 9052 IGFBP3 3486 ITGA3 3675 MTUS1 57509 PHLDA2 7262 PLEKHH1 57475 S100A16 140576 S100A6 6277 SEMA3A 10371 SLC1A1 6505 SPRY4 81848 TFAP2C 7022 TNFRSF12A 51330 TNFRSF21 27242 TPD52L1 7164

[0007] 2. The present invention provides the aforementioned characteristic gene set for detecting pan-cancer BRAF Application of fluorescent quantitative PCR kit in the preparation of mutation activity.

[0008] 3. The present invention provides a method for detecting pan-cancer BRAF The fluorescent quantitative PCR kit for mutation activity contains specific primers for each gene in the aforementioned characteristic gene set.

[0009] As one of the preferred technical solutions, the kit also contains an internal reference gene GAPDH and ACTB .

[0010] 4. The present invention provides the aforementioned characteristic gene set in BRAF Application in the preparation of mutation-related tumor detection kits.

[0011] 5. The present invention provides a BRAF A mutation-related tumor detection kit comprises specific primers for each gene in the aforementioned characteristic gene set.

[0012] As one of the preferred technical solutions, BRAF Mutation-associated tumors include, but are not limited to, colorectal cancer, melanoma, thyroid cancer, lung adenocarcinoma, and pancreatic cancer. Typing different tumor cohorts based on the expression of the described gene set can help identify subtypes with different BRAF pathway activity and be used to enrich for BRAF mutation signature subtypes.

[0013] 6. The present invention provides an SVM classifier constructed based on the aforementioned characteristic gene set.

[0014] 7. The present invention provides a method for constructing the aforementioned SVM classifier, the specific steps of which are as follows: S1. First, remove genes with highly correlated expression in the aforementioned signature gene set and divide the normalized expression matrix of the BRAF mutation-associated tumor cohort into a training set and a test set in an 8:2 ratio. S2. Gene importance was assessed based on its ability to discriminate samples in the training set. Genes were ranked using a random forest algorithm, and gene subsets of varying sizes were constructed incrementally based on this ranking. S3. For each subset, construct an SVM classifier for evaluation on the test set, and obtain the best performing classifier for typing in the external cohort.

[0015] The beneficial effects of the present invention are: The present invention discloses a characteristic gene set for detecting pan-cancer BRAF mutation activity and its application. Specifically, a set of characteristic gene sets that can specifically reflect BRAF mutation activity is constructed, and a customized fluorescence quantitative PCR kit is developed based on this, which is specifically used to evaluate the transcriptional activity status of the BRAF pathway in tumor tissue samples. The characteristic gene set includes the following genes: ADAM9, CCND1, CLDN1, CTDSPL, DUSP6, EPHA2, ERRFI1, ETV4, ETV5, FZD7, GPRC5A, IGFBP3, ITGA3, MTUS1, PHLDA2, PLEKHH1, S100A16, S100A6, SEMA3A, SLC1A1, SPRY4, TFAP2C, TNFRSF12A, TNFRSF21, TPD52L1 .

[0016] Combined with the SVM machine learning model, the characteristic gene set of the present invention can stably reproduce the BRAF activity classification in multiple tumor types and cohorts, and reveal its potential clinical application value in predicting the response to chemotherapy and targeted therapy. Compared with a single BRAF mutation status, the characteristic gene set score of the present invention provides a more biologically functional relevant way of classifying tumor subtypes, providing strong support for precise treatment decisions. Traditional BRAF mutation detection can only show "whether there is a mutation", while the expression profile evaluation of the characteristic gene set of the present invention can show "whether this pathway is truly activated", and can also predict survival, classification and drug sensitivity, with great clinical significance.

[0017] The advantages of the present invention are as follows: 1. Beyond BRAF gene mutations themselves, accurately reflecting the true activation state of the BRAF signaling pathway Traditional BRAF mutation detection methods can only identify typical mutations (such as V600E, V600K), but a large number of BRAF mutation wild-type patients also have pathway activation phenomena, and the treatment responses are significantly different. By constructing a BRAF25 transcriptional expression profile model, the present invention can identify the "functionally activated" BRAF pathway state, greatly improving the detection sensitivity and clinical predictive power.

[0018] 2. The detection dimension is extended from "gene mutation" to "transcriptional activity", achieving a higher-dimensional pathway judgment The present invention uses 25 core downstream target genes as activity markers to reflect the "transcriptional output" of the entire pathway, which can not only identify the activation caused by mutations, but also capture pathway abnormalities caused by other mechanisms (such as upstream MAPK mutations, feedback regulation, etc.).

[0019] 3. Strong clinical accessibility: Based on the qPCR platform, no expensive high-throughput sequencing equipment is required Compared with the RNA-seq or Nanostring platform, the present invention uses a standard fluorescence quantitative PCR method, and the detection can be completed with a pre-packaged array plate. It is applicable to most hospitals and third-party inspection institutions, with low promotion threshold and low detection cost.

[0020] 4. High flux, short time-consuming, and high degree of standardization Through the 96-well array plate design and internal reference calibration process, the complete sample detection can be completed within 2 hours, and the expression analysis of 25 genes can be completed at one time. The highly standardized process is convenient for large-scale and automated applications and is suitable for high-throughput genotyping screening.

[0021] 5. Support intelligent genotyping output and visualization report The present invention is equipped with an SVM classification model. After inputting the original Ct value, the genotyping result can be automatically output, and the drug response and survival prognosis can be predicted, which is convenient for doctors to interpret the results and make clinical decisions.

[0022] 6. Have the potential for extensive indication expansion In addition to colorectal cancer, the present invention can be applied to various BRAF mutation-related tumors including melanoma, thyroid cancer, lung adenocarcinoma, pancreatic cancer, etc., and shows good generality and consistency in a variety of BRAF mutation related indications.

[0023] The present invention can be widely applied to the following scenarios: 1. Clinical molecular diagnosis For BRAF identifying the functional subtypes of pathways in mutant or wild-type tumor patients; making up for the limitations of traditional BRAF mutation detection methods in prognosis prediction and drug selection; providing a genotyping basis for the use of BRAF inhibitors, MEK inhibitors, immunotherapy or chemotherapy.

[0024] 2. Tumor precision drug guidance Applicable to common BRAF mutation-related tumors such as colorectal cancer, melanoma, lung cancer, thyroid cancer, etc.; can assist in identifying BRAF the potential "functionally active" patient population among wild types; used to predict the sensitivity of tumors to various treatment regimens (including chemotherapy, BRAF inhibitors, etc.).

[0025] 3. Tumor prognosis assessment Based on the gene expression characteristics in the characteristic gene set of the present invention, risk stratification and prognosis prediction are carried out; Provide a basis for formulating postoperative adjuvant treatment strategies and assessing recurrence risks.

[0026] 4. Scientific research application and new drug development It can be used in basic research to explore the activation mechanism of the BRAF pathway and its transcriptional regulatory network, and judge the activation degree of the BRAF pathway; Provide a model reference for screening suitable indication populations for targeted drug R & D companies; It can be used as one of the biomarkers for subject enrollment in clinical trials.

[0027] Other advantages, objectives and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following specification. Brief Description of the Drawings

[0028] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where: Figure 1 . Identification and screening process of the BRAF25 gene set; Figure 2 . Evaluation of the typing ability of the BRAF25 gene set in CCLE cell lines; Figure 3 . Design of a fluorescence quantitative PCR kit based on the BRAF25 gene; Figure 4 . Typing of PCR chip data of the BRAF25 gene set for its own cell lines; Figure 5 . Comparison of overexpression in different tumor cells BRAFV600E and after Vemurafenib BRAF . Changes in mutant activity scores; Figure 6 . Validation analysis of the BRAF25 gene set in the TCGA colon cancer cohort. (A) Distribution of BRAF25 activity scores in different BRAF expression subtypes (BAG-0 to BAG-3), showing that the BAG-3 group has significantly higher activity scores; (B) Mutation frequency in each BAG typing; (C) Precision-recall curve (AUPRC) showing BRAF the discriminative ability of mutant activity scores in distinguishing BRAF mutant samples from wild-type samples; BRAF Figure 7 . Construction of an SVM classifier using the BRAF25 gene set (screening process); Figure 8 . Prognosis analysis of chemotherapy and anti-BRAF targeted therapy using the BRAF25 gene set typing, where A is the overall survival of colorectal cancer, B is the overall survival of BRAF inhibitor treatment, and C is the progression-free survival of BRAF inhibitor treatment. Detailed implementation manners

[0029] The present invention will be further described below in conjunction with the detailed implementation manners.

[0030] 1. BRAF Construction of the mutant activity characteristic gene set The present invention integrates the founder gene sets from the following 8 sources to capture the transcriptional signals related to BRAF pathway activity: BRAFV600K characteristic set: It contains 309 genes, which are significantly up-regulated at the initial stage of tumorigenesis after transgenic expression of plasmids in mouse intestinal epithelial cells. (Riemer P, Sreekumar A, Reinke S, Rad R, Schafer R, Sers C, Blaker H, Herrmann BG, Morkel M: Transgenic expression of oncogenic BRAF induces loss of stem cells in the mouse intestine, which is antagonized by beta-catenin activity. Oncogene 2015, 34(24):3164-3175.) BRAFV600K shBRAF characteristic set: This characteristic set consists of 211 genes whose expression is down-regulated after the knockdown of BRAF induced by doxycycline in A375 cells, and is obtained by comparing with the untreated control group. (Hoeflich KP, Herter S, Tien J, Wong L, Berry L, Chan J, O'Brien C, Modrusan Z, Seshagiri S, Lackner M et al: Antitumor efficacy of the novel RAF inhibitor GDC-0879 is predicted by BRAFV600E mutational status and sustained extracellular signal-regulated kinase / mitogen-activated protein kinase pathway suppression. Cancer Res 2009, 69(7):3042-3051) Three tissue-specific BRAF pathway characteristic sets (melanoma, thyroid cancer, colon cancer): These characteristic sets are based on those driven by different tissue-specific promoters ​BRAFV600E Construction of transgenic mouse models with induction in melanocytes, thyroid cells, and intestinal epithelial cells BRAFV600E of the expression, aiming to clarify BRAFV600E the role in driving cancer development. (Rusinek D, Swierniak M, Chmielik E, Kowal M, Kowalska M, Cyplinska R, Czarniecka A, Piglowski W, Korfanty J, Chekan M et al: BRAFV600E-Associated Gene Expression Profile: Early Changes in the Transcriptome, Based on a Transgenic Mouse Model of Papillary Thyroid Carcinoma. PLoS One 2015, 10(12): e0143688.) Lu H, Liu S, Zhang G, Kwong LN, Zhu Y, Miller JP, Hu Y, Zhong W, Zeng J, Wu L et al: Oncogenic BRAF-Mediated Melanoma Cell Invasion. Cell Rep 2016, 15(9): 2012-2024. Leach JDG, Vlahov N, Tsantoulis P, Ridgway RA, Flanagan DJ, Gilroy K, Sphyris N, Vazquez EG, Vincent DF, Faller WJ et al: Oncogenic BRAF, unrestrained by TGFbeta-receptor signalling, drives right-sided colonic tumorigenesis. Nat Commun 2021, 12(1): 3464.) TCGA-BRAFV600E signature: This signature is based on BRAFConstructed from the top 200 genes with the most significant differential expression between mutant samples and wild-type samples. (Yao K, Zhou E, Cheng C: A B-RafV600E gene signature for melanoma predicts prognosis and reveals sensitivity to targeted therapies. Cancer Med 2022, 11(4):1232-1243.) Vemurafenib treatment feature set: Composed of 167 genes, reflecting the changes at the transcriptome level after treatment with the BRAF inhibitor Vemurafenib. (Parmenter TJ, Kleinschmidt M, Kinross KM, Bond ST, Li J, Kaadige MR, Rao A, Sheppard KE, Hugo W, Pupo GM et al: Response of BRAF-mutant melanoma to BRAF inhibition is mediated by a network of transcriptional regulators of glycolysis. Cancer Discov 2014, 4(4):423-433.) MSigDB feature set: This feature set is the "REACTOME_SIGNALING_BY_MODERATE_KINASE_ACTIVITY_BRAF_MUTANTS" meta-feature set in the Molecular Signatures Database (MSigDB https: / / www.gsea-msigdb.org / gsea / msig), containing a curated set of 45 genes representing the BRAF signaling pathways related to mutants with a certain degree of kinase activity.

[0031] 2. Screening and identifying the core gene set BRAF25 (1) Screening of genes with dominant expression: Among the above 8 feature sets, further screen for key genes with dominant expression in tumor cells. Through Loess regression analysis, the log2 mean expression level of all genes is fitted with the coefficient of variation of expression (logCV), and genes with positive regression residuals and log2 expression means greater than 6 are selected as candidate core feature genes.

[0032] (2) Evaluate the recognition of each feature gene set BRAFAbility of mutant cells: Based on the gene expression data of each feature set, hierarchical clustering was performed on tumor cell lines (using the hclust function and Ward.D2 algorithm in R language), and the clustering results were divided into three groups: high expression, low expression, and unclassified. Subsequently, a chi-square test was used to analyze the significance of each feature set in distinguishing cell populations with high and low BRAF mutation ratios to evaluate its classification ability.

[0033] (3) Construction of the BRAF meta-feature set: Summarize the specific driver genes screened by fold change analysis in the shBRAF feature set, TCGA-BRAFV600E feature set, BRAF pathway feature sets in melanoma and colon cancer, and Vemurafenib treatment feature set to form a BRAF meta-feature set containing 134 genes ( Figure 1 in A), providing a basis for subsequent analysis.

[0034] (4) Candidate BRAF feature set obtained by the RRA method and stepwise screening: We used the Robust Rank Aggregation (RRA) method to perform integrated analysis of multiple groups of data using the RobustRankAggreg package in R language. Based on a theoretical model, this method integrates the expression specificity ranking results of 134 genes in tumor cells in different single-cell RNA sequencing analyses (including thyroid cancer, lung cancer, melanoma, colon cancer) and assigns a significance score to each gene ( Figure 1 in B). Finally, a candidate gene set containing 94 genes was obtained. Subsequently, we performed stepwise increasing screening on this gene set to determine the optimal feature set size. The specific method is as follows: Starting from the gene with the highest significance score ranking, add one by one incrementally to construct multiple subsets, and evaluate its classification performance at each step, that is, determine whether it can effectively distinguish BRAF mutant positive samples (active group) from BRAF wild-type samples (silent group). Finally, the top 25 genes with the highest classification accuracy were selected as the candidate feature set, named BRAF25 ( Figure 1 in C). The classification performance of BRAF25 was verified in the CCLE tumor cell line data. The results showed that among 53 BRAF mutant cell lines, 47 were correctly classified as the BRAF high-activity group, 6 were classified as the unclassified group, and none were misclassified as the low-activity group. This result was statistically significant, indicating that BRAF25 can be used as a reliable feature gene set for judging BRAF pathway activity ( Figure 2 ).

[0035] 3. Design of a fluorescence quantitative PCR kit (PCR chip) based on the BRAF25 gene (1)Gene sequence acquisition: Download the RefSeq mRNA sequences of the target genes from the NCBI or Ensembl databases, including 25 characteristic genes related to the BRAF signaling pathway and 2 commonly used internal reference genes ( GAPDH and ACTB ), providing an accurate template for primer design.

[0036] (2)Primer design: Use the OLIGO 7 software to design specific primers for each gene, ensuring high amplification efficiency, low probability of dimer formation, and meeting the experimental requirements of fluorescence quantitative PCR (Table 2).

[0037] Table 2 human NCBI ID number Forward primer Reverse primer product length ADAM9 8754 AACATAGTTGGGGGTGCTGG TGCACTGTCATGTCTCCGAC 90 CCND1 595 GCTGCGAAGTGGAAACCATC CCTCCTTCTGCACACATTTGAA 135 CLDN1 9076 CCTCCTGGGAGTGATAGCAAT GGCAACTAAAATAGCCAGACCT 145 CTDSPL 10217 GCCCGGCTCTTCAGAGAATC TGTCATCGAACCAGGACTGC 166 DUSP6 1848 AGCTCGACCCCCATGATAGA CGACTCGTATAGCTCCTGCG 153 EPHA2 1969 GCCCCACATGAACTACACCT GGCTCTGTCTGGTTGATGCT 108 ERRFI1 54206 GACCCACCGAAGATTAAGAAGG GGTCTAGGAGGTATGGGAACTCT 156 ETV4 2118 CAGTGCCTTTACTCCAGTGCC CTCAGGAAATTCCGTTGCTCT 125 ETV5 2119 CAGTCAACTTCAAGAGGCTTGG TGCTCATGGCTACAAGACGAC 168 FZD7 8324 CAGACGTGCAAGAGCTATGC ACGATCATGGTCATCAGGTACT 104 GPRC5A 9052 CATGCTCACTCTCCCGATCC AAGAGGAAGAAGCGTGTGGG 168 IGFBP3 3486 AGAGCACAGATACCCAGAACT GGTGATTCAGTGTGTCTTCCATT 93 ITGA3 3675 TCAACCTGGATACCCGATTCC GCTCTGTCTGCCGATGGAG 93 MTUS1 57509 CCGGGGAGAGCTAGTCACT CTGCTGGACGAATGCTTCA 94 PHLDA2 7262 CCATCCTCAAGGTGGACTGC GATCTCCTTGTGGTCGGTGG 80 PLEKHH1 57475 CTAGTTCCAGCACGGTCCATT CTGAGCAGTAACCAGACCCTC 155 S100A16 140576 ATGTCAGACTGCTACACGGAG GTTCTTGACCAGGCTGTACTTAG 93 S100A6 6277 GAACAAGGACCAGGAGGTGA CCCTTGAGGGCTTCATTGTA 84 SEMA3A 10371 CTATCTTCCGAACTCTTGGGCA CTTTGGATCATTGAGCCACCT 77 SLC1A1 6505 GAAACCGGCGAGGAAAGGAT GGTAATGCCTAGCACCACCG 94 SPRY4 81848 TAGTCTGCCCACTCCCTACC GCCCCTGCATTGTCTGTTTG 143 TFAP2C 7022 TCAGTCCCTGGAAGATTGTCG CCAGTAACGAGGCATTTAAGCA 112 TNFRSF12A 51330 CTCTGAGCCTGACCTTCGTG GTCTCCTCTATGGGGGTGGT 97 TNFRSF21 27242 ATTGGCACATACCGCCATGTT GGCTTGTGTTGGTACAATGCTC 103 TPD52L1 7164 TTTTGTCAGCGAAAGAAAGGCA GCAGTGGTAGTCTGCATGTCA 118 GAPDH 2597 CAATGACCCCTTCATTGACC GACAAGCTTCCCGTTCTCAG 106 ACTB 60 AGACCTGTACGCCAACACAG TTCTGCATCCTGTCGGCAAT 72 (3)Composition of the fluorescence quantitative PCR kit: The kit includes: 2× qPCR Master Mix, custom primer pairs, ROX reference dye, RNase-free water, and a standard operation manual, enabling one-key high-throughput assessment of BRAF pathway activity.

[0038] The layout of the 96-well PCR chip plate has 3 sets of technical replicates per plate (columns 1–4, 5–8, 9–12), and each set completely covers 27 genes (25 genes related to the BRAF signaling pathway + GAPDH + ACTB ), arranged in the order of rows (A–H). Each well has triple technical replicates to improve the statistical robustness and repeatability of the data ( Figure 3 ).

[0039] (4)Operation procedure: (4-1)Sample preparation: Extract total RNA from tissues / cells using Trizol or a commercial RNA extraction kit. The A260 / A280 ratio should be 1.8 - 2.0, and the total RNA concentration should be greater than 50 ng / µL.

[0040] (4-2)cDNA synthesis and reaction system configuration: Transcribe 1 µg of total RNA into cDNA using a reverse transcription kit. After preparing a 900µL mixture of cDNA and mix and mixing well, add the prepared mixture to the plate at 9µL per well: 5 µL of 2× qPCR Master Mix, 1 µL of cDNA template, 2.8µL of nuclease-free water, and 0.2µL of ROX reference Dye.

[0041] (4-3) was performed on a Bio-Rad CFX96 system with the following program settings: initial denaturation at 95°C for 10 minutes; followed by 40 amplification cycles, each cycle including denaturation at 95°C for 5 seconds, annealing / extension at 60°C for 30 seconds, and real-time fluorescence signal acquisition simultaneously.

[0042] (5) Data analysis: The ΔCt method was used to evaluate the expression level: ΔCt = Ct (target gene) − Ct (reference gene GAPDH or ACTB ); Subsequently, the relative expression was calculated and log2-transformed. Finally, the average log2 expression value of 25 BRAF pathway genes was taken as the " BRAF mutation activity score" for each sample.

[0043] 4. Functional verification of the BRAF25 signature gene set (1) Verification using a cell line PCR array: More than 10 tumor cell lines derived from different organs (covering BRAF mutant and wild-type) were selected, and the BRAF25 gene expression profile was detected using a customized PCR array. Hierarchical clustering analysis was performed using the hclust function and Ward.D2 clustering algorithm in R language. The results showed that this gene set could effectively enrich and distinguish BRAF mutant cell lines ( Figure 4 ).

[0044] (2) BRAFV600E Verification by overexpression and BRAF inhibitor treatment: In BRAF mutant cell lines, gene overexpression and treatment with the BRAF inhibitor Vemurafenib were performed respectively, and a customized PCR array was used for detection. The results showed that the BRAFV600E mutation activity score of the treatment group decreased significantly, while that of the overexpression group increased significantly, both with statistical significance ( BRAF ), further verifying the reliability and sensitivity of the BRAF25 scoring system in reflecting BRAF signal activity. Figure 5

[0045] (3) Verification using large sample data from TCGA colon cancer: To further evaluate the applicability of the BRAF25 signature gene set in clinical samples, the applicant performed hierarchical clustering analysis on the normalized expression matrix of BRAF25 based on the transcriptome data in the TCGA colon cancer (COAD) cohort, using the Ward.D2 clustering method. According to the average expression level of the BRAF25 gene set in each clustering category, all samples were divided into four expression subtypes, named BAG-0 to BAG-3 in sequence, where BAG-3 represents the BRAF pathway high-activity group ( Figure 6 in A). It should be noted that the BRAF in the BAG-3 groupThe mutation frequency increased significantly, reaching 37.8%, which was significantly higher than that of other groups ( Figure 6 in B). Further, we divided the TCGA samples into high-expression and low-expression groups according to the BRAF25 activity score to evaluate their ability to distinguish BRAF mutant 5 from wild-type samples. Considering that BRAF-mutated positive samples only accounted for about 10% in colorectal cancer, and the ratio of positive to negative samples was severely uneven, we used the PRROC package to draw the precision-recall curve (Precision-Recall Curve) and calculated the area under the curve (AUPRC) to obtain a more accurate evaluation index. The analysis results showed that although BRAF the mutation rate was low, the BRAF25 score performed well in distinguishing mutant from non-mutant samples, and its AUPRC reached 0.25 ( Figure 6 in C), which was better than the random prediction level, verifying that this gene set had a certain ability to identify the BRAF mutation activity status.

[0046] 5. Construct a classifier using machine learning methods to expand the application value of the BRAF25 gene set (1) Construction and verification of the SVM classifier: To enhance the application of the BRAF25 gene set in large cohort data, we first removed genes with highly correlated expression in the BRAF25 feature set and divided the standardized expression matrix in the TCGA colon cancer (COAD) cohort into a training set and a test set at a ratio of 8:2. Subsequently, the random forest algorithm was used to rank the remaining genes, and the gene importance was evaluated based on their ability to distinguish samples in the training set, and gene subsets of different scales were constructed accordingly. For each subset, a support vector machine (Support Vector Machine, SVM) classification model was gradually constructed to identify the BRAF high-activity expression subtype BAG-3. The model evaluation results showed that the classifier using all 25 genes achieved the best performance in terms of sensitivity and specificity ( Figure 7 ), and this model could stably reproduce the BAG-3 subtype and significantly enrich BRAF-mutated tumors, indicating its good robustness and generalization ability.

[0047] (2) Application of the SVM classifier model: To further evaluate the clinical application potential of the BRAF25 classifier, we applied this SVM model to Laetitia's colorectal cancer cohort. The patients in this cohort received standard first-line adjuvant chemotherapy and had complete prognostic follow-up information. After model classification, it was found that the overall survival (OS) of patients with the BAG-3 subtype was significantly shorter than that of other subtypes, especially the difference between the BAG-0 and BAG-2 groups was statistically significant ( Figure 8In A). In the multivariate Cox regression analysis incorporating BRAF mutation status and BAG subtypes, BAG classification was confirmed as an independent poor prognostic factor.

[0048] Further evaluate whether the BRAF25 classifier can be used to predict the response to targeted therapy. In three independent external cohorts of melanoma patients treated with BRAF inhibitors, the BAG subtype distribution was successfully reproduced. The results showed that although most patients with BAG-3 subtype melanoma showed a good initial response to BRAF inhibitor treatment, the overall survival (OS, Figure 8 in B) and progression-free survival (PFS, Figure 8 in C) were significantly worse than those of other subtypes, suggesting that this subtype may be associated with the mechanism of acquired resistance.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the present technical solution, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A characteristic gene set for detecting pan-cancer BRAF mutation activity, characterized in that comprises the genes shown in Table 1; Table 1 。 2. Use of the characteristic gene set according to claim 1 in the preparation of a fluorescence quantitative PCR kit for detecting pan-cancer BRAF mutation activity.

3. A fluorescence quantitative PCR kit for detecting pan-cancer BRAF mutation activity, characterized in that specific primers for each gene in the characteristic gene set described in claim 1.

4. Use of the characteristic gene set described in claim 1 in the preparation of a detection kit for BRAF mutation-related tumors.

5. A BRAF tumor detection kit related to mutation, characterized in that specific primers for each gene in the characteristic gene set described in claim 1.

6. The kit according to claim 5, wherein The described BRAF Mutation-related tumors include, but are not limited to: colorectal cancer, melanoma, thyroid cancer, lung adenocarcinoma, pancreatic cancer.

7. An SVM classifier constructed based on the characteristic gene set described in claim 1.

8. The method for constructing the SVM classifier according to claim 7, wherein The specific steps are as follows: S1. First, remove the genes with high expression correlation in the aforementioned characteristic genes, and BRAF The normalized expression matrix in the mutation-related tumor cohort was divided into training and test sets in an 8:2 ratio; S2. Evaluate the gene importance according to its ability to distinguish samples in the training set, rank the genes using the random forest algorithm, and gradually incrementally construct gene subsets of different scales accordingly; S3. For each subset, construct an SVM classifier for the evaluation of the test set, and obtain the classifier with the best performance for the typing of the external cohort.

Images