A method for constructing a POLE mutation discrimination model based on second-generation sequencing results and application thereof
By constructing a POLE mutation discrimination model based on next-generation sequencing results and using specific thresholds and scoring criteria to filter false positives, the problem of high false positives in existing methods is solved, achieving accurate discrimination of POLE mutations and improving the true positive detection rate. This model is applicable to various tumor types, including endometrial cancer.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN HAPLOX BIOTECH
- Filing Date
- 2023-03-02
- Publication Date
- 2026-07-21
AI Technical Summary
Existing POLE mutation discrimination methods cannot accurately determine whether a sample is positive in the absence of clear hyperamplification markers, resulting in a high false positive rate. Furthermore, they lack multi-dimensional judgment thresholds and procedures.
A POLE mutation discrimination model was constructed using next-generation sequencing results, including data preprocessing, sequence alignment, and mutation filtering. Specific thresholds and scoring criteria were set to filter false positive samples. Mutscan or Varscan tools were used to analyze mutation information. Combined with the COSMIC database and tumor burden data, six integer scoring criteria were set to determine pathogenic POLE mutations.
It improves the true positive detection rate of POLE mutations, can accurately distinguish between pathogenic and non-pathogenic POLE mutations, reduces the false positive rate, and is applicable to the identification of POLE mutations in various tumor types.
Smart Images
Figure CN116364183B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of bioinformatics analysis technology, and in particular to a method for constructing a POLE mutation discrimination model based on next-generation sequencing results and its application. Background Technology
[0002] Endometrial carcinoma (EC) is a common gynecological cancer, a group of epithelial malignant tumors that occur in the uterine lining. The POLE hypermutation type of endometrial carcinoma is defined as a tumor caused by a somatic mutation in the DNA polymerase ε (POLE) exonuclease domain. POLE-type endometrial carcinoma is generally more sensitive to chemotherapy and immunotherapy, resulting in good clinical treatment outcomes.
[0003] High tumor burden (TMB>100) and COSMIC-characteristic POLE somatic mutations can serve as important biomarkers for the diagnosis of POLE hyperamplification (mutant) tumors. Furthermore, according to literature reports, the distribution of different types of base substitutions, the ratio of point mutations, and insertion / deletion mutations are also associated with the occurrence of POLE hyperamplification tumors. However, current mainstream diagnostic methods are based on known positive POLE mutation sites (P286R, V411L, S297F, A456P, S459F, M444K). Existing methods have limitations in clinical diagnosis; in the absence of a clearly defined POLE mutation as a hyperamplification marker, it is impossible to determine whether a sample is positive. Moreover, given a known high tumor burden and specific base substitution distribution, there are no statistically significant parameters validated with large sample data to evaluate the positive probability of a sample, nor are there quantifiable evaluation criteria. Therefore, existing methods are prone to false positives.
[0004] Currently, there is no more standardized and multi-dimensional threshold and procedure for determining whether a tumor belongs to the POLE mutation. Summary of the Invention
[0005] The main objective of this invention is to provide a method for constructing a POLE mutation discrimination model based on next-generation sequencing results and its application, aiming to improve the detection rate of true positive POLE mutations.
[0006] The present invention also aims to provide a POLE mutation discrimination model, which is constructed using the aforementioned construction method.
[0007] Another objective of this invention is to provide a computer-readable storage medium.
[0008] POLE mutations (somatic mutations in the DNA polymerase ε[POLE] exonuclease domain) occur in various tumor diseases. POLE mutations, especially pathogenic POLE mutations, are often used as a basis for diagnosing tumor diseases. Current methods for identifying POLE mutations require the identification of POLE mutations as hyperamplification markers, but their false positive rate is relatively high. This invention, by setting specific thresholds, first excludes a portion of false positive samples, and then judges the pathogenicity and potential pathogenicity of POLE mutations according to a set scoring standard, thereby improving the detection rate of true positives for POLE mutations.
[0009] To achieve the above objectives, this invention provides a method for constructing a POLE mutation discrimination model based on next-generation sequencing results, comprising the following steps: Preprocessing of S1 and next-generation sequencing data The second-generation sequencing data of the test samples and control samples were pre-filtered to obtain pileup files. The pileup file is a summary of the results by comparing the overlapping reads at each position in the BAM file, which can be used to determine the probability of mutations at each site. The pre-filtering includes quality control, comparison, and deduplication; The test sample is cell-free DNA (ffpedna) from the tissue of a tumor patient or cell-free DNA (cfdna) from the blood, and the control sample is saliva DNA (sladna) or genomic DNA (gdna) from the blood of a tumor patient. The quality control includes quality assessment, removal of connector sequences, removal of low-quality sequences, and removal of duplicate sequences.
[0010] The alignment process involves comparing quality-controlled reads with a human reference genome to obtain a BAM file. Each quality-controlled read is then aligned with the reference genome and pasted back to its optimal position on the genome. Adapter sequences are removed before alignment. Reads are aligned to the GRCh37 (hg19) version of the human genome using a paired-end alignment method. The aligned reads are then sorted by chromosomal position and converted into a binary BAM file.
[0011] The deduplication process involves removing PCR duplicates, correcting sequencing errors, and generating a pileup file for detection.
[0012] S2, sequence alignment, variant interpretation, and mutation filtering The pileup file obtained in step S1 was analyzed using a mutation detection tool. The mutation information of cancer-related genes in the data was analyzed, the site location and specific base changes of point mutations and insertion / deletion mutations were annotated, POLE mutations were marked, and the mutation number and maximum mutation abundance of all mutations in the sample were obtained. Based on the number of mutations, false positives are filtered out, and DNA samples corresponding to sequencing data with a mutation number of less than or equal to 50 are filtered out. When analyzing tumor mutations, two sets of genomic data need to be collected: tumor genome (tissue DNA [ffpedna] or blood cell-free DNA [cfdna]) and control genome (saliva DNA [sladna] or blood genomic DNA [gdna]) for joint analysis (both genomes need to undergo quality control, comparison, and deduplication). If a mutation exists in both the tumor genome and the control genome, it is a germline mutation; if the mutation exists only in the tumor genome, it is a somatic mutation. The mutations discussed and analyzed in this invention are all somatic mutations, because only somatic mutations are related to the development of cancer.
[0013] The mutation detection tool is either Mutscan or Varscan. Mutscan or Varscan can identify mutations at each site based on the deduplicated pileup file, and the mutation types include SNV and Indel.
[0014] S3. Analysis of mutation filtering data and POLE mutation identification Based on the mutation data filtered in step S2 and the set threshold, the pathogenicity of the POLE mutation is determined.
[0015] Preferably, the analysis of the mutation filtering data specifically involves: analyzing the data filtered in step S2 to obtain the number of mutations where base C is mutated into base A (C>A), the number of mutations where base T is mutated into base G (T>G), the number of mutations where base C is mutated into base G (C>G), the total number of point mutations, the total number of insertion / deletion mutations, and the total number of mutations; Based on the POLE mutations annotated in step S2, obtain the number of times the POLE mutations appear in the corresponding tumors of the samples from the COSMIC database; COSMIC is an online database of somatic mutations found in human cancers. Tumor burden (TMB) data is obtained based on the BED file size of the sequencing region of cancer-related genes (i.e., the BED format file covering the region of cancer-related genes); Tumor burden (TMB) = total number of mutations / sequencing region BED file size, unit: muts / mb; Preferably, the POLE mutation discrimination is specifically as follows: based on the analysis results of the mutation filtering data, six integer scoring criteria are set: 1 point is awarded if the percentage of C-base mutations with A (C>A) is greater than 20% of the total number of point mutations, otherwise no point is awarded; 1 point is awarded if the percentage of T-base mutations with G (T>G) is greater than 4% of the total number of point mutations, otherwise no point is awarded; 1 point is awarded if the percentage of C-base mutations with G (C>G) is less than 0.6% of the total number of point mutations (SNVs), otherwise no point is awarded; 1 point is awarded if the percentage of insertion / deletion sites (Indels) is less than 5% of the total number of mutations, otherwise no point is awarded; 1 point is awarded if the tumor burden (TMB) is greater than 100 muts / mb, otherwise no point is awarded; 1 point is awarded if the number of occurrences of the POLE mutation in the tumor is greater than 1, otherwise no point is awarded. According to the scoring criteria, the POLE mutation score of the sample is statistically analyzed. If the total score is greater than or equal to 4, it is judged to be a pathogenic POLE mutation.
[0016] Preferably, the POLE mutation score of the sample is calculated according to the scoring standard. If the total score is equal to 3, it is judged as an unknown pathogenic POLE mutation; if the total score is less than 3, it is judged as a non-pathogenic POLE mutation.
[0017] Preferably, the cancer-related genes are 680 cancer-related genes.
[0018] Preferably, the BED file size of the sequencing region of the cancer-related genes is the BED file size of the sequencing region of 680 genes (i.e., a BED format file covering the region of 680 genes).
[0019] Preferably, false positives are filtered out based on the number of mutations and the maximum mutation abundance; the sample to be tested is DNA from the tissue of a tumor patient (ffpedna), and samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 5% are filtered out. More preferably, false positives are filtered out based on the obtained number of mutations and maximum mutation abundance; the sample to be tested is cell-free DNA (cfdna) from the blood of a tumor patient, and samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 1% are filtered out. Preferably, the tumor is endometrial cancer, lung cancer, colorectal cancer, melanoma, glioma, ovarian cancer, or urothelial carcinoma.
[0020] More preferably, the tumor is endometrial cancer, lung cancer, or colorectal cancer.
[0021] More preferably, the tumor is endometrial cancer.
[0022] More preferably, the endometrial cancer is POLE-type endometrial cancer.
[0023] To achieve the above objectives, this invention provides a POLE mutation discrimination model, constructed using the aforementioned method. This POLE mutation discrimination model can more accurately identify POLE mutations, distinguish between pathogenic and non-pathogenic POLE mutations, and discover potentially pathogenic POLE mutations.
[0024] To achieve the above objectives, the present invention also aims to provide a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of the above-described construction method.
[0025] In the absence of a clearly defined POLE mutation as a hyperamplification marker, this invention filters out false positives caused by insufficient mutation sites by setting a certain number of mutations and a maximum mutation abundance threshold. Furthermore, it incorporates a POLE mutation discrimination score to assist in the identification of POLE mutations and improve the detection rate of true positives. Simultaneously, the mutation detection scope of this invention covers 680 cancer-related genes; this broad coverage ensures that the given scoring criteria are closer to the predicted assumptions, thus making it more reliable. Attached Figure Description
[0026] Figure 1 This is a flowchart illustrating the method for constructing a POLE mutation discrimination model based on second-generation sequencing results, as described in an embodiment of the present invention.
[0027] Figure 2 This is a schematic diagram of the computer-readable storage medium structure involved in the embodiments of the present invention.
[0028] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0029] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0030] This invention filters out false positives caused by insufficient mutation sites by setting a threshold for the number of mutations and the maximum mutation abundance. It scores POLE mutations in the sequencing data of true positive samples, which can not only distinguish between pathogenic and non-pathogenic POLE mutations, but also discover potential POLE mutations, thus improving the detection rate of true positive POLE mutations.
[0031] The mutation detection scope of this invention covers 680 cancer-related genes. The large detection coverage makes the given scoring criteria closer to the predicted assumptions, and therefore more reliable.
[0032] This invention provides a method for constructing a POLE mutation discrimination model based on next-generation sequencing results, comprising the following steps: Preprocessing of S1 and next-generation sequencing data The second-generation sequencing data of the test samples and control samples were pre-filtered to obtain pileup files. The pre-filtering includes quality control, comparison, and deduplication; The test sample is cell-free DNA (ffpedna) from the tissue of a tumor patient or cell-free DNA (cfdna) from the blood, and the control sample is saliva DNA (sladna) or genomic DNA (gdna) from the blood of a tumor patient. The quality control includes quality assessment, deconnector sequence removal, low-quality sequence removal, and duplicate sequence removal. The alignment process involves comparing the quality-controlled reads with the human reference genome to obtain a BAM file; each quality-controlled read is then aligned with the reference genome and pasted back to its optimal position on the genome; adapter sequences are removed before alignment; the reads are aligned to the GRCh37 (hg19) version of the human genome in a paired-end mode; the aligned reads are then sorted by chromosome position and converted into a binary BAM file. The deduplication process involves removing PCR duplicates, correcting sequencing errors, and generating a pileup file for detection.
[0033] S2, sequence alignment, variant interpretation, and mutation filtering The pileup file obtained in step S1 was analyzed using a mutation detection tool. The mutation information of cancer-related genes in the data was analyzed, the site location and specific base changes of point mutations (SNVs) and insertion / deletion (Indel) mutations were annotated, POLE mutations were labeled, and the mutation number and maximum mutation abundance of all mutations in the sample were obtained. The mutation detection tool is either Mutscan or Varscan; Based on the number of mutations, false positives are filtered out, and DNA samples corresponding to sequencing data with a mutation number of less than or equal to 50 are filtered out. S3. Analysis of mutation filtering data and POLE mutation identification Based on the mutation data filtered in step S2 and the set threshold, the pathogenicity of the POLE mutation is determined.
[0034] Preferably, the analysis of the mutation filtering data specifically involves: analyzing the data filtered in step S2 to obtain the number of mutations where base C is mutated into base A (C>A), the number of mutations where base T is mutated into base G (T>G), the number of mutations where base C is mutated into base G (C>G), the total number of point mutations, the total number of insertion / deletion mutations, and the total number of mutations; Based on the POLE mutations annotated in step S2, obtain the number of occurrences of POLE mutations in the corresponding tumors of the samples from the COSMIC database. Tumor burden (TMB) data is obtained based on the BED file size of the sequencing region of cancer-related genes; Tumor burden (TMB) = total number of mutations / BED file size of sequencing region, unit: muts / mb.
[0035] Preferably, the POLE mutation discrimination is specifically as follows: based on the analysis results of the mutation filtering data, six integer scoring criteria are set: 1 point is awarded if the percentage of C-base mutations with A (C>A) is greater than 20% of the total number of point mutations, otherwise no point is awarded; 1 point is awarded if the percentage of T-base mutations with G (T>G) is greater than 4% of the total number of point mutations, otherwise no point is awarded; 1 point is awarded if the percentage of C-base mutations with G (C>G) is less than 0.6% of the total number of point mutations, otherwise no point is awarded; 1 point is awarded if the percentage of insertion / deletion sites (Indels) is less than 5% of the total number of mutations, otherwise no point is awarded; 1 point is awarded if the tumor burden (TMB) is greater than 100 muts / mb, otherwise no point is awarded; 1 point is awarded if the number of occurrences of the POLE mutation in the tumor is greater than 1, otherwise no point is awarded. According to the scoring criteria, the POLE mutation score of the sample is statistically analyzed. If the total score is greater than or equal to 4, it is judged to be a pathogenic POLE mutation.
[0036] Preferably, the POLE mutation score of the sample is calculated according to the scoring standard. If the total score is equal to 3, it is judged as an unknown pathogenic POLE mutation; if the total score is less than 3, it is judged as a non-pathogenic POLE mutation.
[0037] Preferably, the cancer-related genes are 680 cancer-related genes.
[0038] Preferably, the BED file size of the sequencing region of the cancer-related genes is the BED file size of the sequencing region of 680 genes (i.e., a BED format file covering the region of 680 genes).
[0039] When analyzing tumor mutations, two sets of genomic data need to be collected: tumor genome (tissue DNA [ffpedna] or blood cell-free DNA [cfdna]) and control genome (saliva DNA [sladna] or blood genomic DNA [gdna]) for joint analysis (both genomes need to undergo quality control, comparison, and deduplication). If a mutation exists in both the tumor genome and the control genome, it is a germline mutation; if the mutation exists only in the tumor genome, it is a somatic mutation. The mutations discussed and analyzed in this invention are all somatic mutations, because only somatic mutations are related to the development of cancer.
[0040] Mutation detection tools such as Mutscan or Varscan can identify mutations at each site based on the deduplicated pileup file. The mutation types include point mutations (SNVs) and insertion / deletion (Indel) mutations.
[0041] Preferably, the present invention filters out false positives based on the number of mutations and the maximum mutation abundance; the sample to be tested is DNA from the tissue of a tumor patient (ffpedna), and samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 5% are filtered out. More preferably, the present invention filters out false positives based on the obtained number of mutations and maximum mutation abundance; the sample to be tested is cell-free DNA (cfdna) from the blood of a tumor patient, and samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 1% are filtered out. Preferably, the tumor is endometrial cancer, lung cancer, colorectal cancer, melanoma, glioma, ovarian cancer, or urothelial carcinoma.
[0042] More preferably, the tumor is endometrial cancer, lung cancer, or colorectal cancer.
[0043] More preferably, the tumor is endometrial cancer.
[0044] More preferably, the endometrial cancer is POLE-type endometrial cancer.
[0045] A POLE mutation discrimination model is constructed using the method for constructing a POLE mutation discrimination model based on next-generation sequencing results according to the present invention. This POLE mutation discrimination model can more accurately identify POLE mutations, distinguish between pathogenic and non-pathogenic POLE mutations, and discover potentially pathogenic POLE mutations. The model is applicable to the discrimination of POLE mutations in various tumors, including endometrial cancer, lung cancer, colorectal cancer, melanoma, glioma, ovarian cancer, and urothelial carcinoma.
[0046] The present invention also provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the construction method described in the present invention.
[0047] Example 1: A method for constructing a POLE mutation discriminant model based on next-generation sequencing results and the POLE mutation discriminant model itself. A method for constructing a POLE mutation discrimination model based on next-generation sequencing results, the flowchart of which is shown below. Figure 1 As shown.
[0048] The construction method includes the following steps: Preprocessing of S1 and next-generation sequencing data The second-generation sequencing data of the test samples and control samples were pre-filtered to obtain pileup files. The test sample is cell-free DNA (ffpedna) from the tissue of a tumor patient or cell-free DNA (cfdna) from the blood, and the control sample is saliva DNA (sladna) or genomic DNA (gdna) from the blood of a tumor patient. The pre-filtering includes quality control, comparison, and deduplication; The quality control includes quality assessment, deconnector sequence removal, low-quality sequence removal, and duplicate sequence removal. The alignment process involves comparing the quality-controlled reads with the human reference genome to obtain a BAM file; each quality-controlled read is then aligned with the reference genome and pasted back to its optimal position on the genome; adapter sequences are removed before alignment; the reads are aligned to the GRCh37 (hg19) version of the human genome in a paired-end mode; the aligned reads are then sorted by chromosome position and converted into a binary BAM file. The deduplication process involves removing PCR duplicates, correcting sequencing errors, and generating a pileup file for detection.
[0049] S2, sequence alignment, variant interpretation, and mutation filtering The pileup file obtained in step S1 was analyzed using the mutation detection tools Mutscan or Varscan. The analysis included mutation information of 680 cancer-related genes in the data, annotation of the site location and specific base changes of SNV and Indel mutations, annotation of POLE mutations, and acquisition of the mutation number and maximum mutation abundance of all mutations in the sample. False positives are filtered out based on the number of mutations, and DNA samples corresponding to sequencing data with a mutation number of less than or equal to 50 are filtered out. S3. Analysis of mutation filtering data and POLE mutation identification Based on the mutation data filtered in step S2 and the set threshold, the pathogenicity of the POLE mutation is determined.
[0050] The analysis of the mutation filtering data specifically involves: analyzing the data filtered in step S2 to obtain the number of mutations where base C is mutated into base A (C>A), the number of mutations where base T is mutated into base G (T>G), the number of mutations where base C is mutated into base G (C>G), the total number of point mutations, the total number of insertion / deletion mutations, and the total number of mutations. Based on the POLE mutations annotated in step S2, obtain the number of occurrences of POLE mutations in the corresponding tumors of the samples from the COSMIC database. Tumor burden (TMB) data was obtained based on the BED file size of the sequencing regions of 680 genes; Tumor burden (TMB) = total number of mutations / BED file size of sequencing regions, unit: muts / mb; The POLE mutation discrimination is specifically as follows: Based on the analysis results of the mutation filtering data, six integer scoring criteria are set: A C-type mutation with an A base (C>A) accounting for more than 20% of the total number of point mutations is scored as 1 point, otherwise no score is scored; a T-type mutation with a G base (T>G) accounting for more than 4% of the total number of point mutations is scored as 1 point, otherwise no score is scored; a C-type mutation with a G base (C>G) accounting for less than 0.6% of the total number of point mutations is scored as 1 point, otherwise no score is scored; an insertion / deletion site (Indel) accounting for less than 5% of the total number of mutations is scored as 1 point, otherwise no score is scored; a tumor burden (TMB) greater than 100 muts / mb is scored as 1 point, otherwise no score is scored; and the occurrence of the POLE mutation in the tumor more than once is scored as 1 point, otherwise no score is scored. According to the scoring criteria, the POLE mutation score of the sample is statistically analyzed. If the total score is greater than or equal to 4, it is judged to be a pathogenic POLE mutation. A total score of 3 indicates an unknown pathogenic POLE mutation. A total score of less than 3 indicates a non-pathogenic POLE mutation.
[0051] The POLE mutation discrimination model can be constructed using the above construction method.
[0052] Information on the 680 genes is shown in Table 1 below.
[0053] Table 1
[0054] Example 2: POLE mutation discriminant model test The POLE mutation discrimination model from Example 1 was used for testing. A total of 410 samples potentially exhibiting POLE hypermutation pathogenicity were tested from the endometrial cancer patient database of Shenzhen Haplos Biotechnology Co., Ltd. The model of this invention was used for false positive filtering and POLE mutation score calculation.
[0055] According to the model filtering conditions of this invention, 160 negative samples were filtered out from 410 samples that may have POLE hypermutations. Of the 160 samples, 158 were scored 3 or below according to the scoring standard (out of 6); the remaining 2 were scored 4 or above according to the scoring standard (out of 6). In summary, this filtering removed 158 true negatives and 2 false positives.
[0056] The pathogenicity of the 250 filtered samples was determined based on calculated scores and compared with actual clinical diagnoses. A four-fold table was then created, and the sensitivity and specificity of the diagnostic test evaluation indicators were calculated based on the data from the four-fold table. The calculation formula is as follows: Sensitivity TP / (FN+TP) × 100%; Specificity calculation formula: TN / (TN+FP)×100%.
[0057] The resulting four-cell table is shown in Table 2 below. Table 2
[0058] (Note: Standard set - True means positive, Standard set - False means negative, Test set - True means score greater than or equal to 4, Test set - False means score less than 4) Based on the data in Table 2, the sensitivity of the model of this invention is 96%, and the specificity is 99.5%. That is, the model of this invention has high sensitivity and strong specificity.
[0059] In Table 2, one site (A1814T) that is false in the standard set but has a POLE mutation score of 4 may have unreported potential pathogenicity outside the common detection range (exons 9, 11, 13, 14).
[0060] Comparative Example 1 The second-generation sequencing data were analyzed using a comparative model to calculate the POLE mutation score.
[0061] The only difference between the comparative model and the present invention is that the latter lacks false positive filtering.
[0062] The comparative model construction method includes the following steps: Preprocessing of S1 and next-generation sequencing data The second-generation sequencing data of the test samples and control samples were pre-filtered to obtain pileup files. The test sample is cell-free DNA (ffpedna) from the tissue of a tumor patient or cell-free DNA (cfdna) from the blood, and the control sample is saliva DNA (sladna) or genomic DNA (gdna) from the blood of a tumor patient. The pre-filtering includes quality control, comparison, and deduplication; The quality control includes quality assessment, deconnector sequence removal, low-quality sequence removal, and duplicate sequence removal. The alignment process involves comparing the quality-controlled reads with the human reference genome to obtain a BAM file; each quality-controlled read is then aligned with the reference genome and pasted back to its optimal position on the genome; adapter sequences are removed before alignment; the reads are aligned to the GRCh37 (hg19) version of the human genome in a paired-end mode; the aligned reads are then sorted by chromosome position and converted into a binary BAM file. The deduplication process involves removing PCR duplicates, correcting sequencing errors, and generating a pileup file for detection.
[0063] S2, sequence alignment, variant interpretation, and mutation filtering The pileup file obtained in step S1 was analyzed using the mutation detection tools Mutscan or Varscan. The analysis included mutation information of 680 cancer-related genes, annotation of SNV and Indel mutation sites and specific base changes, annotation of POLE mutations, and acquisition of the number of C mutations to A (C>A), T mutations to G (T>G), C mutations to G (C>G), total number of point mutations, total number of insertion / deletion mutations, and total number of mutations. Based on the POLE mutations annotated in step S2, obtain the number of occurrences of POLE mutations in the corresponding tumors of the samples from the COSMIC database. Tumor burden (TMB) data is obtained based on the BED file size of the sequencing region of 680 genes (i.e., a BED format file covering the region of 680 genes); Tumor burden (TMB) = total number of mutations / sequencing region BED file size, unit: muts / mb; S3 and POLE mutation discrimination Six integer scoring criteria are set: In step S2, if the percentage of C-base mutations that are converted to A-base mutations (C>A) is greater than 20% of the total number of point mutations, 1 point is awarded; otherwise, no points are awarded. If the percentage of T-base mutations that are converted to G-base mutations (T>G) is greater than 4% of the total number of point mutations, 1 point is awarded; otherwise, no points are awarded. If the percentage of C-base mutations that are converted to G-base mutations (C>G) is less than 0.6% of the total number of point mutations, 1 point is awarded; otherwise, no points are awarded. If the percentage of insertion / deletion sites (indels) in the total number of mutations is less than 5%, 1 point is awarded; otherwise, no points are awarded. If the tumor burden (TMB) is greater than 100 muts / mb, 1 point is awarded; otherwise, no points are awarded. If the number of occurrences of POLE mutations in the tumor is greater than 1, 1 point is awarded; otherwise, no points are awarded. According to the scoring criteria, the POLE mutation scores of the sample are statistically analyzed. If the total score is greater than or equal to 4, it is judged as a pathogenic POLE mutation; if the total score is equal to 3, it is judged as an unknown pathogenic POLE mutation; and if the total score is less than 3, it is judged as a non-pathogenic POLE mutation.
[0064] Using Comparative Model 1, without filtering, there might be samples containing only two or a few mutations: C>A, T>G: C>A, 1 mutation; T>G, 1 mutation; C>G, 0 mutations; indels, 0 mutations; these would be scored out of four and classified as positive. However, considering the high tumor burden (numerous mutations) characteristic of POLE hyperamplification, this sample is clearly negative. Therefore, in actual discrimination, a filtering step should be added to identify such false positives.
[0065] Experimental testing using Comparative Model 1 revealed two false positives in the unfiltered results. Therefore, compared to Example 1, Comparative Model 1 exhibits an increased false positive rate.
[0066] Comparative Example 2 The second-generation sequencing data were analyzed using comparative model 2 to calculate the POLE mutation score.
[0067] The only difference between Comparative Model 2 and the present invention is that the conditions for filtering false positives are adjusted. False positives are filtered out based on the number of mutations, and samples corresponding to sequencing data with a mutation number of less than or equal to 25 are filtered out.
[0068] According to the theoretical model, if we assume that the background frequencies of mutation rates for all 12 SNV types are 1:1, it would take 17 samples to detect at least one specific point under the condition of 0.05 significance, while it would take at least 30 samples to detect the difference between the H1 hypothesis and the H0 hypothesis for the percentage of C>A.
[0069] Using Comparative Model 2, with the filtering threshold lowered to 25, one negative sample (number: 73871) was found in the experimental results. Its total mutation count was 36 and its total score was 4, indicating a false positive. Therefore, compared to the model in Example 1, Comparative Model 2 showed an increased false positive rate.
[0070] Comparative Example 3 The second-generation sequencing data were analyzed using comparative model 3 to calculate the POLE mutation score.
[0071] The only difference between Model 3 and the present invention is that the conditions for filtering false positives are adjusted. False positives are filtered out based on the number of mutations, and samples corresponding to sequencing data with a mutation number of less than or equal to 65 are filtered out.
[0072] According to the theoretical model, if we assume that the background frequencies of mutation rates for all 12 SNV types are 1:1, it would take 17 samples to detect at least one specific point under the condition of 0.05 significance, while it would take at least 30 samples to detect the difference between the H1 hypothesis and the H0 hypothesis for the percentage of C>A.
[0073] Using Comparative Model 3 with a filtering threshold of 65, one negative sample (number: 107878) was found in the experimental results. Its total mutation count was 59 and its total score was 5, making it a true positive. Therefore, compared to the model in Example 1, Comparative Model 3 showed an increased false negative rate.
[0074] Example 3: A computer-readable storage medium A computer-readable storage medium, the structural diagram of which is shown below. Figure 2 As shown, it stores a computer program / instruction, which, when executed by a processor, implements the steps of the construction method described in Embodiment 1 of the present invention.
[0075] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.
[0076] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0077] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a computer device to execute the methods described in the various embodiments of the present invention.
[0078] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.
Claims
1. A method for constructing a POLE mutation discrimination model based on second-generation sequencing results, characterized in that, Includes the following steps: Preprocessing of S1 and next-generation sequencing data The second-generation sequencing data of the test samples and control samples were pre-filtered to obtain pileup files. The pre-filtering includes quality control, comparison, and deduplication; The sample to be tested is DNA from the tissue or cell-free DNA from the blood of a tumor patient, and the control sample is DNA from the saliva or genomic DNA from the blood of a tumor patient. S2, sequence alignment, variant interpretation, and mutation filtering The pileup file obtained in step S1 was analyzed using a mutation detection tool. The mutation information of cancer-related genes in the data was analyzed, the site location and specific base changes of point mutations and insertion / deletion mutations were annotated, POLE mutations were marked, and the mutation number and maximum mutation abundance of all mutations in the sample were obtained. Based on the number of mutations, false positives are filtered out. When the sample to be tested is DNA from the tissue of a tumor patient, samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 5% are filtered out. When the sample to be tested is cell-free DNA from the blood of a tumor patient, samples corresponding to sequencing data with a mutation number less than or equal to 50 and a maximum mutation abundance less than or equal to 1% are filtered out. The number of mutations includes the number of mutations from C to A, the number of mutations from T to G, the number of mutations from C to G, the total number of point mutations, the total number of insertion / deletion mutations, and the total number of mutations. Based on the POLE mutations annotated in step S2, obtain the number of occurrences of POLE mutations in the corresponding tumors of the samples from the COSMIC database. Tumor burden data is obtained based on the BED file size of the sequencing region of cancer-related genes; Tumor burden = total number of mutations / BED file size of sequencing region; S3. Analysis of mutation filtering data and POLE mutation identification Based on the filtered mutation data from step S2 and the set thresholds, the pathogenicity of POLE mutations is determined. Six integer scoring criteria are set: 1 point is awarded if the percentage of C-type A-type mutations in the total number of point mutations is greater than 20%, otherwise no point is awarded; 1 point is awarded if the percentage of T-type G-type mutations in the total number of point mutations is greater than 4%, otherwise no point is awarded; 1 point is awarded if the percentage of C-type G-type mutations in the total number of point mutations is less than 0.6%, otherwise no point is awarded. A score of 1 is awarded if the percentage of insertion / deletion sites in the total number of mutations is less than 5%, otherwise no score is awarded; a score of 1 is awarded if the tumor burden is greater than 100 muts / mb, otherwise no score is awarded; a score of 1 is awarded if the POLE mutation occurs more than once in the tumor, otherwise no score is awarded. According to the scoring criteria, the POLE mutation score of the sample is statistically analyzed. If the total score is greater than or equal to 4, it is judged to be a pathogenic POLE mutation. The tumor in question is endometrial cancer.
2. The construction method as described in claim 1, characterized in that, The mutation detection tool is either Mutscan or Varscan.
3. The construction method as described in claim 1, characterized in that, The endometrial cancer mentioned is POLE-type endometrial cancer.
4. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the construction method according to any one of claims 1 to 3.