System of ovarian cancer risk model based on pan-apoptosis related gene

Abstract

The invention provides a system of an ovarian cancer risk model based on generic apoptosis related genes. The system comprises a data acquisition and preprocessing module, a clustering analysis module, a prognosis feature construction module, a high-risk and low-risk group tumor microenvironment and drug sensitivity evaluation module, a statistical analysis module and a risk model based on differential expression genes DEGs. According to the technical scheme, key genes in the cell apoptosis process are used for ovarian cancer risk prediction, and the blank of an existing risk model in biomarker selection is filled up.

Description

Technical Field

[0001] The invention relates to the field of biomedical technology, in particular to a system of ovarian cancer risk model based on pan-apoptosis related genes. Background Art

[0002] Ovarian cancer (OC) is the most malignant tumor in the female reproductive system, with an overall morbidity and mortality rate that is gradually increasing, posing a serious threat to human health. Due to the lack of typical clinical manifestations in the early stages of OC and the lack of effective screening methods and therapeutic targets at this stage, patients are often diagnosed in the late stages. Although modern medicine has made progress in early detection and treatment, early diagnosis of ovarian cancer still faces challenges. Current risk assessment methods mostly rely on clinical characteristics and traditional biomarkers, but these methods often have limited accuracy. Therefore, it is of great clinical value to develop new risk assessment tools to improve early screening of ovarian cancer.

[0003] Currently, the most commonly used tumor markers for ovarian cancer screening are serum carbohydrate antigen 125 (CA125) and human epididymis protein 4 (HE4). However, many ovarian cancer patients do not have significantly elevated serum CA125 levels in the early stages of the disease, and CA125 and HE4 are not only elevated in ovarian cancer, but also abnormally elevated in other cancers or non-malignant inflammatory reactions. Using only HE4 or CA125 for early ovarian cancer diagnosis may result in a high rate of misdiagnosis or missed diagnosis.

[0004] Pan-apoptosis (PANoptosis) is a cell programmed death mode that combines the characteristics of apoptosis, pyroptosis, and necroptosis. Due to the complex mechanism of PANoptosis, it has been widely studied in recent years. Studies have found that PANoptosis disorders are associated with a variety of human diseases, including autoinflammatory diseases, cancer, infectious diseases, and metabolic disorders. Studies have confirmed that PANoptosis can induce cell death in different cancer cell lines, and its abnormal expression may be related to the risk of ovarian cancer. Therefore, the establishment of an ovarian cancer risk model based on pan-apoptosis-related genes may provide a new direction for the early diagnosis and personalized treatment of ovarian cancer. At present, the prognostic value of PANoptosis genes in tumors such as colon cancer, lung cancer, and breast cancer has been gradually carried out, but there are few pan-apoptosis-related prognostic models for ovarian cancer patients, and the various tumor prognostic models based on PANoptosis characteristics are mostly constructed from a single aspect, rarely combining clinical characteristics, molecular landscapes, and gene mutation characteristics. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a system of ovarian cancer risk model based on pan-apoptosis-related genes, which uses key genes in the cell apoptosis process to predict ovarian cancer risk, filling the gap in biomarker selection in existing risk models.

[0006] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a system for an ovarian cancer risk model based on pan-apoptosis-related genes, comprising a data acquisition and preprocessing module, a cluster analysis module, a prognostic feature construction module, a tumor microenvironment and drug sensitivity assessment module for high-risk and low-risk groups, a statistical analysis module, and a risk model based on differentially expressed genes DEGs.

[0007] In a preferred embodiment, the data acquisition and preprocessing module collects transcriptome and clinical data of 412 cases of ovarian cancer from the Cancer Genome Atlas TCGA database; the data acquisition and preprocessing module acquires the mRNA expression of 88 normal ovarian samples from the GTEx database; the data acquisition and preprocessing module uses the expression spectrum and clinical information of the GSE30161 dataset as a validation set.

[0008] In a preferred embodiment, the data acquisition and preprocessing module identifies 19 pan-apoptosis-related genes PRGs, namely CASP8, FADD, CASP6, NLRP3, TAB2, TAB3, PSTPIP2, TNFAIP3, CASP7, PARP1, GSDMD, MLKL, IRF1, AIM2, ZBP1, CASP1, RIPK1, RIPK3, and TRADD.

[0009] In a preferred embodiment, the cluster analysis module uses the ConsensusClusterPlus software package to perform cluster analysis on the ovarian cancer OC patient data in TCGA to clarify the PRGs-related ovarian cancer molecular subtypes, namely Cluster1 and Cluster2; the Kaplan-Meier survival curve is used to perform survival analysis of different molecular subtypes of ovarian cancer; the cluster analysis module uses the limma software package to establish a linear model for each gene by taking Cluster1 as the experimental group and Cluster2 as the control group, and estimates the expression difference of each gene between different groups through the linear model; the cluster analysis module uses the lmFit function to calculate the coefficient estimate and standard error of each gene; the cluster analysis module uses the eBayes function to perform Bayesian adjustment on the model; for each gene, the t statistic and p value are calculated to evaluate whether the expression difference between different groups is significant; the cluster analysis module uses the topTable function to extract gene DEGs with significant differences, and the cluster analysis module uses "|log fold change (log2FC)|>2" and "corrected P<0.05" as the screening criteria for differentially expressed genes.

[0010] In a preferred embodiment, the prognostic feature module is constructed to perform univariate Cox regression analysis on the differentially expressed genes DEGs between molecular subtypes of ovarian cancer to evaluate the relationship between gene expression and overall survival (OS); the model form is: h(t)=h0(t)·exp(β·X), where h(t) is the risk function under a certain gene expression, h0(t) is the baseline risk function, β is the regression coefficient, and X is the gene expression level; for each gene, a COX regression model is constructed separately, with the gene expression level as the only independent variable, the survival time and the survival status as the dependent variable, the effect of gene expression on survival risk is calculated, the regression coefficient β of each gene and its standard error are calculated, and then the P value is obtained for evaluation Whether gene expression is significantly correlated with survival; the exponential value of the regression coefficient exp(β) represents the hazard ratio HR; HR>1 indicates that increased gene expression is associated with increased risk, and HR<1 indicates that increased gene expression is associated with decreased risk; genes significantly associated with survival were screened out based on P value and HR value, and prognosis-related DEGs were identified with p.value<0.05 as the standard; subsequently, the screened prognosis-related DEGs were subjected to LASSO regression analysis and multivariate Cox regression analysis using the R packages "survival", "survminer" and "glmnet" to select model genes with independent prognostic value; the calculation formula of PRGs risk score is: risk score = ∑(β i *Exp i ), where β irepresents the LASSO coefficient of the i-th gene, Exp i represents the expression level of the ith gene; patients were divided into high-risk group and low-risk group according to the median risk score, and their prognostic results were evaluated by Kaplan-Meier analysis; the stability of the prognostic model was evaluated using the time-dependent receiver operating characteristic ROC curve and the area under the curve AUC.

[0011] In a preferred embodiment, the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups uses CIBERSORT to evaluate the immune cell infiltration in the gene expression data based on TCGA; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups uses the "ESTIMATE" R package to calculate the immune score, stromal score and ESTIMATE score, which represent immune infiltration, stromal cell infiltration and overall tumor purity, respectively; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups calculates the immune cell infiltration score, and uses single sample gene set enrichment analysis ssGSEA to evaluate immune function; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups uses the "pRRophetic" R package to predict the drug response of cancer patients, and evaluates the drug sensitivity of different patient subgroups based on cell line screening data.

[0012] In a preferred embodiment, the statistical analysis module is performed using R software; the differences between normally distributed data groups are analyzed using t-test, and the differences between data groups that do not obey the normal distribution are analyzed using Wilcoxon rank sum test; the survival curve is generated using the Kaplan-Meier method, and the inter-group comparison is performed using the log-rank test; and the multivariate survival analysis uses Cox proportional hazard regression analysis.

[0013] In a preferred embodiment, the risk model based on differentially expressed gene DEGs was performed by performing univariate and multivariate Cox regression analysis on the differentially expressed genes between Cluster 1 and Cluster 2 to screen out genes with independent prognostic value, including CD38, CXCL11, CXCL10, CD2 and CXCL13, which were used for LASSO regression analysis.

[0014] In a preferred embodiment, the goal of LASSO regression analysis is to minimize the following objective function: y i is the survival time of the ith sample, x iis the gene expression of the ith sample, β is the regression coefficient, n is the number of samples, p is the number of genes, λ is the regularization parameter, which controls the strength of the L1 norm penalty. λ=0.0172 is selected through cross-validation, and the optimization algorithm is used to minimize the objective function to obtain the regression coefficient β. The regression coefficient of each gene is checked, and the genes assigned with non-zero coefficients are identified. These genes are important features in the model, indicating that the gene has a significant effect on the predicted survival time.

[0015] In a preferred embodiment, four model construction genes are finally determined, including CD38, CXCL11, CD2 and CXCL13. According to the β coefficient and the gene expression levels of CD38, CXCL11, CD2 and CXCL13 in each sample, the risk score of each sample is determined as: (-0.27750442)*CD38+(-0.04081566)*CXCL11+(0.10028621)*CD2+(-0.07546636)*CXCL13. The TCGA samples are divided into high-risk group and low-risk group according to the median value of the risk score. The correlation between PANclutser, high-risk and low-risk groups and survival outcomes is shown using a Sankey diagram. Through univariate and multivariate Cox and LASSO regression analysis, a nomogram is established based on age, grade, stage and risk score to improve the clinical applicability of the model.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] 1. Risk assessment based on pan-apoptosis-related genes: Using key genes in the cell apoptosis process to predict ovarian cancer risk fills the gap in biomarker selection in existing risk models.

[0018] 2. Integration of multiple data sources: Combining genomic data with clinical data enhances the predictive power and clinical applicability of the model.

[0019] 3. Application of machine learning algorithms: Using advanced machine learning techniques to optimize risk models improves prediction accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a box plot of differentially expressed genes in ovarian cancer and adjacent tissues according to a preferred embodiment of the present invention.

[0021] Figure 2 This is the mutation information of 19 PRGs in ovarian cancer samples according to a preferred embodiment of the present invention.

[0022] Figure 3 The positions of CNV changes of 19 PRGs on chromosomes in the preferred embodiment of the present invention are shown.

[0023] Figure 4 In a preferred embodiment of the present invention, unsupervised cluster analysis based on 19 PRGs was performed to divide ovarian cancer samples into two categories (cluster 1 and cluster 2).

[0024] Figure 5 This is a prognostic difference diagram for two groups of people in a preferred embodiment of the present invention.

[0025] Figure 6 The LASSO regression analysis screens and obtains prognosis-related genes in the preferred embodiment of the present invention.

[0026] Figure 7 The Sankey diagram, which is a preferred embodiment of the present invention, shows the distribution of PAN clusters, gene clusters, risk score groups, and survival outcomes.

[0027] Figure 8 Multivariate regression analysis, which is a preferred embodiment of the present invention, showed the independent prognostic predictive value of the risk score.

[0028] Fig. 9 This is a nomogram and calibration diagram for predicting 5-year, 8-year, and 10-year OS of ovarian cancer patients using the risk score according to a preferred embodiment of the present invention.

[0029] Fig.10 It is the survival status distribution between high and low risk scores and the expression heat map of four characteristic genes in the preferred embodiment of the present invention.

[0030] Fig.11 This is the prognostic difference among different risk groups according to a preferred embodiment of the present invention.

[0031] Fig.12 This is a ROC analysis of the OS risk score of the TCGA ovarian cancer cohort at 5, 8, and 10 years of follow-up in a preferred embodiment of the present invention.

[0032] Fig.13 It is the survival status distribution between high and low risk scores and the expression heat map of four characteristic genes in the validation set of the preferred embodiment of the present invention.

[0033] Fig.14 This is the prognostic difference between different risk groups in the validation set of the preferred embodiment of the present invention.

[0034] Fig.15 This is the ROC analysis of the OS risk score of the ovarian cancer cohort in the validation set of the preferred embodiment of the present invention after 5, 8 and 10 years of follow-up.

[0035] Fig.16Figure 1 is a comprehensive assessment of the immune cell landscape of the risk model of the preferred embodiment of the present invention. (A) Association between risk score and immune cell type. (B) Correlation between model genes and immune cell type. (C) Differences in immune cell abundance between different risk models. (D) Immune cell infiltration between different risk model groups. (E) Immune score, matrix score, and ESTIMATE score between different risk models.

[0036] Fig.17 This is a chemotherapy drug sensitivity analysis of different risk groups of people according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0038] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0039] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0040] A system for ovarian cancer risk model based on pan-apoptosis-related genes, reference Figure 1-17 ,include:

[0041] Data acquisition and preprocessing module

[0042] Transcriptome and clinical data of 412 cases of ovarian cancer were collected from the Cancer Genome Atlas (TCGA) database (https: / / portal.gdc.cancer.gov / ). The mRNA expression data of normal ovarian samples were obtained from the GTEx database (https: / / www.gtexportal.org / ), which included 88 normal ovarian samples. The expression profile and clinical information of the GSE30161 dataset were used as a validation set. Based on previous studies and literature, 19 pan-apoptosis-related genes (PRGs) were identified, namely CASP8, FADD, CASP6, NLRP3, TAB2, TAB3, PSTPIP2, TNFAIP3, CASP7, PARP1, GSDMD, MLKL, IRF1, AIM2, ZBP1, CASP1, RIPK1, RIPK3, and TRADD.

[0043] Clustering pattern module based on PRGs

[0044] Based on the TCGA expression profile of PRGs, the “ConsensusClusterPlus” software package was used to cluster the ovarian cancer (OC) patient data in TCGA to identify the molecular subtypes of ovarian cancer (Cluster1, Cluster2) associated with PRGs. The survival analysis of different molecular subtypes of ovarian cancer was performed using the Kaplan-Meier (KM) survival curve. The “limma” software package established a linear model for each gene by using Cluster1 as the experimental group and Cluster2 as the control group, and the expression difference of each gene between different groups was estimated by the linear model. The lmFit function calculated the coefficient estimate and standard error of each gene. The Bayes function was used to perform Bayesian adjustment on the model. For each gene, the t statistic and p value were calculated to evaluate whether its expression difference between different groups was significant. The topTable function was used to extract genes with significant differences (DEGs), and “|log fold change(log2FC)|>2” and “corrected P<0.05” were used as the screening criteria for differentially expressed genes.

[0045] Constructing a prognostic feature module for ovarian cancer patients

[0046] First, univariate Cox regression analysis was performed on differentially expressed genes (DEGs) between molecular subtypes of ovarian cancer to evaluate the relationship between gene expression and overall survival (OS). The model form is: h(t) = h0(t) exp(β · X), where h(t) is the risk function under a certain gene expression, h0(t) is the baseline risk function, β is the regression coefficient, and X is the gene expression. For each gene, a separate COX regression model was constructed, with gene expression as the only independent variable, survival time and survival status as dependent variables, and the effect of gene expression on survival risk was calculated. The regression coefficient (β) and its standard error of each gene were calculated, and then the P value was obtained to evaluate whether gene expression was significantly associated with survival. The exponential value of the regression coefficient (exp(β)) represents the hazard ratio (HR). HR>1 indicates that an increase in gene expression is associated with an increased risk, and HR<1 indicates that an increase in gene expression is associated with a decreased risk. Genes significantly associated with survival were screened based on P and HR values, and DEGs associated with prognosis were identified with p.value<0.05 as the standard. Subsequently, the screened prognosis-related DEGs were subjected to LASSO regression analysis and multivariate Cox regression analysis using the R packages “survival”, “survminer”, and “glmnet” to select model genes with independent prognostic value. The calculation formula of PRGs risk score is: risk score = ∑(β i *Exp i ), where β i represents the LASSO coefficient of the i-th gene, Exp i represents the expression level of the ith gene. Patients were divided into high-risk and low-risk groups according to the median risk score, and their prognostic results were evaluated by Kaplan-Meier analysis. The stability of the prognostic model was evaluated using the time-dependent receiver operating characteristic (ROC) curve and the area under the curve (AUC).

[0047] Tumor microenvironment and drug sensitivity assessment module for high-risk and low-risk groups

[0048] The immune cell infiltration in TCGA-based gene expression data was evaluated using CIBERSORT. Next, the “ESTIMATE” R package was applied to calculate the immune score, stromal score, and ESTIMATE score, which represent immune infiltration, stromal cell infiltration, and overall tumor purity, respectively. To further characterize the immune cell infiltration, the immune cell infiltration score was calculated, and the immune function was evaluated using single-sample gene set enrichment analysis (ssGSEA). In addition, the “pRRophetic” R package was used to predict the drug response of cancer patients, and the drug sensitivity of different patient subgroups was evaluated based on cell line screening data.

[0049] Statistical analysis

[0050] Statistical analysis was performed using R software (version 4.1.3). Differences between groups were analyzed using the t test for normally distributed data, and the Wilcoxon rank-sum test for data that did not follow a normal distribution. Survival curves were generated using the Kaplan-Meier method, and the log-rank test was used for intergroup comparisons. Multivariate survival analysis was performed using Cox proportional hazards regression analysis. Two-sided P values ​​less than 0.05 in all tests were considered statistically significant.

[0051] Genetic variation of pan-apoptosis-related genes in ovarian cancer

[0052] Based on the results of previous studies, 19 PRGs were included. 412 ovarian cancer samples and 88 normal ovarian tissue samples from TCGA and GTEx databases were used to analyze the 19 PRGs ( Figure 1 ). There was no significant difference in the expression of TAB2, MLKL and PSTPIP2 genes between the two groups, while TNFAIP3, TAB3 and GSDMD genes were highly expressed in normal ovarian tissues, while the other genes were highly expressed in ovarian cancer tissues (p<0.05). By analyzing the somatic mutation rate of 19 PRGs in ovarian cancer patients, it was found that 44 out of 462 ovarian cancer samples had PRGs mutations, with a mutation frequency of 9.52%, among which the NLRP3 gene had the highest mutation frequency ( Figure 2 ). Copy number variations (CNVs) of PRGs were observed on human chromosomes, with a high frequency of copy number amplification in GSDMD, RIPK1, and NLRP3 genes, and a high frequency of copy number deletion in TAB2, TRADD, and TNFAIP3 genes ( Figure 3 ).

[0053] Identification and analysis of molecular subtypes associated with pan-apoptotic genes

[0054] Based on 19 PRGs, consensus clustering classification was performed on ovarian cancer samples from the TCGA dataset. After unsupervised clustering analysis, the samples were divided into two clusters, named Cluster 1 (n = 211) and Cluster 2 (n = 201) ( Figure 4 ). Prognostic analysis showed that the difference in survival time between the two groups was statistically significant, and the prognosis of Cluster 2 was significantly better than that of Cluster 1 ( Figure 5 ).

[0055] Establishment and validation of risk model based on differentially expressed genes (DEGs)

[0056] Univariate and multivariate Cox regression analysis was performed on the 135 differentially expressed genes between Cluster 1 and Cluster 2 to screen out genes with independent prognostic value (p < 0.05), including CD38, CXCL11, CXCL10, CD2, and CXCL13. These potential marker genes were used for LASSO regression analysis (e.g. Figure 6 ), the goal of LASSO regression analysis is to minimize the following objective function: y i is the survival time of the ith sample, x i is the gene expression of the ith sample, β is the regression coefficient, n is the number of samples, p is the number of genes, λ is the regularization parameter, which controls the strength of the L1 norm penalty. λ=0.0172 is selected through cross-validation, and the optimization algorithm is used to minimize the objective function to obtain the regression coefficient β. The regression coefficient of each gene is checked, and the genes assigned with non-zero coefficients are identified. These genes are important features in the model, indicating that the gene has a significant effect on the predicted survival time. Finally, four genes for model construction were determined, including CD38, CXCL11, CD2, and CXCL13. According to the β coefficient and the gene expression of CD38, CXCL11, CD2, and CXCL13 in each sample, the risk score of each sample was determined as: (-0.27750442)*CD38+(-0.04081566)*CXCL11+(0.10028621)*CD2+(-0.07546636)*CXCL13. According to the median value of the risk score, the TCGA samples were divided into high-risk and low-risk groups. The Sankey diagram was used to show the correlation between PANclutser, high-risk and low-risk groups, and survival outcomes ( Figure 7 ). Multivariate regression analysis showed that the risk score was an independent indicator of poor prognosis ( Figure 8 ). A nomogram was established based on age, grade, stage, and risk score through univariate and multivariate Cox and LASSO regression analysis to improve the clinical applicability of the model. The nomogram helps to accurately predict the 5-year, 8-year, and 10-year overall survival of patients. The calibration curve showed that the survival rate predicted by the nomogram was highly consistent with the actual observation results ( Fig. 9 ). Through the median risk score, it was found that the risk score of ovarian cancer patients increased from left to right. The higher the risk score, the more deaths and the poorer the patient's prognosis. The heat map showed that the model gene expression in the low-risk group was higher than that in the high-risk group, indicating that the high expression of the model gene indicates that the patient has a better prognosis ( Fig.10 ). Through prognostic analysis, it was found that the prognosis of patients in the low-risk group was significantly better than that in the high-risk group ( Fig.11The sensitivity and specificity of the prognostic model were evaluated by ROC curve, and the results also proved that the model has good predictive ability. The AUCs for predicting 5-year, 8-year, and 10-year prognostic outcomes for ovarian cancer patients were 0.66, 0.68, and 0.66, respectively ( Fig.12 ). Using the GEO30161 database as the validation set, the same prediction results were obtained ( Fig.13 , 14 and 15).

[0057] Evaluation of the tumor microenvironment in two risk subgroups

[0058] Immune infiltration analysis revealed differences in TME between the high-risk group and the low-risk group ( Fig.16 A). As the risk score increases, the infiltration of most immune cells decreases, such as macrophages, NK cells, CD4 and other killer cells, and only the initial B cells increase with the increase of risk score, which may be due to the stimulation of the generation of initial immune cells by the tumor. In addition, the model gene is positively correlated with plasma cells, CD8 cells, CD4 cells, follicular helper T cells and macrophages ( Fig.16 B). The relationship between risk group and immune cell abundance and infiltration was also discussed, and it was found that there were significant differences in immune infiltrating cells between high-risk and low-risk groups ( Fig.16 C), the infiltration levels of Tfh, CD8+T cells, and M1-macrophages were higher in the low-risk group ( Fig.16 D). It was also found that the Stromal score, Immune score and Estimate score were significantly different between different subgroups, and the scores of the high-risk group were lower than those of the low-risk group ( Fig.16 E).

[0059] Predictive effect of risk model on sensitivity to chemotherapy drugs

[0060] To verify the value of risk score in predicting chemotherapy response, the sensitivity of the two groups to commonly used ovarian cancer chemotherapy drugs was compared ( Fig.17 ). It was found that the IC50 values ​​of bleomycin, gemcitabine and paclitaxel in the high-risk group were significantly higher than those in the low-risk group, indicating that the low-risk group was more sensitive to these drugs, while the high-risk group was more sensitive to vorinostat, X17.AAG and imatinib. Overall, the low-risk group was more sensitive to common ovarian cancer chemotherapy drugs, which provides a theoretical basis for the application of the model in personalized drug therapy.

Claims

1. A system for ovarian cancer risk model based on pan-apoptosis-related genes, characterized in that: It includes data acquisition and preprocessing module, cluster analysis module, prognostic feature construction module, tumor microenvironment and drug sensitivity assessment module for high-risk and low-risk groups, statistical analysis module and risk model based on differentially expressed gene DEGs.

2. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The data acquisition and preprocessing module collects transcriptome and clinical data of 412 cases of ovarian cancer from the Cancer Genome Atlas TCGA database; the data acquisition and preprocessing module obtains the mRNA expression of 88 normal ovarian samples from the GTEx database; the data acquisition and preprocessing module uses the expression spectrum and clinical information of the GSE30161 dataset as a validation set.

3. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 2, characterized in that: The data acquisition and preprocessing module identifies 19 pan-apoptosis-related genes PRGs, namely CASP8, FADD, CASP6, NLRP3, TAB2, TAB3, PSTPIP2, TNFAIP3, CASP7, PARP1, GSDMD, MLKL, IRF1, AIM2, ZBP1, CASP1, RIPK1, RIPK3, and TRADD.

4. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The cluster analysis module uses the ConsensusClusterPlus software package to perform cluster analysis on the ovarian cancer OC patient data in TCGA to clarify the PRGs-related ovarian cancer molecular subtypes, namely Cluster1 and Cluster2; the Kaplan-Meier survival curve is used to perform survival analysis of different molecular subtypes of ovarian cancer; the cluster analysis module uses the limma software package to establish a linear model for each gene by taking Cluster1 as the experimental group and Cluster2 as the control group, and estimates the expression difference of each gene between different groups through the linear model; the cluster analysis module uses the lmFit function to calculate the coefficient estimate and standard error of each gene; the cluster analysis module uses the eBayes function to perform Bayesian adjustment on the model; For each gene, the t statistic and p value were calculated to assess whether the expression difference between different groups was significant; the cluster analysis module used the topTable function to extract gene DEGs with significant differences, and the cluster analysis module used "|log fold change (log2FC)|>2" and "corrected P<0.05" as the screening criteria for differentially expressed genes.

5. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The prognostic feature module is constructed to perform univariate Cox regression analysis on differentially expressed genes DEGs between molecular subtypes of ovarian cancer, and evaluate the relationship between gene expression and overall survival (OS); the model form is: h(t)=h0(t)·exp(β·X), where h(t) is the risk function under a certain gene expression, h0(t) is the baseline risk function, β is the regression coefficient, and X is the gene expression level; for each gene, a COX regression model is constructed separately, with the gene expression level as the only independent variable, and the survival time and survival status as the dependent variable, and the effect of gene expression on survival risk is calculated, and the regression coefficient β and its standard error of each gene are calculated, and then the P value is obtained to evaluate whether the gene expression is significantly correlated with the survival period; The exponential value of the regression coefficient exp(β) represents the hazard ratio HR; HR>1 indicates that an increase in gene expression is associated with an increased risk, and HR<1 indicates that an increase in gene expression is associated with a decreased risk; genes significantly associated with survival were screened based on P value and HR value, and prognosis-related DEGs were identified with p.value<0.05 as the standard; subsequently, the screened prognosis-related DEGs were subjected to LASSO regression analysis and multivariate Cox regression analysis using the R packages "survival", "survminer" and "glmnet" to select model genes with independent prognostic value; the calculation formula for PRGs risk score is: risk score = ∑(β i *Exp i ), where β i represents the LASSO coefficient of the i-th gene, Exp i represents the expression level of the ith gene; patients were divided into high-risk group and low-risk group according to the median risk score, and their prognostic results were evaluated by Kaplan-Meier analysis; The stability of the prognostic model was evaluated using the time-dependent receiver operating characteristic (ROC) curve and the area under the curve (AUC).

6. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups used CIBERSORT to evaluate the immune cell infiltration in the gene expression data based on TCGA; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups applied the "ESTIMATE" R package to calculate the immune score, stromal score and ESTIMATE score, which represent immune infiltration, stromal cell infiltration and overall tumor purity, respectively; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups calculated the immune cell infiltration score, and used single sample gene set enrichment analysis ssGSEA to evaluate immune function; the tumor microenvironment and drug sensitivity assessment module of the high-risk and low-risk groups used the "pRRophetic" R package to predict the drug response of cancer patients, and evaluated the drug sensitivity of different patient subgroups based on cell line screening data.

7. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The statistical analysis module was performed using R software; differences between groups of normally distributed data were analyzed using the t-test, and differences between groups of data that did not conform to the normal distribution were analyzed using the Wilcoxon rank-sum test; survival curves were generated using the Kaplan-Meier method, and inter-group comparisons were performed using the log-rank test; multivariate survival analysis used Cox proportional hazard regression analysis.

8. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 1, characterized in that: The risk model based on differentially expressed genes DEGs was developed by performing univariate and multivariate Cox regression analysis on the differentially expressed genes between Cluster 1 and Cluster 2, and genes with independent prognostic value were screened, including CD38, CXCL11, CXCL10, CD2, and CXCL13, which were used for LASSO regression analysis.

9. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 8, characterized in that: The goal of LASSO regression analysis is to minimize the following objective function: y i is the survival time of the ith sample, x i is the gene expression of the ith sample, β is the regression coefficient, n is the number of samples, p is the number of genes, λ is the regularization parameter, which controls the strength of the L1 norm penalty. λ=0.0172 is selected through cross-validation, and the optimization algorithm is used to minimize the objective function to obtain the regression coefficient β. The regression coefficient of each gene is checked, and the genes assigned with non-zero coefficients are identified. These genes are important features in the model, indicating that the gene has a significant effect on the predicted survival time.

10. The system of the ovarian cancer risk model based on pan-apoptosis-related genes according to claim 9, characterized in that: Finally, four genes for model construction were determined, including CD38, CXCL11, CD2, and CXCL13. According to the β coefficient and the gene expression levels of CD38, CXCL11, CD2, and CXCL13 in each sample, the risk score of each sample was determined as follows: (-0.27750442)*CD38+(-0.04081566)*CXCL11+(0.10028621)*CD2+(-0.07546636)*CXCL13. According to the median value of the risk score, the TCGA samples were divided into high-risk and low-risk groups. The correlation between PANclutser, high- and low-risk groups, and survival outcomes was demonstrated using a Sankey diagram. A nomogram was established based on age, grade, stage, and risk score by univariate and multivariate Cox and LASSO regression analysis to improve the clinical applicability of the model.

Images