| Background: Ovarian cancer is a common malignant tumor in the female reproductive system,with the fifth highest incidence in the world,and the highest mortality in gynecological malignant tumors.According to the survey,there were313,959 cases globally in 2020,and 207,252 deaths of ovarian cancer worldwide,accounting for 3.4% of new cases and 4.7% of all deaths.At present,At present,the 5-year survival rate of early ovarian cancer can exceed 90%,but it is often ignored because there are no obvious signs and symptoms.As a result,the overall 5-year survival rate of patients with early ovarian cancer could be more than90%,but it is often ignored because there are no obvious signs and symptoms.As a results,most of the patients（about 70%）are in the advanced stage of cancer,while the5-year survival rate is only about 30%.Therefore,it is particularly important to find the key genes and therapeutic targets that affect the prognosis of ovarian cancer.With the development of genomics technology,it is of great significance to find reliable molecular biomarkers by analyzing tumor biological information through gene expression profile.With the useful application of many disease predictive risk models in the diagnosis,treatment evaluation and prognosis of clinical patients,it is urgent to build a model that can predict the survival rate and therapeutic effect of patients with ovarian cancer.Objective: The purpose of this study was to screen out the prognostic risk model of ovarian cancer by using GEPIA database and the gene set related to pan-RNA modification of ovarian cancer,analyze the relationship between prognostic risk model and clinical indicators,immune cell infiltration and tumor mutation load,and search for intervention targets.In addition,the prognostic prediction model of ovarian cancer was constructed according to the prognostic risk model and clinical indicators,in order to lay a research foundation for improving the prognosis of patients.Methods: The differential genes of patients with TCGA-OV were obtained from GEPIA database,the differentially expressed genes were collected and overlapped with pan-RNA epigenetic genes,obtained the pan-RNA epigenetic differential genes associated with ovarian cancer and analyzed by GO and KEGG.The gene sequencing data and clinical data of OV patients in TCGA database were used as the analysis data source and verification set,which were divided into 70%training set and 30%verification set respectively.In the training set,the prognosis-related genes were screen by the univariate and multivariate Cox regression analysis with R software,to establish the pan-RNA modification related gene risk score and to construct the prognostic gene risk model.Finally,to draw ROC curve and Kaplan-Meier survival analysis curve,and verify at the same time.The patient of ovarian cancer were divided into high risk group and low risk group according to risk core.The differences between high risk groups in clinical indicators,tumor mutation burden,immune checkpoint molecules,immune score,stroma score,tumor-infiltrating lymphocyte and GSEA enrichment analysis were compared to low risk groups explore the correlation between risk model and the above indicators.clinical index,tumor mutation burden,immune checkpoint molecule,immune score,matrix score,tumor-infiltrating lymphocytes and GSEA enrichment analysis were compared to explore the correlation between the risk model and the above indexes,and to further evaluate the credibility of the prediction results of the model.Finally,the relationship between the prognostic risk model and clinical indicators of ovarian cancer was analyzed by univariate and multivariate COX regression analysis,and the Nomogram diagram including prognostic risk model and clinical factors was constructed and the calibration curve was drawn to further evaluate the predictive efficiency of the model.Results: In this study,a total of 13,690 differential genes for ovarian cancer were obtained and 43 pan-RNA epigenetic modification genes related to ovarian cancer were screened.In the training set,the analysis showed that HNRNPA2B1 and YBX1 genes were associated with the prognosis of ovarian cancer patients（P < 0.05）.Next,we constructed the prognostic risk model.We divided the patients into the high-low risk group based on the median risk score.It was found that patients with low risk of ovarian cancer had high survival rate and long survival cycle.The AUC of 1,3 and 5years survival rate predicted by the model was 0.536,0.566 and 0.620.The results show that the prognostic risk model has certain judgment value in survival prediction.And the verification set test is performed and verified.Correlation analysis of clinical indicators based on the risk model showed that there were no significant differences in age,tumor stage,residual tumor size,lymph node metastasis and vascular invasion except for tumor grade among ovarian cancer patients.We continued to analyze the tumor microenvironment based on the risk model,and we found significant immune differences between patients in the high and low risk groups.Among the ovarian cancer patients,the immune score,matrix score and immunoassay molecular points C10orf54,CD27,TMIGD2,TNFRSF14,TNFRSF18 and TNFRSF4 in high-risk group were higher than those in low-risk group.There was no statistical significance in TMB scores between the two groups.Finally,we combined clinical data and Risk Score to conduct variable screening,and found that age,residual tumor size and Risk Score could be independent predictors of prognosis of ovarian cancer patients.In addition,age,residual tumor size and Risk Score were included in the multivariate Cox regression Risk scale model,and a Nomogram was drawn to predict the individual survival rate of ovarian cancer patients.Calibration curve,Kaplan-Meier survival analysis curve and ROC curve were used to evaluate the survival rate.Conclusion: In this study,we obtained 43 pan-RNA epigenetic genes from ovarian cancer based on TCGA database and pan-RNA epigenetic gene set.Then,we used Cox regression analysis on these 43 genes and screened HNRNPA2B1 and YBX1 genes as potential influences on the prognosis of ovarian cancer patients.Therefore,we constructed a prognostic risk model of ovarian cancer and verified it well.And there are significant immune differences between ovarian cancer patients.Finally,we constructed a prognostic model for ovarian cancer by combining risk score and clinical indicators. |
PDF Full Text Request