Bioinformatic Analysis Of Colorectal Cancer Prognosis Based On Oxidative Stress-related Genes

Objective:This study aimed to screen out oxidative stress-related differentially expressed genes in CRC and to find key genes associated with prognosis to construct a prognostic model that can guide clinical practice.Methods:This dataset was screened for differentially expressed oxidative stress genes(DEOSGs)and underwent GO and KEGG enrichment analysis using R software.Resources from the String website were used to build protein-protein interaction(PPI)networks and Cytoscape software were used to screen key genes.Then,the corresponding prognostic genes were obtained by univariate and multivariate Cox regression analysis,and the prognostic model was constructed.Patients were divided into high risk and low risk groups by risk score,and internal and external evaluation were conducted by KaplanMeier survival analysis and subject working characteristic curve(ROC)analysis.Multivariate Cox regression analysis was used to identify whether the built model served as an independent prognostic factor,and finally the model was verified by an external GSE12945 cohort.Results:12 prognosis-related oxidative stress genes were initially obtained by univariate Cox regression analysis,the 5 genes were finally determined by a multivariate Cox regression analysis.After constructing the prognostic model,the survival time found that the low-risk group was much longer than the high-risk group(P<0.001),and the AUC under the 1,3 and 5 years of the prognostic model was large.Multivariate Cox regression analysis further confirmed the independent prognostic value of oxidative stress-related genes,combined with a prognostic model ROC curve of 0.629.Finally,the GSE12945 cohort also validated the effectiveness of the oxidative stress-related genes.Conclusion:A colorectal cancer prognosis model composed of five oxidative stress-related genes was successfully constructed,which has some value in predicting the prognosis of patients.