A Glycolysis-related Gene Signature Predicts Prognosis In Gastric Cancer

Objective: To screen glycolysis-related m RNAs associated with the prognosis of GC and construct a risk score model and evaluate the predictive value for GC prognosis by downloading m RNAs transcriptome data of gastric cancer(GC)in the Cancer Genome Atlas(TCGA)database for data mining.Methods: The m RNAs expression profiles and clinical data of GC samples were downloaded from the TCGA database.We run GSEA using m RNA expression data downloaded from TCGA database and integrate the glycolysis-related genes obtained from GSEA with clinical data.Then it was divided into training and testing sets randomly by the use of R software.Univariate regression analysis and multivariate regression analysis were applied to screen the genes related to the prognosis of GC and a risk score model was constructed in the training cohort.And then by using Kaplan–Meier analysis,receiver operating characteristic curve(ROC)used to calculate the area under the curve(AUC)to evaluate the prediction accuracy of the model in the training cohort,testing cohort and entire cohort.Results: Enrichment analysis results showed that 178 genes related to GO＿GLYCOLYTIC＿PROCESS and REACTOME＿GLYCOLYSIS were significantly enriched in tumor tissues(P<0.05).Univariate and multivariate Cox regression analysis were applied to screen out three genes including STC1,PLOD2 and CXCR4 and construct a risk model: STC1 * 0.01542+PLOD2 * 0.03486+ CXCR4 * 0.00668 in the training cohort.And then the risk score of each sample was calculated and the samples were divided into a high-risk group and a low-risk group according to the median risk score.Kaplan–Meier curves show that the survival rate of patients in the high-risk group is significantly shorter than those in the low-risk group(P<0.05)in the training cohort,testing cohort and entire cohort.The results of stratified analysis showed that the risk score model could predict survival of patients within the same clinical features.The ROC curves and AUC calculated were used to evaluate the prediction efficiency of the model.The results show that the AUC is 0.735 and 0.629 in the training cohort and the AUC is 0.766 and0.612 in the testing cohort and the AUC is 0.728 and 0.602 in the entire cohort.The results of independent prognosis analysis also show the model can be as an independent prognostic factor for GC(P<0.05).Conclusion: STC1,PLOD2 and CXCR4 are related to the Prognosis of GC.The glycolysis-related genes risk score model is useful in clinical medicine and its prediction accuracy is better than TNM staging.And it also can be an independent model for survival prediction of GC patients.