Construction Of A Prognostic Risk Prediction Model For Colorectal Cancer Patients Based On Autophagy-related Genes

Objective: To explore the relationship between ARGs and the prognosis of patients with colorectal cancer by data mining of autophagy-relatedgenes(ARGs)associated with colorectal cancer in the Cancer Genome Atlas(TCGA)database by bioinformatics.To construct a prognostic prediction model for patients with colorectal cancer based on ARGs for evaluating the prognosis and efficacy of patients with colorectal cancer.Methods: 1.All transcriptome profiles and autophagy gene data of colorectal cancer were downloaded from the TCGA database and the Human Autophagy Database(HADb),respectively,and differentially expressed autophagy genes(DE-ARGs)between normal and colorectal cancer tissues were selected.2.GO(Gene Ontology)and KEGG(Kyotoencyclopediaofgenesandgenomes)enrichment analysis of DE-ARGs were performed to determine the biological functions of DE-ARGs and related signaling pathways.3.Multivariate Cox regression analysis was performed after screening DE-ARGs associated with overall survival(OS)by univariate Cox regression analysis to establish a prognostic risk prediction model for colorectal cancer patients with autophagy-related genes.4.Calculate the risk score,divide the patients into high and low risk groups according to the median risk value,and verify the reliability of the model with survival analysis(Kaplan-Meier,K-M)and receiver operating characteristic(ROC)analysis.5.The relationship between the risk score and the clinical characteristics of the patients was compared using the t-test method,including age,gender,pathological stage of the tumor,primary tumor,lymph node status,and distant metastasis status.6.The accuracy of the risk assessment model in this study was verified using the findings in the Gene Expression Profilling Interactive Analysis(GEPIA)as well as the Human Protein Atlas(HPA)database.Results: A total of 66 DE-ARGs were selected,including 27 significantly up-regulated and 39 significantly down-regulated DE-ARGs.2.Both GO and KEGG enrichment analysis showed that they were related to autophagy.3.Univariate Cox regression analysis identified 15 DE-ARGs associated with OS,and multivariate Cox regression analysis finally identified four autophagy-related genes,SLC6A1,CDKN2 A,PPARGC1A,and REP15,for constructing a prediction model.K-M survival curves showed that patients in the high-risk group had poorer overall survival(OS)compared with patients in the low-risk group,and the area under the curve(AUC)of the time-dependent receiver operating characteristic curve(ROC)at 1,3,and 5 years was 0.70,0.72,and 0.73,respectively,confirming that the prediction model had good accuracy.4.The results of correlation analysis between risk score and clinical characteristics showed that risk score was correlated with primary tumor stage,lymph node metastasis status,distant metastasis status and tumor pathological stage.5.The results of GEPIA and HPA databases were basically consistent with the results of this study: in colorectal tissues,the expression of SLC6A1 and CDKN2 A was significantly increased in colorectal tissues,while the expression levels of PPARGC1 A and REP15 were relatively decreased;the overall survival of patients with high expression of SLC6A1 and CDKN2 A was shorter,and the overall survival of patients with high expression of PPARGC1 A and REP15 was longer.Conclusions: In this study,SLC6A1,CDKN2 A,PPARGC1A and REP15,four autophagy-related genes related to the prognosis of colorectal cancer patients,were finally selected for the construction of prognostic risk prediction model,which can be used to evaluate the prognosis and efficacy of colorectal cancer patients and help to further guide clinical treatment.