In this study, we performed bioinformatics and statistical analyses to investigate the prognostic significance of metabolic genes in clear cell renal cell carcinoma (ccRCC) using the transcriptome data of 539 ccRCC and 72 normal renal tissues from TCGA database. We identified 79 upregulated and 45 downregulated (n=124) metabolic genes in ccRCC tissues. Eleven prognostic metabolic genes (NOS1, ALAD, ALDH3B2, ACADM, ITPKA, IMPDH1, SCD5, FADS2, ACHE, CA4, and HK3) were identified by further analysis. We then constructed an 11-metabolic gene signature-based prognostic risk score model and classified ccRCC patients into high- and low-risk groups. Overall survival (OS) among the high-risk ccRCC patients was significantly shorter than among the low-risk ccRCC patients. Receiver operating characteristic (ROC) curve analysis of the prognostic risk score model showed that the areas under the ROC curve for the 1-, 3-, and 5-year OS were 0.810, 0.738, and 0.771, respectively. Thus, our prognostic model showed favorable predictive power in the TCGA and E-MTAB-1980 ccRCC patient cohorts. We also established a nomogram based on these eleven metabolic genes and validated internally in the TCGA cohort, showing an accurate prediction for prognosis in ccRCC.
Keywords: bioinformatics; clear cell renal cell carcinoma; metabolic genes; prognostic model.