Construction and evaluation of a prognostic model for renal clear cell carcinoma based on metabolism-related genes in TCGA database

Abstract

Objective To investigated the expression of metabolism-related genes (MRGs) in kidney renal clear cell carcinoma (KIRC) and its correlation with prognosis of patients so as to search for potential intervention targets. Methods Bioinformatics were used to mine KIRC transcriptome data in the Cancer Genome Atlas Program (TCGA) database, and the MRGs abnormally expressed were obtained, among which the ones closely associated with overall survival of KIRC patients were screened by LASSO and Cox regression. Furthermore, Cox regression analysis was adopted to construct a risk-score model to predict the prognosis of KIRC patients, and its predictive ability was evaluated. Finally, the expression characteristics of prognostic related MRGs were verified using RT-qPCR in KIRC cell line A498 and renal tubular epithelial cell line HK-2, respectively. Results A total of 875 differentially expressed MRGs were found between KIRC and adjacent tissues. Univariate Cox regression analysis showed that 38 MRGs significantly affected the prognosis of patients, and 7 of them (CNGB1, CACNA1I, KCNT1, KCNH1, PLA2G4D, CHST6, and ATP2A1) were further selected as prognostic risk factors by LASSO and multivariate Cox regression. Subsequently, a risk-score model was constructed based on the above 7 factors. The score obtained from the model could be used as an independent risk factor to predict the prognosis of KIRC patients (risk ratio: 4.671, P 0.05). Conclusion There are several MRGs with abnormal expression in KIRC, and the prognostic risk model based on 7 genes screened from the MRGs can effectively predict the prognosis of KIRC patients, providing a new basis for future diagnosis and treatment of KIRC.

Keywords

• kidney renal clear cell carcinoma
• metabolism-related genes
• prognostic model