Kidney renal cell carcinoma (KIRC), which is the most common subtype of kidney cancer, has a poor prognosis and a high mortality rate. In this study, a multi-omics analysis is performed to build a multi-gene prognosis signature for KIRC. A combination of a DNA methylation analysis and a gene expression data analysis revealed 863 methylated differentially expressed genes (MDEGs). Seven MDEGs (BID, CCNF, DLX4, FAM72D, PYCR1, RUNX1, and TRIP13) were further screened using LASSO Cox regression and integrated into a prognostic risk score model. Then, KIRC patients were divided into high- and low-risk groups. A univariate cox regression analysis revealed a significant association between the high-risk group and a poor prognosis. The time-dependent receiver operating characteristic (ROC) curve shows that the risk group performs well in predicting overall survival. Furthermore, the risk group is contained in the best multivariate model that was obtained by a multivariate stepwise analysis, which further confirms that the risk group can be used as a potential prognostic biomarker. In addition, a nomogram was established for the best multivariate model and shown to perform well in predicting the survival of KIRC patients. In summary, a seven-MDEG signature is a powerful prognosis factor for KIRC patients and may provide useful suggestions for their personalized therapy.
International journal of molecular sciences. 2019 Nov 14*** epublish ***
Fuyan Hu, Wenying Zeng, Xiaoping Liu
Department of Statistics, Faculty of Science, Wuhan University of Technology, 122 Luoshi Road, Wuhan 430070, China., Department of Water Resources and Hydro-elctricity Engineering, College of Water Resources and Architectural Engineering, Northwest A&F University, No.3 Taicheng Road, Yangling 712100, China., School of Mathematics and Statistics, Shandong University at Weihai, Weihai 264209, China.