Fig. 2

Construction and validation of a M2 MRGs-based risk prediction model. (A) A venn diagram between M2 MRGs and differentially expressed genes (DEGs) (left panel), and univariate-Cox regression analysis of cross-set genes (right panel). (B) Least absolute shrinkage and selection operator (LASSO) regression analysis determines the number of factors to build a model. (C) Partial likelihood bias coefficient distribution plot for selected genes. (D) The overall survival (OS) status of the internal validation set. (E) The OS status of the external validation set. (F) Multivariate-Cox regression analysis of our model’s risk score for ccRCC patients. (G) Plots of risk score and survival status distribution. (H) Receiver operating curve (ROC) analysis compared the prognostic predictive ability of risk score and clinical factors. (I) Time-dependent ROC analysis of 1-, 3-, and 5-year survival for ccRCC patients. (J-M) Survival analysis of our constructed model combined with clinical factors (age, gender, grade, and stage)