Fig. 1: Stratification of ccRCC, pRCC, and chRCC risk by Sig27.

a–c Sig27 risk scores for individual RCCs were calculated; cutoff points towards poor prognosis (fatality) were estimated using the empirical, kernel, normal, maxstat method. Kaplan–Meier survival curves and log-rank test were performed using the R survival package. The relevant p value, cutoff point, sensitivity, and specificity are included. The individual TCGA Pancancer datasets were used in these analyses. Discriminative analysis of Sig27-derived stratification of OS in ccRCC (d) and pRCC (e). f Evaluate the discriminative power of Sig27 towards OS in ccRCC, pRCC, and chRCC using time-dependent ROC. The individual values of area under the curve (AUC) in the indicated follow-up time frame were graphed. Univariate and multivariate Cox analyses of Sig27, age at diagnosis (Age), Sex, and Stage in predicting OS probability in ccRCC (g) and pRCC (h) were performed with the R survival package in the respective TCGA Pancancer cohort. In ccRCC, Stage II, III, and IV tumors are compared to Stage I tumors; in pRCC, Stage III + IV in comparison to Stage I + II; HR hazard ratio, CI confidence interval.