Fig. 4

The assessment of the predictive power of the model using the KM analysis and ROC curve in the GSE21257 validation set (A-B) The number of dead patients increased with the increase of the risk score. (C) The risk heat map shows the expression of the three genes from low risk to high risk in the GSE21257 validation set. (D) Survival analysis using KM methods in the GSE21257 validation set (P < 0.05). (E) An analysis of ROC curves that predict survival rates over 1, 3, and 5 years (AUCs of 0.857,0.737,0.730, respectively) in time-dependent scenarios for survival prediction of OS patients on the GSE21257 validation set. (F) Total points are obtained by incorporating the corresponding points of age, gender, site, and risk score on the point scale. (G) The multivariate cox regression analysis has identified that risk score is an independent prognostic factor in OS