Figure 3.

Recurrence prediction with ROC analysis and Kaplan-Meier survival analysis with the five-gene panel in the TCGA-STAD project. (a) Recurrence prediction with ROC (receiver operating characteristic) analysis in the TCGA-STAD project. The baseline model was established according to the patient's age, tumor stage, grade, gender, and race, and ROC curve was generated accordingly. The AUC (area under curve) was 0.699. After adding the five-gene prognostic panel to the baseline model, the AUC was 0.813, and the p value was 0.007. (b) Kaplan-Meier survival analysis with the five-gene prognostic panel in the TCGA-STAD project. Survival of patients with at least one mutation in the five genes (no less than one gene) showed significant longer survival time than those without mutations in these five genes (no mutant genes), p value =0.198.