Fig. 3

Machine learning builds a recurrence prediction model based on angiogenesis-related lncRNA. (A) Twenty vascular-related lncRNAs were significantly associated with recurrence by univariate cox analysis, with a p value of less than 0.001 as the inclusion criterion; (B) Random survival forest analysis identified 4 key lncRNAs in TCGA, namely LUCAT1, AC010761.1, AC006504.7, and MIR210HG; (C, E) KM survival curve validation of predictive effect in training cohort; (C) and validation cohort (E); (D, F) Time-dependent ROC curves of 4-lncRNAs-based classifiers; time-dependent ROC curves of multiple lncRNA-based risk models for recurrence