Research Paper Volume 12, Issue 16 pp 16457—16475

Figure 1. Risk stratification of CM patients using the 21-gene signature. (A) Cross-validation error plot for tuning parameter selection in the proportional hazards model. The R package glmnet returns a sequence of lambda values and cross-validation chooses the optimal value. The plot includes the cross-validation curve (red dotted line) and upper and lower standard deviation (error bars) along the lambda sequence. We used the lambda value with a minimum standard error of the mean (red vertical line) across 100 runs to choose the 21-gene signature. (B) Risk score distribution with patient survival status. The x axis is sorted by risk score values. Red indicates dead patients and blue indicates those still alive. (C) The Kaplan–Meier survival analysis of the signature. Patients were divided into high- and low-risk groups using the cut-off value estimated by the R package survminer. (D) ROC analysis of the signature in predicting the OS of patients.