Skip to main content
. 2021 Aug 20;11(11):jkab296. doi: 10.1093/g3journal/jkab296

Figure 3.

Figure 3

Risk prediction model construction based on optimal genes. OS time of patients in (A) TCGA training, (B) GSE62452, (C) GSE79668, and (D) GSE28735 validation datasets assigned to low-risk group (black) compared with high-risk group (red). (E) ROC curve for all datasets based on optimal gene combination. Black, red, green, and purple curves represent TCGA training, GSE62452, GSE79668, and GSE28735 validation datasets, respectively.