Fig. 7
From: Gene clusters based on OLIG2 and CD276 could distinguish molecular profiling in glioblastoma
Survival prediction model based on genes in clusters. Multivariate Cox regression analysis revealed the association of 26 genes with OS and PFS (A). The predictive value of genes in cluster for OS and PFS is shown in 2-dimension plot (B). A signature of five genes is obtained by LASSO regression algorithm (C, D). The coefficients of five genes in the signature is shown (E). The full view of the risk score and the survival status based on five genes signature (F). The survival prediction model is tested in training cohort and test cohort (G)