Fig. 1

The GSAGI predicts prognosis in LUAD patients. A Volcano map of differentially expressed genes in LUAD patients in high and low somatic mutation groups. B Pathway enrichment analysis of GO and KEGG for 347 candidate differential genes, p-values of pathways shown are all < 0.05. C Hazard ratio and P-value of constituents involved in multivariate Cox regression. D PCA of TCGA-LUAD training set to distinguish between high-risk and low-risk groups. E, G, I, K, M, O Kaplan-Meier survival curves in the TCGA-LUAD training set (E), GSE30219 testing set (G), GSE50081 testing set (I), GSE31210 testing set (K), GSE41271 testing set (M), and GSE42127 testing set (O) for patients in the high-risk and low-risk groups differentiated by the GSAGI. F, H, J, L, N, P ROC curves for patients in the TCGA-LUAD training set (F), GSE30219 testing set (H), GSE50081 testing set (J), GSE31210 testing set (L), GSE41271 testing set (N), and GSE42127 testing set (P). AUCs at 1, 2, and 3 years are shown in the figures