Research Paper Volume 15, Issue 21 pp 12330—12368
A robust six-gene prognostic signature based on two prognostic subtypes constructed by chromatin regulators is correlated with immunological features and therapeutic response in lung adenocarcinoma
Figure 8. Identification of key differentially expressed genes (DEGs) between two clusters. (A) Distribution of DEGs between two clusters. Totals of 466 upregulated and 557 downregulated DEGs were identified in the cluster 1. (B) Identification of key DEGs between two clusters. An overlap analysis showed that totals of 821 dysregulated DEGs were important between two clusters. (C) LASSO Cox analysis. A 10-round cross validation was performed to prevent overfitting and 10 DEGs (CENPH, RHOV, ANLN, MDFI, TPSB2, CPS1, ANGPTL4, CCL20, CENPK, GJB3) had the most important feature when lambda was equal to 0.0606. (D) Univariate Cox regression analysis. Six DEGs including TPSB2, CPS1, ANGPTL4, CCL20, CENPK and GJB3 had significant prognostic values (all p < 0.001). (E) Heat map of expressions of six DEGs between lung adenocarcinoma (LUAD) and normal lung tissues. Five DEGs including CPS1, ANGPTL4, CCL20, CENPK and GJB3 were upregulated expressed and TPSB2 was downregulated expressed in LUAD tissue. (F) Heat map of expressions of six DEGs between two clusters. Five DEGs including CPS1, ANGPTL4, CCL20, CENPK and GJB3 were upregulated expressed and TPSB2 was downregulated expressed in the cluster 1. (G–I) The expression levels of 6 prognostic genes between LUAD and normal lung tissues based on GEPIA (gene expression profiling interactive analysis), real RNA-seq and GSE19804 datasets.