Identification of prognostically relevant gene signatures in lung squamous cell carcinoma by Lung Cancer Explorer

Abstract

Background: Biological features and key genes are urgently needed to reveal the transcriptomic landscape and guide diagnosis and therapy for lung squamous cell carcinoma (LUSC). However, most papers have only focused on highly expressed genes correlated with poor survival in LUSC, which limits the understanding of this cancer type.Methods: Meta-analysis results of genes with tumour-normal differential expression and survival association in LUSC patients provided by the web-tool Lung Cancer Explorer (LCE) were used to determine the differentially expressed genes (DEGs) and prognostically relevant genes (PRGs); the intersected genes were divided into groups, and their biological functions were explored by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. The network of each group was visualized, and the top-ranked genes were selected by the 'degree' method and further tested for their survival association using the Kaplan-Meier (KM) Plotter web tool. Paired cancer and adjacent tissues from LUSC patients were used to confirm the differential expression.Results: A total of 506 genes, as both DEGs and PRGs, were categorized into 4 gene signatures representing distinct biological features, including regulation of the immune system, epithelium development, small GTPase-mediated signal transduction, Adenosine diphosphate (ADP) metabolic process, and glycolysis. Finally, 9 hub genes were identified and subsequently retested to be correlated with survival through univariate and multivariate Cox regression analyses in KM plotter. Paired clinical samples validated the changes in mRNA expression of NTRK2, RAB11A, and CD52. Conclusions:This study provides more insight into the functional and molecular features of LUSC through meta-analysis of available data and identifies therapeutic and diagnostic markers for LUSC.