Identification of prognostic candidate signatures by systematically revealing transcriptome characteristics in lung adenocarcinoma with differing tumor microenvironment immune phenotypes

Accumulated evidence shows that tumor microenvironment plays crucial roles in predicting clinical outcomes of lung adenocarcinoma (LUAD). The current study aimed to identify some potentially prognostic signatures by systematically revealing the transcriptome characteristics in LUADs with differing immune phenotypes. LUAD gene expression data were retrieved from the public TCGA and GEO databases, and the transcriptome characteristics were systematically revealed using a comprehensive bioinformatics method including singlesample gene set enrichment analysis, differentially expressed gene (DEG) analysis, protein and protein gene coexpression network analysis and prognostic model establishment. Finally, 1169 key DEGs associated with LUAD immune phenotype, including 88 immune DEGs, were excavated. Five essential and eight immune essential DEGs were separately identified by constructing two PPI networks based on the above DEGs. Totals of 1085 key DElncRNAs and 45 key DEmiRNAs were excavated and one ceRNA network consisting of 26 DEmRNAs, 3 DEmiRNAs and 57 DElncRNAs were established. The most significant gene coexpression module (cor=0.63 and p=3e-55) associated with LUAD immune phenotypes and three genes (FGR, BTK, SPI1) related to the immune cell infiltration were identified. Three robust prognostic signatures including a 9-lncRNA, an 8-lncRNA and an 8mRNA were established. The areas under the curves of 5-year correlated with overall survival rate were separately 0.7319, 0.7228 and 0.713 in the receiver operating characteristic curve. The findings provide novel insights into the immunological mechanism in LUAD biology and in predicting the prognosis of LUAD patients.

Automated summary

This study reveals the transcriptome characteristics in LUADs with different immune phenotypes and identifies key genes and non-coding RNAs associated with LUAD immune phenotypes. Gene coexpression networks and prognostic signature models are established. The findings provide novel insights into the immunological mechanism in LUAD biology and predicting the prognosis of LUAD patients.