A Seven Immune-Related lncRNAs Model to Increase the Predicted Value of Lung Adenocarcinoma
Jianping Li, Rui Li, Xiao Liu, Huo Chen, Tingting Liu, Jie Yao, Yi-Qing Qu
Abstract
Background: Recent researches showed that immune-related lncRNA plays a crucial part in tumor immune microenvironment. This study tried to identify immune-related lncRNAs and construct a robust prediction model to increase predicted value of lung adenocarcinoma. Methods: RNA expression data of lung adenocarcinoma (LUAD) were download from The Cancer Genome Atlas (TCGA) database. Immune genes were acquired from The Molecular Signatures Database (MSigDB). The immune gene related lncRNAs were acquired by “limma R” package and Cytoscape3.7.1. Cox regression analysis were applied to construct this forecast model. The prognostic model was validated by the testing cohort which was acquired by bootstrap method. Results: A total of 551 lncRNA expression profiles including 497 LUAD tissues and 54 non-LUAD tissues were obtained. 331 immune genes were acquired. The result of Cox regression analysis showed that seven lncRNAs (AC022784-1, NKILA, AC026355-1, AC068338-3, LINC01843, SYNPR-AS1 and AC123595-1) can be performed to construct the predictive model to forecast the prognosis of LUAD. Kaplan-Meier curves indicated that our predicted model can distribute LUAD patients into two different risk groups (high and low) with significant statistical significance (P = 1.484e−07). Cox analysis and independent analysis illustrated that the seven-lncRNAs predicted model was an isolated factor by comparing with other clinical variables. We validated the accuracy of our model in the testing dataset. Furthermore, the prognostic model also showed higher predictive efficiency than three published prognostic models. The two different survival groups represented diverse immune feature according to principal components analysis (PCA). GSEA analysis (Gene set enrichment analysis) indicated that seven-lncRNAs signature may be involved in the progress of tumorigenesis. Conclusions: We have established a seven-immune related lncRNAs predicted model. This prognostic model had significant clinical significance to increase predicted value and guide the personalized treatment for LUAD patients.