Construction of a potential long noncoding RNA prognostic model involved competitive endogenous RNA for patients with...

Increasing evidence has underscored the role of long noncoding RNAs (lncRNAs) make up the major proportion of the competing endogenous RNAs (ceRNAs) network and can regulate gene expression by competitively binding to miRNAs in the development and progression of tumors. Nevertheless, the role of lncRNA-mediated ceRNAs in gastric cancer (GC) and their regulatory mechanisms have not been elucidated to some extent. This study is aimed at constructing a prognostic risk model for GC based on lncRNAs. A TCGA (The Cancer Genome Atlas) dataset was analyzed using edgeR to identify differentially expressed lncRNAs (DElncRNAs) in GC tissues vs normal tissues. Subsequently, DElncRNAs that could predict GC prognosis were determined using a training set. A prognostic risk model based on the DElncRNAs was then constructed. The performance of the model was tested using a test set. The functions of these lncRNAs in GC were investigated using a lncRNA-miRNA-mRNA network. Analysis of lncRNA expression in 407 TCGA GC cases identified 3 lncRNAs that significantly correlated with prognosis. GC cases with high-risk scores showed markedly poor prognosis relative to those with low-risk scores in both the training and test sets. Univariate and multivariate Cox regression analysis of the relationship between various clinical features and prognosis found that these lncRNAs and stage significantly correlated with GC prognosis. A lncRNA-miRNA-mRNA network based on 3 lncRNAs and functional enrichment analysis of interacting mRNA indicated that these genes are enriched in various intracellular receptor signaling pathways, including regulation of muscle system process, and protein deubiquitylation. The current study provides novel insights into the lncRNA-related ceRNA network in GC and sheds lights on underlying 3 lncRNA biomarkers may be independent prognostic signatures in predicting the survival of GC patients.