Comprehensive Analysis of eRNAs and Genes Involved in Gastric Cancer and Constructing a Prognostic Model Predicting the Overall Survival of Patients

Gastric cancer (GC) is one of the most prevalent cancers and the first cause of cancer-related deaths in Iran. These patients are diagnosed at advanced stages, which leads to poor prognosis because there is inadequate knowledge of molecular factors to provide diagnostic/prognostic biomarkers. The aim of the current study was the comprehensive analysis of enhancer RNAs (eRNAs) to determine their molecular interactions with other genes and propose a survival-related model for patients.

RNA-sequencing raw data of GC patients were downloaded from The Cancer Genome Atlas (TCGA), and differentially-expressed eRNAs and genes in tumor compared to non-tumor samples were extracted. The target genes of each eRNA were then identified based on physical distance and correlation and a regulatory network was constructed with these elements. A prognostic model for predicting the overall survival of patients was eventually established by performing univariate and multivariate Cox regression analyses, and the performance of the model was surveyed.

By performing differential expression analysis, 69 and 2606 differentially-expressed eRNAs and genes were extracted, respectively and by identifying the relationships between these elements, a regulatory network consisting of 84 nodes and 119 edges was constructed. A three-components' survival-related model subsequently was established which had the ability to predict patients' outcomes.

Based on the results, the 3-component constructed model, including an eRNA and two other genes, can be considered as a possible prognostic tool; however, further research is needed to clinically implement it.