Integrated analysis of differentially expressed profiles and construction of a competing endogenous long non-coding RNA network in renal cell carcinoma

BACKGROUND: Long non-coding RNAs (lncRNAs) play crucial roles in the initiation and progression of renal cell carcinoma (RCC) by competing in binding to miRNAs, and related competitive endogenous RNA (ceRNA) networks have been constructed in several cancers. However, the coexpression network has been poorly explored in RCC. METHODS: We collected RCC RNA expression profile data and relevant clinical features from The Cancer Genome Atlas (TCGA). A cluster analysis was explored to show different lncRNA expression patterns. Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses and gene set enrichment analysis (GSEA) were performed to analyze the functions of the intersecting mRNAs. Targetscan and miRanda bioinformatics algorithms were used to predict potential relationships among RNAs. Univariate Cox proportional hazards regression was conducted to determine the RNA expression levels and survival times. RESULTS: Bioinformatics analysis revealed that the expression profiles of hundreds of aberrantly expressed lncRNAs, miRNAs, and mRNAs were significantly changed between different stages of tumors and non-tumor groups. By combining the data predicted by databases with intersection RNAs, a ceRNA network consisting of 106 lncRNAs, 26 miRNAs and 69 mRNAs was established. Additionally, a protein interaction network revealed the main hub nodes (VEGFA, NTRK2, DLG2, E2F2, MYB and RUNX1). Furthermore, 63 lncRNAs, four miRNAs and 31 mRNAs were significantly associated with overall survival. CONCLUSION: Our results identified cancer-specific lncRNAs and constructed a ceRNA network for RCC. A survival analysis related to the RNAs revealed candidate biomarkers for further study in RCC.