Identification of prognostic biomarkers of prostate cancer with long non-coding RNA-mediated competitive endogenous RNA network

Prostate cancer (PCa) is the second most frequently diagnosed male cancer, and no treatments exist for effective inhibition of metastatic spread of PCa. Long non-coding RNA (lncRNA) plays key roles in pathogenesis and development of various cancers through competing with endogenous RNAs (ceRNAs), but at present research on lncRNA functions in PCa is still very limited. Hence, this aspect was investigated using bioinformatics methods. Firstly, the functional lncRNA-mediated ceRNA network associated with PCa was constructed by the multi-step computational approach. Then the cytoscape software was used to analyze the node degree and betweenness centrality (BC) value of lncRNAs and mRNAs in the interaction. Finally, the lncRNAs were screened in the central region of the network by the node degree and BC value, and the functional enrichment of mRNAs was evaluated with the Gene Ontology (GO) database. In our results, LINC00476, MALAT1, SNHG11, LINC00649, and ILF3-AS1 are the lncRNAs which have the most nodes and higher BC values and considered as prognostic markers in PCa. GO analysis suggested that the function of screened lncRNAs was obviously focused on intracellular receptor signaling pathway, which indicated these lncRNAs might be potential biomarkers for diagnosis, evaluation and gene-targeted therapy of PCa.