Computational Derivation of the Human microRNA Regulome

Abstract

MicroRNAs are known to have pervasive effects on gene expression and exercise post-transcriptional control over most eukaryotic genomes. Even though several miRNA functions have already been documented, their mechanisms of action as a whole remain unclear. In this study, 1,213 genes were found to be targeted by 1,048 human rniRNAs obtained from miRBase. They were computationally identified from the intersection of results of three prediction algorithms in miRWalk: miRanda, PicTar and TargetScan. The genes were filtered to include only those with targeted frequency of at least 5, yielding 126 gene targets. These were then compared for common genes with experimentally validated gene targets, generating 60 genes. Using DAVID, the involvement of these 60 genes in the overall function of the human miRNA regulome was elucidated. From groups of enriched GO terms with enrichment scores greater than 1.3, it was found that the miRNA regulome frequently targets genes which are mostly involved in transcription regulation, along with a substantial portion involved in axial determination during embryogenesis. These findings suggest that rniRNAs may be involved in the evolution of complex life.