Optimisation of miRNA-mRNA relationship prediction using biological features.

MicroRNAs are small (approx. 22nt) non-coding RNAs that regulate the expression of genes by either degrading messenger-RNA (mRNA) that has already been transcribed or by repressing the translation of mRNA, thus inhibiting protein production. This mechanism of gene regulation by binding of the miRNA to 3-prime-untranslated region of target mRNAs has been recently discovered. This sequence-specific post-transcriptional gene regulation process affects large set of genes involved in number of biological pathways. Mapping of 7nt long miRNA seed sequence to the target gene has been a standard way of predicting miRNA targets. In this study, we develop a framework to enrich the human miRNA-mRNA relationship based on genomic and structural information.