Searching for microRNA prostate cancer target genes

MicroRNAs are a class of short, non-coding RNA regulating the expression of genes involved in several biological processes. This paper addresses the computational identification of target genes for microRNAs, in particular those involved in the development of prostate cancer. The available tools, in particular the miRanda program, need to be validated against biological evidence. We identify several refinement of the algorithm, and provide an optimization of the parameters involved, by means of a genetic algorithm, so as to maximize the adherence of results to biological evidence. The increased selectivity of the method may also be used to guide experimental validation in a more focused direction.