MicroRNAs (miRNAs) are small, non-coding RNAs of ~23 nt that repress the translation of messenger RNAs by basepairing with their 3' UTRs. The mature RNA species are derived from primary transcripts that can fold into hairpin secondary structures for recognition and cleavage by the RNase III enzymes Drosha and Dicer. While other requirements of the primary transcripts are somewhat mysterious, all miRNAs derive from the hairpin portions of those transcripts. MicroRNA gene annotation has therefore been simplified as the identification of a mature miRNA sequence along with definition of the hairpin from which it is derived.
Mirscan is intended to identify the most microRNA-like hairpin sequences from an arbitrary set of candidates. It does so by comparing a set of features in a foreground set of previously annotated miRNA hairpins to those features of a background set of candidates. Those candidates that most resemble the foreground set can be retained, while the rest can be discarded. This process of discarding candidates from a background set that bear the least resemblence to the foreground set can be repeated iteratively, as each repetition will change the properties of the background set and thereby adjust the relative importance of various features in distinguishing between foreground and background.
Software and documentation written by J. Graham Ruby. For further discussion, see the supplemental methods document that accompanies Ruby et al., "Evolution, biogenesis, expression, and target predictions of a substantially expanded set of Drosophila microRNAs". Genome Research, 2007.