DuplexFinder: predicting the miRNA-miRNA* duplex from the animal precursors

Abstract

It is an important work to search unknown microRNA gene from genome or EST sequences. There are several ab initio methods to classify microRNA precursors from random hairpins. However, the exact position of the mature part in the precursor remains unknown only from these ab initio classification methods. In this paper, we propose four significant position distribution features, and then give a computational method to predict the miRNA?miRNA* duplex from the animal precursors. It is helpful for classifying miRNA precursor and detecting unknown miRNA genes.