Prediction of Residues in Protein-RNA Interaction Sites by Neural Networks

Abstract

Structural studies of protein-RNA interactions have been focused on analyzing intraand intermolecular interactions or recognizing binding patterns [3, 4, 5]. It has been examined that specific nucleotide-protein secondary structures are favored in protein-RNA interactions. The primary focus of our study is to investigate a predicting problem of the protein-RNA interaction sites upon a protein chain. A prediction system based on artificial neural networks was developed to identify accurate residues involved in protein-RNA interaction sites from combining the residue sequence and the protein secondary structure information.