PETcofold: predicting conserved interactions and structures of two multiple alignments of RNA sequences

• DOI: 10.1093/bioinformatics/btq634
• Publication Date: 2010-11-19T03:58:31Z
• Authors: Stefan E. Seemann; Andreas S. Richter; Tanja Gesell; Rolf Backofen; Jan Gorodkin
• Keywords: Nucleic acid structure

Predicting RNA-RNA interactions is essential for determining the function of putative non-coding RNAs. Existing methods prediction are all based on single sequences. Since comparative have already been useful in RNA structure determination, we assume that conserved also imply function. Of these, further a non-negligible amount existing acquired compensating base changes throughout evolution. We implement method, PETcofold, can take covariance information intra-molecular and inter-molecular pairs into account to predict secondary structures two multiple alignments sequences.PETcofold's ability was evaluated carefully curated dataset 32 bacterial small RNAs their targets, which manually extracted from literature. For evaluation both interaction prediction, were able extract only few high-quality examples: one vertebrate nucleolar four these show be improved by our approach. Furthermore, PETcofold controlled data with phylogenetically simulated sequences enriched patterns at sites. observed increased performance amounts covariance.The program available as source code downloaded http://rth.dk/resources/petcofold.