Biocomputational prediction of small non-coding RNAs in Streptomyces coelicolor

Josef Pánek, Jan Bobek, Karel Mikulík & Jiří Vohradský

Laboratory of Bioinformatics, Institute of Microbiology, Academy of Sciences of the Czech Republic, Prague

A prediction of small non-coding RNAs (ncRNAs, sRNAs) in gram-positive differentiating bacteria of Streptomyces is presented. The prediction is based on sequence conservation within intergenic regions (IGRs) of S. coelicolor and S. avermitilis genomes, and predicted termination factors.

58 novel Streptomyces sRNAs were predicted and tested for expression. 15 of them have been shown experimentally to be expressed during cell development at standard conditions.

Among the predicted Streptomyces sRNAs, m1-like, 4.5S-like and spot42-like RNA candidates were identified computationally using sequence and / or structural similarity with known sRNAs of different bacteria. The already known Streptomyces tmRNA was also identified as a positive control.

As termination factors for computational identification of the sRNAs, either predicted Rho-independent terminators or new putative transcription terminators were used. The new termination factor was proposed as it most likely terminates m1-like sRNA and other two sRNAs that were computationally identified and experimentally validated with high reliability.