An artificial intelligence approach to motif discovery in protein sequences: application to steriod dehydrogenases

MEME (Multiple Expectation-maximization for Motif Elicitation) is a unique new software tool that uses artificial intelligence techniques to discover motifs shared by a set of protein sequences in a fully automated manner. This paper is the first detailed study of the use of MEME to analyse a large, biologically relevant set of sequences, and to evaluate the sensitivity and accuracy of MEME in identifying structurally important motifs. For this purpose, we chose the short-chain alcohol dehydrogenase superfamily because it is large and phylogenetically diverse, providing a test of how well MEME can work on sequences with low amino acid similarity. Moreover, this dataset contains enzymes of biological importance, and because several enzymes have known X-ray crystallographic structures, we can test the usefulness of MEME for structural analysis. The first six motifs from MEME map onto structurally important alpha-helices and beta-strands on Streptomyces hydrogenans 20beta-hydroxysteroid dehydrogenase. We also describe MAST (Motif Alignment Search Tool), which conveniently uses output from MEME for searching databases such as SWISS-PROT and Genpept. MAST provides statistical measures that permit a rigorous evaluation of the significance of database searches with individual motifs or groups of motifs. A database search of Genpept90 by MAST with the log-odds matrix of the first six motifs obtained from MEME yields a bimodal output, demonstrating the selectivity of MAST. We show for the first time, using primary sequence analysis, that bacterial sugar epimerases are homologs of short-chain dehydrogenases. MEME and MAST will be increasingly useful as genome sequencing provides large datasets of phylogenetically divergent sequences of biomedical interest.