Bioinformatical recognition of Potassium-channel sequences.
Burkhard Heil (1), Jost Ludwig (1), Thomas Lengauer (2), Hella Lichtenberg-Fraté (1)
(1) IZMB, Molekulare Bioenergetik, Universität Bonn, Kirschallee 1, Bonn, 53115, Germany;
(2) Max Planck Institut für Informatik Stuhlsatzenhausweg 85 , 66123 Saarbrücken, Germany
Potassium (K+-)channels comprise a large and diverse class of membrane proteins with a great variety of cellular functions. K+-channel-subunits consist of at least two transmembrane regions framing the ion-conducting pore-domain, the only conserved region among all K+-channels. At the extracellular opening of the pore a 'selectivity-filter' provides the K+ specificity. Here we introduce a new bioinformatical method based on a distinctive signature and a new algorithm that analyses physico-chemical properties of given residues. The signature comprises 25 amino acids covering the pore region and the selectivity filter. In contrast to conventional sequence signature based methods, it represents a broad spectrum of amino acids that may appear at a given position instead of using only very few but highly conserved residues. On the basis of this signature the algorithm creates in a second step a highly-specific string, detailing the properties of the pore region and the selectivity filter. Identified hits represent rather a match to a stringent order of properties than a biased amino acid sequence. The method was validated using cross-validation and statistical analysis of the results. An application approach on yeast genomes identified undoubtlessly in S. cerevisiae the TOK1 channel and not the transporters TRK1,2 despite their conserved K selectivity. Results will be presented concerning similar queries on the S. pombe genome. The project is funded by the EC QLK3-2001-00401.
Return to YGM 2004 Home at SGD