Phi-psi conformational pattern clustering of protein amino acid residues using the potential function method.

This paper describes phi-psi conformational pattern clustering of protein amino acid residues. The method is based on the potential function method and mode seeking technique. phi-psi conformational pattern distribution maps and their three-dimensional potential surface maps were computed for 20 different kinds of amino acid residues, using 67 proteins (14,723 residues) taken from the Protein Data Bank. It was found that glycine residues of proteins have five major clusters in the phi-psi conformational space, and two or three major clusters were found for the other kinds of residues. Mode seeking on the potential surface of glycine residues identified their representative phi-psi conformational patterns: (82.64 degrees, 9.84 degrees), (-62.63 degrees, -41.71 degrees), (-85.33 degrees, 176.78 degrees), (91.88 degrees, 178.14 degrees) and (167.06 degrees, -175.33 degrees). The details of the method and the results are discussed with new classifications of phi-psi conformational patterns of amino acid residues of proteins.