PREDICTION METHOD AND APPARATUS FOR A SECONDARY STRUCTURE OF PROTEIN

A prediction method and apparatus for secondary structures of a protein in which accuracy in prediction of a formation of a .beta.-sheet is increased and which can applied to any type of protein including an .alpha.-helix and a .beta.-sheet. Formation of the .alpha.-helix is predicted with respect to each amino acid residue in a sequence of amino acid residues. Then, formation of the .beta.-sheet is predicted with respect to all pairs of residues which were not predicted to form the .alpha.-helix. Results of prediction of the .alpha.-helix and the .beta.-sheet are combined to obtain a result of prediction of the secondary structure of the protein.