The exploitation of assembly language instructions in biological text manipulation: II. Amino acid sequences

Amino acid residues may be divided into groups according to similarity of function, or evolutionary history, or other useful criteria. A grouping of amino acids into the eight sets based upon functionality allows a representation involving a three-bit code that can be of value in string matching searches. An amino acid residue may be identified uniquely by employing a further two bits. We propose that amino acid sequence data and search strings be preprocessed to form strings of highest bits, strings of the next highest bits, and so on. Machine assembly language instructions on the separate bit-strings provide a hierarchical measure of homology. We study a number of preprocessing strategies arranged to accord with the kind of search contemplated.