Recognition of Patterns in Protein Sequences by Metric Analysis

Abstract

Metric analysis is a series of mathematical procedures useful for assigning the order of fragments generated during protein sequence determination and for measuring the similarity of the final amino acid sequence to other sequences. This analysis uses theoretically proven combinatorial algorithms that are implemented by digital computer programs. For a given sequence, the entire specified sequence is denoted as S and the set of all unspecified intervals present in S is designated as U. Algorithm SS calculates the evolutionary distance and all metric alignments between two specified sequences (1). Evolutionary distance is a metric function defined as the minimum number of nucleotide mutations, insertions, and deletions needed to interconvert two nucleic acid sequences that could code for the amino acid sequences. Algorithm SU finds all of the locally best ways to align a specified sequence (a pattern) with unspecified intervals of a longer sequence (2,3). Algorithm UU identifies the locally best ways to align unspecified intervals of one sequence with those of another (3).