Similarity of symbolic sequences

TL;DR

This paper introduces a novel similarity measure for symbolic sequences based on their common subsequences, effective for both short and very long sequences, with potential applications in evolutionary biology.

Contribution

It proposes a new similarity measure using sequence decomposition into patterns, improving analysis of symbolic sequences like DNA and proteins.

Findings

Effective for sequences of varying lengths

Applicable to nucleotide and protein sequences

Potential to trace evolutionary relationships

Abstract

A new numerical characterization of symbolic sequences is proposed. The partition of sequence based on Ke and Tong algorithm is a starting point. Algorithm decomposes original sequence into set of distinct subsequences - a patterns. The set of subsequences common for two symbolic sequences (their intersection) is proposed as a measure of similarity between them. The new similarity measure works well for short (of tens letters) sequences and the very long (of hundred thousand letters) as well. When applied to nucleotide or protein sequences may help to trace possible evolutionary of species. As an illustration, similarity of several sets of nucleotide and amino acid sequences is examined.