Exact algorithm for segment based sequence alignment

Abstract

In bioinformatics, proteins, large biological molecules consisting long chain of amino acids, are described by a sequence of 20 amino acids. To analyze these protein sequences pairwise alignment is being used, which identify regions of similarity that may be a result of structural, functional and/or evolutionary relationship between them. Traditional pairwise alignment algorithms work on residue level; it does not account structural or functional information that protein carries. A new approach for protein sequence analysis is being proposed here, pairwise alignment of two protein sequences based on segments. Segments of the sequences can be formed on the basis of protein feature, i.e., functional sites or secondary structure of the protein. Each segment carries a type and weight for the alignment process. Algorithm should align two sequences such that segments with weight higher than threshold value must align with the similar type of segment and score for the alignment must be maximal for given scoring function. Here, we are proposing a generic framework to understand, explore and experiment proteins based on their features, i.e. structure, function, and evolution.