Parallel Recombinative Simulated Annealing Technique for PSP using the Subsequence Secondary Structures: A Novel Protein Fold Lock Model

Abstract

The Subsequence Secondary Structures are a set of regular patterns (common motif) frequently occurring in the amino acid sequence of proteins [9]. These motifs are seen to be conserved among species, thus, forming a group of functionally related sequences, sharing a specific biological function. Determination of compact groups of these common patterns is a prerequisite to ably simulate a protein structure prediction algorithm (PSP). In this work, a novel parallel recombinative simulated annealing technique is proposed for an efficient prediction of minimum energy structure on a HP lattice model. The implementation comprising of hybrid heuristics, combines the population based evolutionary Genetic Algorithm and the cooling scheme of Simulated Annealing amid provision of protein subsequence secondary fold lock. A GUI implementation is done for user based parameter testing with 5 lock options={α1,α2,β,loop,core} derived from [5] [7]. The empirical results are compared with the existing EMC method [5] on similar datasets. It is found that the PRSA technique outperforms its ancestor with drastic improvement in computational speed to reach the minimum energy protein structure.

Keywords: Protein Secondary Structure, Parallel Recombinative Simulated Annealing, Protein Fold Lock Model, Genetic Algorithm, Simulated Annealing.