A new greedy heuristic for 3DHP protein struture prediction with side chain

Abstract

In spite of the fact that many models of protein structure prediction have been proposed and have also been widely studied in the last years, little attention has been given to the discrete models with side chains. Few papers present algorithms that try to predict the 3 dimensional structures of protein from their amino acid sequences represented by a backbone and the side chains (hydrophobic or hydrophilic). In this paper, we propose a new greedy heuristic with a pull-move set for finding these structures to the 3DHP-SC model, i.e. for a three-dimensional model on a cubic lattice, with side chains. To demonstrate the performance of our method, we have used 25 benchmark instances from the literature. For the instances tested, the proposed technique matched the best known results for 12 instances and obtained better results for the other 13. The computational resources that we have used have been relatively limited in comparison with other studies in the literature, and the quality of our results shows the potential of the approach both in terms of quality and total computation time.