Open Access

Protein structure prediction with the 3D-HP side-chain model using a master–slave parallel genetic algorithm

  • César Manuel Vargas Benítez1Email author and
  • Heitor Silvério Lopes1
Journal of the Brazilian Computer Society201016:2

DOI: 10.1007/s13173-010-0002-6

Received: 29 September 2009

Accepted: 23 February 2010

Published: 24 April 2010

Abstract

This work presents a master-slave parallel genetic algorithm for the protein folding problem, using the 3D-HP side-chain model (3D-HP-SC). This model is sparsely studied in the literature, although more expressive than other lattice models. The fitness function proposed includes information not only about the free-energy of the conformation, but also compactness of the side-chains. Since there is no benchmark available to date for this model, a set of 15 sequences was used, based on a simpler model. Results show that the parallel GA achieved a good level of efficiency and obtained biologically coherent results, suggesting the adequacy of the methodology. Future work will include new biologically-inspired genetic operators and more experiments to create new benchmarks.