Energy minimization of protein tertiary structures by local search algorithm and parallel simulated annealing using genetic crossover

In this paper, the local search algorithm to improve the searching capability of parallel simulated annealing using genetic crossover (PSA/GAc) for the energy minimization of protein tertiary structures is proposed. Our previous research shows that PSA/GAc is effective for the energy minimization of the small proteins. However, because the energy minimization of larger proteins requires larger number of calculations required to reach the global optimum. In this paper, the local search algorithm to search α-helix efficiently is proposed and is applied to the energy minimization of proteins. Also, for the verification of the algorithm, the test function which has a similar characteristic to the energy functions of proteins that have α-helix structures is proposed. Finally, PSA/GAc with the proposed local search is applied to the same proteins and its capability is discussed. The result indicates that as for the target proteins of this paper, PSA/GAc with local search has obtained the more accurate solutions and additionally, total number of evaluations required to reach the optimum can reduced. From the results, the possibility of effectiveness of proposed local search algorithm on the energy minimization of the proteins with α-helix has been verified.