Get access

Studies of molecular docking between fibroblast growth factor and heparin using generalized simulated annealing

Authors

  • Samuel Silva da Rocha Pita,

    Corresponding author
    1. Laboratório de Modelagem e Dinâmica Molecular (LMDM), Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro (UFRJ), Brazil
    • Laboratório de Modelagem e Dinâmica Molecular (LMDM), Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro (UFRJ), Brazil
    Search for more papers by this author
  • Tácio Vinício Amorim Fernandes,

    1. Laboratório de Modelagem e Dinâmica Molecular (LMDM), Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro (UFRJ), Brazil
    Search for more papers by this author
  • Ernesto Raul Caffarena,

    1. Programa de Computação Científica (ProCC), Fundação Oswaldo Cruz, Rio de Janeiro, Brazil
    Search for more papers by this author
  • Pedro Geraldo Pascutti

    1. Laboratório de Modelagem e Dinâmica Molecular (LMDM), Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro (UFRJ), Brazil
    Search for more papers by this author

Abstract

Since the middle 70s, the main molecular docking problem consists in limitations to treat adequately the degrees of freedom of protein (or a receptor) due to the energy landscape roughness and the high computational cost. Until recently, only few algorithms considering flexible simultaneously both ligand and receptor at low computational cost were developed. As a recent proposed Statistical Mechanics, generalized simulated annealing (GSA) has been employed at diverse works concerning global optimization problems. In this work, we used this method exploring the molecular docking problem taking into account the FGF-2 and heparin complex. Since the requirements of an efficient docking algorithm are accuracy and velocity, we tested the influence of GSA parameters qA (new configuration acceptance index), qV (energy surface visiting index), and qT (temperature decreasing control) on the performance of GSADOCK program. Our simulations showed that as temperature parameter qT increases, qA parameter follows this behavior in the interval ranging from 1.1 to 2.3. We found that the GSA parameters have the best performance for the qA values ranging from 1.1 to 1.3, qV values from 1.3 to 1.5, and qT values from 1.1 to 1.7. Most of good qV values were equal or next the good qT values. Finally, the implemented algorithm is trustworthy and can be employed as a tool of molecular modeling methods. The final version of the program will be free of charge and will be accessible at our home-page or could be requested to the authors for e-mail. © 2008 Wiley Periodicals, Inc. Int J Quantum Chem, 2008

Ancillary