Use of decoys to optimize an all-atom force field including hydration
Yelena A. Arnautova 1 and Harold A. Scheraga 1*
1 Cornell University
* To whom correspondence should be addressed. E-mail: has5{at}cornell.edu.
Submitted on March 14, 2008
Revised on April 14, 2008
Accepted on 7 May 2008
 |
Abstract |
|---|
A novel method of parameter optimization is proposed. It makes use of large sets of decoys generated for six non-homologous proteins with different architecture. Parameter optimization is achieved by creating a free energy gap between sets of nativelike and non-native conformations. The method is applied to optimize the parameters of a physics-based scoring function consisting of the all-atom ECEPP05 force field coupled with an implicit solvent model (a solvent accessible surface area model). The optimized force field is able to discriminate near-native from non-native conformations of the six training proteins when used either for local energy minimization or for short Monte Carlo simulated annealing runs that follow local energy minimization. The resulting force field is validated with an independent set of six non-homologous proteins and appears to be transferable to proteins not included in the optimization; i.e., for five out of the six test proteins, decoys with 1.7-4.0 Å all-heavy-atom root-mean-square deviation (rmsd) emerge as those with the lowest energy. Additionally, we examined the set of misfolded structures created by Park and Levitt with a four-state reduced model [Park B, Levitt M. J Mol Biol 1996;258:367-392]. The results from these additional calculations confirm the good discriminative ability of the optimized force field obtained with our decoys sets.
Key Words:
parameter optimization, protein, solvation, surface area model
