Protein Force Field Improved by using a New Backbone-Torsion-Energy Term

Abstract

Bimolecular simulations are performed by using potential energy functions with force-field parameters such as AMBER, CHARMM, OPLS, GROMOS, and ECEPP. We have performed detailed comparisons of three version of AMBER (ff94, ff96, and ff99), CHARMM22, OPLS-AA/L, and GROMOS96 by generalized-ensemble simulations of two small peptides [1]. As results of these comparisons, we saw that these force fields showed clearly different behavior of peptides, especially, about secondary-structure-forming tendencies. These results imply that it is necessary to refine and improve the existing force-field parameters. Among the energy terms, the torsion-energy term is the most problematic. We have proposed force-field refinement methods. This method consists of minimizing the sum of the square of the force acting on each atom in the proteins with the structures from the Protein Data Bank (PDB) [2]. Additionally, we also proposed a new backbone-torsion-energy term, which is represented by a double Fourier series in two variables, the backbone dihedral angles phi and psi [3,4].In this poster, we apply our optimization method and backbone-torsion-energy term to AMBER ff94(ff96) force field for molecular simulation of protein systems. The result implies that the new force-field parameters give structures of two peptides more consistent with the experimental implications for the secondly-structure-forming tendencies than the original force field.