(663e) A Parameterization of the Generalized Born Model for Simulations with Only Torsional Degrees of Freedom | AIChE

(663e) A Parameterization of the Generalized Born Model for Simulations with Only Torsional Degrees of Freedom

Authors 

Curtis, R. - Presenter, The University of Manchester
Magee, J. - Presenter, University of Manchester


In all-atomistic simulations of peptides, the amount of phase space to be sampled can be reduced by constraining all but torsional degrees of freedom. Most forcefields have been developed for non-constrained systems, with a prominent exception being ECEPP. Polar solvation and screening effects are incorporated into ECEPP using solvent accessible surface area (SASA) models. A recent, more rigorous method to incorporate these effects is through use of Generalized Born (GB) models. In this work, a GB model is parameterized for the ECEPP forcefield by matching the electrostatic solvation free energy with that calculated from solving the Poisson equation for a trial set of energy-minimized amino acid structures. We compare results determined from simulations of homopolypeptides using the GB model with those determined using various SASA models and distance dependent dielectric. Further, different methods for incorporating salt effects into the GB model are investigated. Finally, we use Monte-Carlo simulations to determine the effect of salt concentration on the structures of self-assembling peptides such as poly-lysine and the octapeptide, AEAEAKAK.