PNNL

Abstract

Peptoids (N-substituted glycines) are a class of biomimetic polymers that have attracted significant attention due to their accessible synthesis and enzymatic and thermal stability relative to their naturally occurring counterparts (polypeptides). While these polymers provide the promise of more robust functional materials via hierarchical approaches, they present a new problem for the computational prediction of structure for material design. For proteins, structure prediction based on sequence and de novo design has made dramatic progress in recent years. The reliability of calculations hinges on the accuracy of interactions represented in the force field used to model peptoids. Current efforts to develop and implement peptoid specific force fields are spread out resulting in a replication of efforts and a fragmented collection of parameterized sidechains. Here, we developed a peptoid specific force field containing 70 different side chains, using GAFF2 as starting point. The new model is validated based on the generation of Ramachandran-like plots from DFT optimization compared against force field reproduced potential energy and free energy surfaces, as well as the reproduction of equilibrium cis/trans values (Kct) for some residues experimentally known to form helical structures. Kct distributions are estimated for all parameterized residues to identify which residues have an intrinsic propensity to cis or trans structures in the monomeric state.