Folding Trp-cage to atomic resolution: Searching parameter space for a free-energy minimum

COMP 180

Olgun Guvench, guvench@scripps.edu and Charles L. Brooks III. Department of Molecular Biology, TPC6, The Scripps Research Institute, 10550 N. Torrey Pines Rd, La Jolla, CA 92037
We present an all-atom implicit-water protein force-field and its parameterization using a self-consistent approach, with the target of parameterization being the folding of the 20-residue mini-protein Trp-cage. The force-field includes interaction terms that are all short-ranged in nature, making it computationally expedient. The terms accounting for intramolecular and protein water hydrogen bonding, as well as the hydrophobic effect, are all of the form k*f(x,) where f(x) is a function of the atomic coordinates and k is an energetic parameter subject to optimization. Using a self-consistent Z-score optimization protocol in conjunction with replica-exchange molecular dynamics (REMD) for enhanced thermodynamic sampling, we show that it is possible to start with a set of parameters (k's) that lead to rapid denaturation of the protein at laboratory temperature and to algorithmically optimize these parameters such the native state becomes a 2 kcal/mol free-energy minimum at laboratory temperature.