Improved methods for predicting the structure and function of G protein-coupled receptors

COMP 91

Ravinder Abrol and William A Goddard III, wag@wag.caltech.edu. Materials and Process Simulation Center, California Institute of Technology, MC 139-74, Beckman Institute, Pasadena, CA 91125
G protein-coupled receptors (GPCRs) are intrinsic membrane proteins with 7 transmembrane helices, and modulated by diverse bioactive molecules (biogenic amines, peptides, lipids) to regulate essential physiological processes (neurotransmission, metabolism, secretion, growth, immunity). Thus, GPCRs have been implicated in all major disease areas. Development of drugs with reduced side effects has been hampered by the availability of 3D structure for only one GPCR. Our research group has been developing computational methods based on first principles to predict the structure and function of GPCRs. The methods have been validated for many systems against experimental mutagenesis data. We will present some of the recent method developments aimed at more accurate and faster predictions of GPCR structure and function. These include the use of implicit membrane (multi-dielectric) solvation for scoring monte carlo generated helical packings, accurate placement of protein side chains, use of thermodynamic hydrophobicities, and generation of ligand rotamer libraries for flexible docking.