Accurate protein folding software to predict ligand-induced conformational changes in G protein-coupled receptors

Period of Performance: 09/01/2016 - 06/30/2017


Phase 2 SBIR

Recipient Firm

Dnastar, Inc.
Madison, WI 53705
Principal Investigator


Abstract In humans and animals, G protein-coupled receptors (GPCRs) are embedded on cell surfaces and function as key regulators of physiological events by transmitting signals from extracellular stimulants across the cell membrane into the cell. Impaired or abnormal GPCR function can result in disordered physiological processes causing a broad and diverse range of diseases. For this reason, GPCRs are targeted for therapeutic intervention by over 40% of the FDA-approved drugs on the market as well as many of those in development today. However, biomedical studies of GPCRs are hampered by a lack of atomic structures due to the difficulty of experimentally determining membrane protein structures. Thus far, drug discovery efforts have gone without the benefit of software tools that can accurately model the 3D structures of GPCRs at the atomic-level. Such tools, if they existed, could provide an in-depth understanding of how new drugs will interact with GPCRs and support the ability to predict their therapeutic benefit. Here we propose to advance GPCR drug discovery by developing highly accurate software tools built on the success of the iterative threading assembly refinement (I- TASSER) algorithm for protein structure prediction. This proposal seeks to develop new computational methodologies for the accurate, comprehensive, and more powerful generation of atomic-level GPCR models. The aims of the project focus on developing more accurate and effective GPCR structure predictions through better modeling of the challenging loop domains, which play important but often poorly understood roles in GPCR function, and of the overall structural changes induced by drug or ligand binding, which controls GPCR signal transmission into the cell. In particular, overall predictive capability will be improved by incorporating new concerted loop modeling and transmembrane helix packing methods into full-chain GPCR structure predictions, and also by introducing ligand-binding interactions into the fully flexible ligand-GPCR complex structure assembly simulations when constructing the structure of a ligand bound to its receptor. Additionally, we will replace software dependencies that impede the commercial distribution of this GPCR structure prediction tool, thereby accelerating pharmaceutical structure-based drug design. The project goal is to deliver advanced GPCR structure prediction software that is powerful, accurate, and easy to use for both academic and commercial use, which will accelerate GPCR drug discovery by enabling, for the first time, detailed and accurate GPCR structure predictions.