Toward the development of an accurate virtual screening protocol: Recent improvements in protein ensemble docking

COMP 163

Luciano Mueller, luciano.mueller@bms.com, Department of Macromolecular Structure, Bristol Myers Squibb Pharmaceutical Research Institute, Route 206 & Provinceline Road, P. O. Box 4000, Princeton, NJ 08543-4000 and Daniel L. Cheney, daniell.cheney@bms.com, Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 4000, Princeton, NJ 08543.
In earlier reports we have shown that using multiple, representative conformations of a target protein significantly enhanced sampling over a range of protein targets and ligand chemotypes, and using a variety of docking programs in “high accuracy” mode, and that refinement and re-scoring of the “raw” complexes lead to significant improvements in the ability to predict accurately the binding pose. In the present study, we examine the results of protein ensemble docking using an expanded dataset of ligands and targets, and also examine alternative methods of scoring and protein ensemble generation.