Algorithmic improvements in DOCK6 for enhanced conformational sampling

COMP 204

Sudipto Mukherjee, sudipto.mukherjee@gmail.com, Trent E. Balius, tbalius@ams.sunysb.edu, and Robert C. Rizzo, rizzorc@gmail.com. Department of Applied Mathematics and Statistics, Stony Brook University, Stony Brook, NY 11794-3600
In an effort to optimize docking and virtual screening protocols for the DOCK6 program we have performed a critical analysis of three core docking experiments. Results will be presented from experiments which aim to recover the native bound structure of ligands using rigid (RGD), fixed anchor (FAD), and flexible (FLX) docking. Success rates for individual protein families and specific considerations such as average ligand flexibility, presence of ions etc. will be examined. Growth trees were implemented to visualize partially grown ligand conformations and trace the behavior of the core DOCK anchor-and-grow algorithm, which in turn facilitated debugging and optimization of input parameters. Aggressive pruning of unfavorable ligand geometries, using the repulsive term from the ligand intramolecular van der Waals energy, yielded faster run times and enhanced sampling over baseline DOCK protocols.
 

Poster Session
6:00 PM-8:00 PM, Tuesday, August 18, 2009 Walter E. Washington Convention Center -- Ballroom A, Poster

Division of Computers in Chemistry

The 238th ACS National Meeting, Washington, DC, August 16-20, 2009