OEDocking
OEDocking is a suite of well validated molecular docking tools and their associated workflows. Each tool is specifically designed to address its own unique application to the docking problem.
OEDocking features POSIT for informed pose prediction as well as FRED and HYBRID as complementary tools for virtual screening.
Much of the functionality available in OEDocking is also available in toolkit form via the OEDocking TK.
Pose Prediction
POSIT - Ligand guided pose prediction
POSIT is pose prediction tool that uses a combined shape and MMFF force field to predict poses in the context of a binding site with known bound ligand(s). POSIT fits a new molecule to the known binding mode using the shape and chemical features of the known ligand, while minimizing the strain of the predicted pose. Each pose generated by POSIT is provided with a prospective estimation of the probability that this predicted pose will be within 2.0 Angstroms RMSD of the actual experimental structure.
The probability is computed using both the 2D and the 3D similarity of the ligand being fit to the known bound ligand. In addition, if provided with a selection of receptors from a crystallographic series, POSIT will automatically determine which receptor is best suited for pose prediction.
Virtual Screening
FRED - Fast exhaustive docking
FRED is among the best, fastest, and shows the lowest variability in its virtual screening results among a large number of docking programs examined in two different exhaustive studies [1,2].
FRED performs a systematic, exhaustive, nonstochastic examination of all possible poses within the protein active site, filters for shape complementarity [3] and pharmacophoric features before selecting and optimizing a single pose based upon a consensus of structure-based scoring functions including ChemGuass, ChemScore, PLP, and ShapeGauss.
HYBRID - Ligand guided docking
HYBRID is the docking tool of choice for virtual screening when there is a known reference ligand for the system of choice.
Like FRED, HYBRID exhaustively examines all possible poses within the protein active site, but instead of initially filtering the poses based on their shape complementarity to the active site, the poses are filtered based on their shape and chemical complementarity to a known bound ligand. The poses which pass the ligand-based filtering are then scored and optimized using a consensus of structure-based scoring functions.
This ligand-guided docking process provides statistically significantly better enrichment over similar docking tools [2].
HYBRID is currently available as a run-time option within FRED; however, it will be provided as a separate but complementary application to FRED in the next release.
References
-
G.B. McGaughey, R.P. Sheridan, C.I. Bayly, J.C. Culberson, C. Kreatsolas, S. Lindsley, V. Maiorov, J.-F. Truchon and W.D. Cornell, "Comparison of Topological, Shape, and Docking Methods in Virtual Screening", J. Chem. Inf. Model., 2007, 47 (4), pp 1504-1519
- M. McGann, "FRED Pose Prediction and Virtual Screening Accuracy", J. Chem. Inf. Model., 2011, 51 (3), pp 578-596
- M.R. McGann, H.R. Almond, A. Nicholls, J.A. Grant and F.K. Brown, "Gaussian docking functions", Biopolymers, 2003, 68 (1), pp 76-90
