CSAR Benchmark of Flexible MedusaDock in Affinity Prediction and Nativelike Binding Pose Selection

While molecular docking with both ligand and receptor flexibilities can help capture conformational changes upon binding, correct ranking of nativelike binding poses and accurate estimation of binding affinities remains a major challenge. In addition to the commonly used scoring approach with interm...

Full description

Saved in:
Bibliographic Details
Published in:Journal of chemical information and modeling 2016-06, Vol.56 (6), p.1042-1052
Main Authors: Nedumpully-Govindan, Praveen, Jemec, Domen B, Ding, Feng
Format: Article
Language:English
Subjects:
Benchmarking
Benchmarks
Binding sites
Biochemistry
Comparative analysis
Ligands
Molecular chemistry
Molecular Docking Simulation
Protein Binding
Protein Conformation
Proteins - chemistry
Proteins - metabolism
Ratings & rankings
Structure-Activity Relationship
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:While molecular docking with both ligand and receptor flexibilities can help capture conformational changes upon binding, correct ranking of nativelike binding poses and accurate estimation of binding affinities remains a major challenge. In addition to the commonly used scoring approach with intermolecular interaction energies, we included the contribution of intramolecular energies changes upon binding in our flexible docking method, MedusaDock. In CSAR 2013–2014 binding prediction benchmark exercises, the new scoring function MS complex was found to better recapitulate experimental binding affinities and correctly identify ligand-binding sequences from decoy receptors. Our further analysis with the DUD data sets indicates significant improvement of virtual screening enrichment using the new scoring function when compared to the previous intermolecular energy based scoring method. Our postanalysis also suggests a new approach to select nativelike poses in the clustering-based pose ranking approach by MedusaDock. Since the calculation of intramolecular energy changes and clustering-based pose ranking and selection are not MedusaDock specific, we expect a broad application in force-field based estimation of binding affinities and pose ranking using flexible ligand–receptor docking.
ISSN:1549-9596
1549-960X
DOI:10.1021/acs.jcim.5b00303