ShapeFit is a ligand-centric pose prediction methodology that performs both molecular alignment and scoring, during pose prediction, against the bound-ligand of the input receptor. ShapeFit aligns a molecule to the bound-ligand using a flexible fitting algorithm that combines rigid overlay as performed in ROCS, and intra-molecular interactions of docked molecules as defined by a force field. ShapeFit is part of OpenEye’s pose prediction tool POSIT, and is preferred for molecules with a ROCS similarity of 1.1 or more with the bound ligand.  In this study, the latest OpenFF force field, Sage, is introduced to ShapeFit. The adiabatic optimization process of slowly mixing Shape with a force field is also replaced by a single optimization of both ROCS overlay and intra-molecular energies.<br/> <br/>Pose prediction accuracy of the new algorithm, and the effect of Sage force field on it is studied by cross-docking experiments on a large database of kinase structures. When the fraction of accurate predictions is compared, results from single optimization against adiabatic optimization, as well as between the use of MMFF vs Sage force field are comparable. However, both the new algorithm and the Sage force field generate poses that show lower local strain compared to predictions using their counterparts.<br/> <br/>To further enhance the pose predictive capability with ShapeFit, current efforts are being directed towards subsequent optimization of the generated pose in presence of the protein.

ShapeFit: Pose prediction using flexible fitting with shape, color, and force field

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.