Reliable predictions of binding poses and binding affinities of small/drug-like molecules to target proteins based on their crystal structures remains a challenge in computational chemistry despite the increase in available computational power and vast algorithmic improvements. The ability to accurately predict binding is of particular interest in the area of computer aided drug-design and can reduce production costs of new drug molecules significantly.<br/><br/>Here, we will present a high-throughput pipeline for reliably setting-up and carrying out alchemical binding free energy calculations. FESetup serves as an automated ligand and protein setup tool in conjunction with the molecular library Sire/OpenMM for molecular modelling of protein-ligands binding using alchemical free energy calculations. To illustrate this pipeline we will discuss results obtained by application of this pipeline to selected protein-ligand systems in retrospective predictions. We will also discuss coupling of the software with docking protocols for blind predictions of binding affinities for a large dataset of ligands binding to the chaperone protein HSP90 (heat shock protein 90) in the context of the D3R grand challenge.

Software pipelines for high throughput alchemical binding free energy calculations

251st National Meeting (2016)

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