The large number of available protein-ligand X-ray structures represents a wealth of information on non-bonded interactions. The challenge is to convert this huge resource of structural data into a form that is useful for structure-based design purposes. The most obvious application is to derive a fully knowledge-based scoring function or pseudo force field from a statistical analysis of protein-ligand interactions. In theory, such force fields would have some significant advantages, but in practice it is far from trivial to construct them. Some of the key issues that need to be overcome are the choice of the reference state, biases caused by secondary contacts, and accurate atom typing.<br/>We have addressed each of these issues and developed a Protein-Ligand Informatics force field (PLIff). It is a surface-area based function, where only primary contacts - derived by a Voronoi tessellation algorithm – are considered. PLIff describes non-bonded interactions in a fully geometry-dependent manner, and the potentials are smooth, interpretable and make intuitive sense. As such, PLIff has a wide range of applications, from protein-ligand docking and virtual screening to druggability and solvent mapping.<br/>Here we will describe the PLIff methodology and present its performance in docking and virtual screening validation experiments and we will describe its use for solvent mapping. Finally, we will illustrate how we have incorporated PLIff as part of our integrated fragment-based drug discovery platform.

Toward a complete, fully knowledge-driven pseudo force field for protein-ligand interactions

249th National Meeting (2015)

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