Expanding FTMap for the fragment-based identification of pharmacophore regions in ligand binding sites

Abstract:
In fragment-based drug discovery, the binding mode of a fragment bound to a hot spot is expected to be conserved as it is optimized into a larger ligand.1,2 Therefore, predicting the locations of intermolecular interactions that are conserved in fragment-lead pairs is of great importance in the context of pharmacophore generation. To aid in the identification of pharmacophore regions in ligand binding sites we have developed E-FTMap, a computational solvent mapping algorithm which exhaustively maps binding sites with dozens of small organic probes, and identifies important interaction sites as atomic consensus sites (ACSs) where similar chemical groups bind. We validate E-FTMap against a set of 109 experimentally derived structures of fragment-lead pairs, finding highly ranked pharmacophore features that overlap with corresponding atoms in both fragment and lead compounds. Additionally, we compare mapping results to pharmacophores derived from ensembles of bound ligands, revealing that E-FTMap results tend to sample highly conserved protein-ligand interactions. Furthermore, we explore an application of E-FTMap in the context of virtual ligand screening by using mapping results to score and rank ligands on the basis of their binding affinity.