In an attempt to identify new ligands for the 5-HT(7) receptor (5-HT(7)R), we developed and tested a hierarchical multi-step strategy of virtual screening (VS) based on two-dimensional (2D) pharmacophore similarity, physicochemical scalar descriptors, an ADME/Tox filter, three-dimensional (3D) pharmacophore searches and a docking protocol. Six chemical classes of 5-HT(7)R antagonists were used as query structures in a double-path virtual screening scheme. The Enamine screening database, consisting of approximately 730,000 commercially available drug-like compounds, was adopted and used as a source of structures. A biological evaluation of 26 finally selected virtual hits resulted in finding two benzodioxane derivatives with significant affinity (K(i)=197 and 265 nM). The approach described in this case study can be easily used as a general rational drug design tool for other biological targets.
Copyright 2010 Elsevier Ltd. All rights reserved.