Consensus Computational Ligand-Based Design for the Identification of Novel Modulators of Human Estrogen Receptor Alpha

Abstract: We describe the first targeted validation offFLASH, a molecular similarity program from IBM that hasbeen previously proposed as suitable for the virtual screen-ing (VS) of compound libraries based on explicit 3D flexiblesuperimpositions, as part of its deployment within a novelconsensus ligand-based virtual screening cascade. A virtualscreening protocol using fFLASH for the human estrogenreceptor alpha (ERa) was advanced and benchmarkedagainst screens completed using established commercialscreening softwares – Catalyst and ROCS. The optimisedprotocol was applied to a ~6000 member physical screen-ing collection and virtual ‘hits’ sourced and biologically as-sayed. The approach identified a novel, potent and highlyselective partial antagonist of the ERa. This study firstly vali-dates the clique detection algorithm utilised by fFLASH andsecondly, emphasises the benefits of the consensus ap-proach of employing more than one program in a VS pro-tocol. Keywords: Consensus screening · Estrogen receptor · Ligand-based drug design · Pharmacophore matching · Virtual screening