Development of a novel virtual screening strategy for the discovery of inhibitors

Potential inhibitors for specific proteins are identified via an integrated approach comprising pharmacophore modeling, docking and substructure search. The pharmacophore models, which are generated either based on the binding site structure or based on a set of previously reported inhibitors, are used to filter the ZINC database based on the 3D similarity to the selected pharmacophore hypothesis. Molecular docking is applied to predict the binding modes of the compounds in the filtered database. Common scaffolds in the top scoring hits are identified and substructure search with the selected scaffold is carried out to filter the database and obtain other compounds. The surviving hits are further analyzed via induced fit docking and molecular dynamics simulations are proposed as potential inhibitors.