| Abstract
| - Three-dimensional pharmacophore models for peptidic and small organic nonpeptidic inhibitors of the humanrhinovirus 3C protease were generated in a structure-based as well as in a ligand-based approach, using thesoftware package Catalyst. The inhibitors possess an electrophilic moiety, often a Michael acceptor function,which covalently binds to a cysteine in the active site of the enzyme. Since this process presents the keystep for virus inactivation, the creation of a new function in Catalyst was required in order to include thisdecisive functionality into the pharmacophore models. In the present study we focus on this feature definitionprocess because it presents an innovative strategy to expand the pharmacophore description ability of theCatalyst software to also include covalent bonds between ligand and binding site. The resulting hypotheseswere then used for virtual screening of 3D databases in order to verify their quality and to search forstructurally diverse, possible new lead substances.
|
