Virtual Screening as a Computational Tool to Rapidly Profile Target Drug-Candidates
Laboratory of Molecular Modeling and Design (M/C-781)
College of Pharmacy University of Illinois at Chicago Chicago, IL 60612-7231
The most fundamental reason for doing computational chemistry is to build models that provide insight and understanding of chemical systems. Perhaps the best computational models are those that permit quantitative predictions of the properties of chemical systems. The biopharmaceutical sciences continue to focus on exploiting the use of computational models to streamline the drug discovery process. A computational biopharmaceutical model usually takes the form of a quantitative structure-activity relationship, QSAR, and when the QSAR is used in a predictive mode, the process is referred to as virtual screening. This talk will describe how virtual screening can be used to profile key potency, transport and toxicity properties of target drug-candidates before they are made and tested. Two QSAR paradigms developed in our laboratory will be highlighted. The 4D-QSAR formalism can be applied to construct a predictive computational model in terms of a 3D-pharmacophore defining how a drug-candidate binds to its target receptor macromolecule, normally a protein. The MI-QSAR paradigm permits construction of models for estimation of absorption, distribution and toxicity of drug-candidate compounds in order to evaluate their overall “drug-likeness”. A full-scale applications study will be presented for the in silico design of glucose analog inhibitors of glycogen phosphorylase-b for treatment of metabolic and secretory diseases including type-1 diabetes.