A common first step in drug repurposing is the large-scale screening of compound libraries to identify therapeutic leads. Computational approaches to quickly screen a library of putative drugs for their potential to bind protein targets are becoming increasingly popular. A pharmacophore is a set of chemical features necessary for binding of a ligand to a protein. Ligand-based pharmacophore mapping is a computational method that identifies prominent chemical features of ligands and generates models based on the common chemical features responsible for activity. When the receptor protein's structure is unknown, known ligands of a receptor protein provide a common pharmacophore hypothesis to model and search compound libraries for putative drugs.
Human CC-Chemokine Receptor 4(CCR4) antagonists represent a novel therapeutic intervention in diseases where CCR4 has a central role in pathogenesis, such as asthma and various other allergic diseases, the mosquito-borne tropical diseases, cancer and HIV/AIDS. One way to antagonize CCR4 activity is via allostery. In order to perform virtual screening of large compound databases for new leads of allosteric CCR4 antagonists, we built pharmacophore models using two sets of CCR4 allosteric antagonists found in the literature: indazole arylsulfonamides and heteroarylpyrazole arylsulfonamides. Both sets of compounds bind to the same allosteric site. We built models for each set and validated their accuracy computationally. Pharmacophore modeling validation was done in two opposite directions. Each of the arylsulfonamide groups was used in modeling while the other provided a base to build a validation set. In my talk, I will discuss about the models and the validation results.