The authors use machine learning of compound-protein interactions to explore drug polypharmacology and to efficiently identify bioactive ligands, including novel scaffold-hopping compounds for two pharmaceutically important protein families: G-protein coupled receptors and protein kinases.

Structure-activity relationship studies on CXCR4 antagonists, which were previously found by using cyclic pentapeptide libraries, were performed to optimize side-chain functional groups, involving conformationally constrained analogues. In addition, a new lead of cyclic pentapeptides with the introduction of a novel pharmacophore was developed.

Several low molecular weight nonpeptide compounds having the dipicolylamine-zinc(II) complex structure were identified as potent and selective antagonists of the chemokine receptor CXCR4. These compounds showed strong inhibitory activity against CXCL12 binding to CXCR4, and the top compound exhibited significant anti-HIV activity. Zinc(II)-dipicolylamine unit-containing compounds proved to be useful and attractive lead compounds for chemotherapy of these diseases as nonpeptide CXCR4 antagonists possessing the novel scaffold structure.