Analysis of HIV protease binding pockets based on 3D shape and electrostatic potential descriptors

Abstract

Using a shape analysis technique, a Dataset of 79 (19 wild-type and 60 mutated) HIV-1 protease crystal structures was analyzed for the shape changes in their binding pockets. The structures are reported with different bound inhibitors and consist of a variety of mutations. Several 3D-shape descriptors based on the volumetric shape function are calculated, in addition to the electrostatic potential (EP)-based descriptors. A cluster analysis is performed using the calculated descriptors of the binding pocket of the proteins to investigate the effect of mutations or bound inhibitors on the shape of the binding pocket of proteins. We observed that mutations and/or bound ligands influence the 3D shape of the binding pocket of these HIV proteases. This study shows that the shape changes in the binding pocket caused by mutations or bound ligands can be quantitatively captured using 3D shape- and EP-based descriptors.