Exploring Druggable\nBinding Sites on the Class A GPCRs\nUsing the Residue Interaction Network and Site Identification by Ligand\nCompetitive Saturation

G protein-coupled receptors (GPCRs) play a central role\nin cellular\nsignaling and are linked to many diseases. Accordingly, computational\nmethods to explore potential allosteric sites for this class of proteins\nto facilitate the identification of potential modulators are needed.\nImportantly, the availability of rich structural data providing the\nlocations of the orthosteric ligands and allosteric modulators targeting\ndifferent GPCRs allows for the validation of approaches to identify\nnew allosteric binding sites. Here, we validate the combination of\ntwo computational techniques, the residue interaction network (RIN)\nmodel and the site identification by ligand competitive saturation\n(SILCS) method, to predict putative allosteric binding sites of class\nA GPCRs. RIN analysis identifies hub residues that mediate allosteric\nsignaling within a receptor and have a high capacity to alter receptor\ndynamics upon ligand binding. The known orthosteric (and allosteric)\nbinding sites of 18 distinct class A GPCRs were successfully predicted\nby RIN through a dataset of 105 crystal structures (91 ligand-bound,\n14 unbound) with up to 77.8% (76.9%) sensitivity, 92.5% (95.3%) specificity,\n51.9% (50%) precision, and 86.2% (92.4%) accuracy based on the experimental\nand theoretical binding site data. Moreover, graph spectral analysis\nof the residue networks revealed that the proposed sites were located\nat the interfaces of highly interconnected residue clusters with a\nhigh ability to coordinate the functional dynamics. Then, we employed\nthe SILCS-Hotspots method to assess the druggability of the novel\nsites predicted for 7 distinct class A GPCRs that are critical for\na variety of diseases. While the known orthosteric and allosteric\nbinding sites are successfully explored by our approach, numerous\nputative allosteric sites with the potential to bind drug-like molecules\nare proposed. The computational approach presented here promises to\nbe a highly effective tool to predict putative allosteric sites of\nGPCRs to facilitate the design of effective modulators.