Computational Drug Repurposing Studies on SARS-CoV-2 Protein Targets

To contribute to the combat of COVID-2019, we applied structure-based computational docking screens using flexible docking protocol of Rosetta GALigandDock against multiple potential SARS-CoV-2 protein targets, including the Nsp5 3-chymotrypsin-like protease (3CLpro), the Nsp3 ADP ribose phosphatase, the Nsp15 Endoribonuclease, the RNA binding domain of nucleocapsid phosphoprotein, the Nsp16 2'-O-MTase, Nsp14, and Nsp12 RNA-dependent RNA polymerase. Screening against a re-purposing library of 8,395 FDA approved drugs at various stages of drug development and various natural products from DrugBank, we found a total of 124 putative inhibitors with predicted binding ∆G less than -8.9 kcal/mol, including HIV-AIDS drugs Nelfinavir and Tipranavir, targeting 3Clpro with ∆G=-18.8 kcal/mol and ∆G=-16.6 kcal/mol respectively. These primarily involve binders to the Nsp5 3CLpro (37 hits) and the Nsp3 ADP ribose phosphatase (36 hits), with smaller numbers of hits to other targets. These small molecule putative inhibitors suggest a possible avenue for drug repurposing, and the identified compounds should serve as a high-priority list for experimental validation via co-crystallization, enzymatic and cell based assays.