Artificial Neural Network Based Study Predicting GS-441524 As Potential Inhibitor of SARS COV-2 Activator Protein Furin: A Polypharmacology Approach

Abstract

Furin, a pro-protein convertase, plays a significant role of biological scissor in bacterial, viral, and even mammalian substrates which in turn decides the fate of many viral and bacterial infections along with the numerous ailments caused by cancer, diabetes, inflammations, and neurological disorders. In the wake of the current pandemic caused by the virus SARS COV-2, furin has become the center of attraction for researchers. In the present work, we have searched for novel inhibitors against this interesting human target from FDA-approved antivirals. To enhance the selection of new inhibitors we employed Kohonen’s-artificial neural network-based self-organizing maps for ligand based virtual screening. Promising results were obtained which can help in drug repurposing and network pharmacology studies addressing the errors due to promiscuity/polypharmacology. We found 15 existing FDA antivirals having the potential to inhibit furin. Among these, six compounds have targets on other important human proteins (LDLR, FCGR1A, PCK1, TLR7, DNA and PNP) also. These 15 drugs inhibiting furin could be studied in patients having many viral infections including SARS COV-2, which is known to have many interacting motifs like NSPs, ORFs, and spike protein. We also propose two promising candidate FDA drugs GS-441524 and Grazoprevir (MK-5172) to repurpose as inhibitors of furin. The best results were observed with GS-441524.