Atomistic De-novo Inhibitor Generation-Guided Drug Repurposing for SARS-CoV-2 Spike Protein with Free-Energy Validation by Well-Tempered Metadynamics

Computational drug design is increasingly becoming important with new and unforeseen diseases like COVID-19. In this study, we present a new computational de novo drug design and repurposing method and applied it to find plausible drug candidates for the receptor binding domain (RBD) of SARS-CoV-2 (COVID-19). Our study comprises three steps: atom-by-atom generation of new molecules around a receptor, structural similarity mapping to existing approved and investigational drugs, and validation of their binding strengths to the viral spike proteins based on rigorous all-atom, explicit-water well-tempered metadynamics free energy calculations. By choosing the receptor binding domain of the viral spike protein, we showed that some of our new molecules and some of the repurposable drugs have stronger binding to RBD than hACE2. To validate our approach, we also calculated the free energy of hACE2 and RBD, and found it to be in an excellent agreement with experiments. These pool of drugs will allow strategic repurposing against COVID-19 for a particular prevailing conditions.

Automated summary

Computational drug design is crucial for tackling emerging diseases like COVID-19. This study introduces a new method for de novo drug design and repurposing, successfully identifying potential drug candidates for the receptor binding domain of SARS-CoV-2. By generating new molecules, mapping their structural similarity to existing drugs, and validating their binding strengths, the study demonstrates the effectiveness of the approach in strategic drug repurposing against COVID-19.