Multitarget Virtual Screening for Drug Repurposing in COVID19

Abstract

Therapeutic or preventive research for coronavirus SARS-CoV2 is an extremely active topic of research since its outbreak in January 2020. In this paper we report the results from a virtual drug screening analysis that, to the best of our knowledge, is the widest work in terms of target proteins and compound library. Our study was focused on the repurposing of currently commercialized drugs, and especially those that can interact with multiple viral proteins and several binding sites within each protein. Additionally, we performed a second virtual screening analysis in which we compared our results to the predicted binding affinities for the drugs currently in clinical trials. We show that the best molecules in our screening compares favorably to those in clinical trials, suggesting their suitability for therapeutic or preventive applications.

In silico analysis of SARS-CoV-2 proteins as targets for clinically available drugs

“…Previous virtual drug repurposing screens against multiple SARS-CoV-2 targets suffer from several limitations, including the use of relatively few sites and/or too many sites obtained from automated methods, reliance on docking scores alone to prioritize drugs that favorably bias large molecular weight ligands, biased comparison of drugs between binding sites, the absence of molecular dynamics simulations to further prioritize lead docked structures 37 , 38 , and a lack of use of known biological data on anti-SARS-CoV-2 29 , 30 activity. Since many of the putative sites and drug-target interactions we have identified are novel we were unable to retrospectively benchmark our methodology.…”

“…This provides a resource for further investigating potential SARS-CoV-2 drugs identified in phenotypic screens without an obvious target, where target deconvolution remains a major barrier 50 . Moreover, several of the identified drugs are predicted to act at multiple sites across the same or different proteins, suggesting drug combinations may produce additive or even synergistic antiviral activity 29 , 30 .…”

“…Similarly, attractive targets include less studied domains/proteins, such as the Mac1 domain of Nsp3 26 , 27 and the endonuclease Nsp15 28 , both likely involved in blocking the host immune response. Docking-based virtual screens enable the quick identification of leads by screening many drugs across single targets 26 , 27 , or single drugs across many targets 29 , or a combination of both 30 . For example, virtual screens have been used to identify plant products that may inhibit important viral proteins such as the main protease (Nsp 5 or Mpro) 31 , Nsp1 32 , RNA-dependent polymerase 33 , the spike protein binding domain 34 , Nsp15 33 , 35 and Nsp16 36 .…”