Strengths and Weaknesses of Docking Simulations in the SARS-CoV-2 Era: the Main Protease (Mpro) Case Study

Abstract

The scientific community is working against the clock to arrive at therapeutic interventions to treat patients with COVID-19. Among the strategies for drug discovery, virtual screening approaches have the capacity to search potential hits within millions of chemical structures in days, with the appropriate computing infrastructure. In this article, we first analyzed the published research targeting the inhibition of the main protease (Mpro), one of the most studied targets of SARS-CoV-2, by docking-based methods. An alarming finding was the lack of an adequate validation of the docking protocols (i.e., pose prediction and virtual screening accuracy) before applying them in virtual screening campaigns. The performance of the docking protocols was tested at some level in 57.7% of the 168 investigations analyzed. However, we found only three examples of a complete retrospective analysis of the scoring functions to quantify the virtual screening accuracy of the methods. Moreover, only two publications reported some experimental evaluation of the proposed hits until preparing this manuscript. All of these findings led us to carry out a retrospective performance validation of three different docking protocols, through the analysis of their pose prediction and screening accuracy. Surprisingly, we found that even though all tested docking protocols have a good pose prediction, their screening accuracy is quite limited as they fail to correctly rank a test set of compounds. These results highlight the importance of conducting an adequate validation of the docking protocols before carrying out virtual screening campaigns, and to experimentally confirm the predictions made by the models before drawing bold conclusions. Finally, successful structure-based drug discovery investigations published during the redaction of this manuscript allow us to propose the inclusion of target flexibility and consensus scoring as alternatives to improve the accuracy of the methods.

A Review of the Current Landscape of SARS-CoV-2 Main Protease Inhibitors: Have We Hit the Bullseye Yet?

“…Using two different SARS-CoV-2 M-pro structures and five protein-ligand docking methods, we have recently shown that docking scores or the Gibbs free energy (∆G) calculated with an MM-GBSA method [ 18 ] do not correlate with bioactivity [ 19 ], probably because of the inability of common docking programs to correctly reproduce the binding modes of SARS-CoV-2 M-pro inhibitors [ 20 ]. This reinforces the idea that it is essential to validate the results obtained by protein-ligand docking or any other computational tool, especially when analyzing SARS-CoV-2 M-pro inhibitors [ 19 , 21 , 22 , 23 ]. The results of protein-ligand docking can be computationally validated by re-docking, cross-docking and applying the same protocol to a set of known active compounds and a set of decoy or inactive compounds [ 19 ].…”

Epicatechin is a promising novel inhibitor of SARS-CoV-2 entry by disrupting interactions between angiotensin-converting enzyme type 2 and the viral receptor binding domain: A computational/simulation study

“…In contrast to these encouraging data, it has recently been proven that the docking score is not good enough to fully support as a cutoff value for selecting possible inhibitors for SARS-CoV-2 infection because no good correlations have been found between docking scores and pIC50 values for these inhibitors [ 70 ]. Furthermore, it has been found that even though all tested docking protocols have a good pose prediction, their screening accuracy is quite limited as they fail to correctly rank a test set of compounds [ 71 ]. Though increasing docking search algorithms have recently been improved, it is still questioned if there are any reliable approach in which molecular docking have aided to bring a drug to the market.…”

Machine learning prediction of 3CL SARS-CoV-2 docking scores

“…Therefore, the tools of trade still need to be developed and established [32] to provide a much faster response to treat new potential pandemic risks [33] , [34] , [35] , [36] . The popularity of molecular docking and connected software were slowly diminishing in the 2010s due to the simplicity of its approach, while methods based on a more robust interpretation of drug-target interactions gained more popularity.…”