Isabela de Souza Gomes, Charles Abreu Santana, Leandro Soriano Marcolino, Leonardo Henrique França de Lima, Raquel Cardoso de Melo-Minardi, Roberto Sousa Dias, Sérgio Oliveira de Paula, Sabrina de AzevedoSilveira
Abstract
The development of new drugs is a very complex and time-consuming process, and for this reason, researchers have been resorting heavily to drug repurposing techniques as an alternative for the treatment of various diseases. This approach is especially interesting when it comes to emerging diseases with high rates of infection, because the lack of a quickly cure brings many human losses until the mitigation of the epidemic, as is the case of COVID-19. In this work, we combine an in-house developed machine learning strategy with docking, MM-PBSA calculations, and metadynamics to detect potential inhibitors for SARS-COV-2 main protease among FDA approved compounds. To assess the ability of our machine learning strategy to retrieve potential compounds we calculated the Enrichment Factor of compound datasets for three well known protein targets: HIV-1 reverse transcriptase (PDB 4B3P), 5-HT2A serotonin receptor (PDB 6A94), and H1 histamine receptor (PDB 3RZE). The Enrichment Factor for each target was, respectively, 102.5, 12.4, 10.6, which are considered significant values. Regarding the identification of molecules that can potentially inhibit the main protease of SARS-COV-2, compounds output by the machine learning step went through a docking experiment against SARS-COV-2 Mpro. The best scored poses were the input for MM-PBSA calculations and metadynamics using CHARMM and AMBER force fields to predict the binding energy for each complex. Our work points out six molecules, highlighting the strong interaction obtained for Mpro-mirabegron complex. Among these six, to the best of our knowledge, ambenonium has not yet been described in the literature as a candidate inhibitor for the SARS-COV-2 main protease in its active pocket.
Introduction
During an outbreak, it is necessary to quickly respond to the unknown pathogen to avoid the uncontrolled spread of the disease. Just like what happened with the novel COVID-19 disease, caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV-2), that appeared for the first time in Wuhan, China at the end of 2019, spreading quickly all over the globe. In approximately one year, the infection reached more than 170 million cumulative confirmed cases and caused over than 3.5 million deaths (https://covid19.who.int/; as of Jun 2021). Considering that the development of new drugs is expensive and time-consuming, in this scenario computational strategies can potentially speed up the process of drug discovery. The repurposing of existing drugs to treat new diseases can accelerate the approval process, giving a quick response against unknown pathogens [1]. The use of the antiviral Remdesivir is an example, as this drug was initially indicated for the treatment of Ebola, and in October 2020, FDA approved it for use against COVID-19 [2]. However, the mortality rate for patients treated with Remdesivir is still quite high and does not differ significantly from placebo treatment in clinical trials [3, 4]. This shows that treatment with this antiviral alone is still not enough and further research to identify other promising drugs should continue.
Materials and methods
This section details our computational strategy to predict candidate inhibitors for SARS-COV-2 Mpro. (Fig 1) presents the workflow that outlines the process. We explain data collection and preprocessing, prediction of molecules through the in-house developed supervised learning strategy, and refinement of these molecules through docking and molecular dynamics.
Results and discussion
This section discusses detailed results for each step of our method. First, we show that our supervised learning strategy is able to enrich compound datasets for three well known targets. Then our strategy is applied to point out potential molecules to inhibit SARS-COV-2 Mpro that will be refined through docking and molecular dynamics. Next, molecules that went through docking simulation and their respective scores are presented. Finally, the binding energies calculated through MM-PBSA and metadynamics simulations for two force fiels are shown.
Conclusion
This work brings together an in-house developed machine learning strategy, docking, MM-PBSA calculations and metadynamics to identify FDA approved molecules that can potentially inhibit the main protease of SARS-COV-2. First, we devised a machine learning strategy that couples different molecule fingerprints to perform a first step of LBVS. Next, the resulting molecules go through SBVS steps, which consists of docking these molecules against the target protein (SARS-COV-2 Mpro) using AutodockVina and then selecting the poses with lowest vina scores. Finally, we selected the best-scored poses to perform MM-PBSA calculations and metadynamic simulations using CHARMM and AMBER force fields to predict the binding energy for each complex.
Citation: Gomes IdS, Santana CA, Marcolino LS, Lima LHFd, Melo-MinardiRCd, Dias RS, et al. (2022) Computational prediction of potential inhibitors for SARS-COV-2 main protease based on machine learning, docking, MM-PBSA calculations, and metadynamics. PLoS ONE 17(4): e0267471. https://doi.org/10.1371/journal.pone.0267471
Editor: AlessioLodola, University of Parma, ITALY
Received: August 9, 2021; Accepted: April 6, 2022; Published: April 22, 2022
Copyright: © 2022 Gomes et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The source code and datasets are available at: https://github.com/IsabelaGomes/Prediction_SARSCOV2_inhibitors.
Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 23038.004007/2014-82 - Grant 051/2013; ConselhoNacional de DesenvolvimentoCientífico e Tecnológico (CNPq); Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG). The funding agencies had no role in study design, data collection, analysis and interpretation, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0267471#abstract0
