Identification of Potential Inhibitors of SARS-CoV-2 Using Machine Learning, Molecular Docking and MD Simulation

Abstract

The advent of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of the coronavirus disease 2019 (COVID-19) pandemic, has promoted physical and mental health worldwide. Due to the unavailability of effective antiviral drugs, there is an unmet demand for a robust therapeutic approach for the development of anti-COVID-19 drugs. Multiple investigations have established that ACE2 is the primary receptor for the causal virus of COVID-19, and this amalgamation of ACE2 with the spike protein of the coronavirus is essential for viral entry into host cells and inducing infection. As a result, limiting or restricting the accessibility of the virus to ACE2 offers a different tactical approach to averting this illness. Therefore, this study aimed to identify the most effective inhibitors with an augmented affinity for the ACE2 protein and evaluate their pharmacological efficacy. The pre-established repurposed viral compounds, Birinapant, Remdesivir, etc., and the ACE2-inhibiting compounds, Quninaprill, Moexipril, etc., were taken as test datasets, and machine learning algorithms were employed to govern the novel compounds. Furthermore, comparative analyses were also performed for both the new machine-learning compounds and pre-established compounds via the MD simulation approach to confirm the structural stability. The study concluded that the machine learning compound (CID: 23658468) could be a potential drug for the treatment of COVID-19.