Cheminformatics and machine learning approaches for repurposing anti-viral compounds against monkeypox virus thymidylate kinase.

One of the emerging epidemic concerns is Monkeypox disease which is spreading globally. This disease is caused by the monkeypox virus (MPXV), with an increasing global incidence with an outbreak in 2022. One of the novel targets for monkeypox disease is thymidylate kinase, which is involved in pyrimidine metabolism. In this study, docking-based virtual screening and molecular dynamics techniques were employed in addition to the machine learning (ML) model to investigate the potential anti-viral natural small compounds to inhibit thymidylate kinase of MPXV. Several potential hits were identified through high-throughput virtual screening, and further top three candidates were selected, which ranked using the ML model. These three compounds were then examined under molecular dynamics simulation and MM/GBSA-binding free energy analysis. Among these, Chlorhexidine HCl showed high potential for binding to the thymidylate kinase with stable and consistent conformation with RMSD < 0.3 nm. The MM/GBSA analysis also showed the minimum binding free energy (ΔG) of -62.41 kcal/mol for this compound. Overall, this study used structure-based drug design complemented by machine learning-guided ligand-based drug design to screen potential hit compounds from the anti-viral natural compound database.