Machine learning-guided in Silico identification of Na⁺-NQR inhibitors from Berberis vulgaris and Hydrastis Canadensis phytochemicals against Vibrio cholerae

Abstract

Abstract Cholera, a life-threatening diarrheal disease caused by Vibrio cholerae, remains a global health concern. Traditional medicinal plants such as Berberis vulgaris (barberry) and Hydrastis canadensis (goldenseal) have long been used for their antimicrobial properties. This study employed integrated computational approaches to identify potential anti-cholera compounds from these plants by targeting sodium-pumping NADH-ubiquinone oxidoreductase (Na⁺-NQR). Virtual screening identified compounds 122623 , 197835 , and 638024 , along with the control Korormicin, exhibiting favorable binding affinities and hydrogen bonding interactions. Machine learning-based prediction models were further applied to assess functional activity and binding affinity. The study incorporated ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) profiling and density functional theory (DFT) calculations to evaluate physicochemical properties and electronic characteristics of the compounds. Molecular dynamics (MD) simulations demonstrated stable RMSD values for the ligands, with 122623 exhibiting the lowest RMSD (0.3–0.5 nm), closely resembling the control (0.3–0.4 nm), suggesting stable binding. Principal component analysis (PCA) showed tight conformational clusters for complexes with 122623 and 197835 , while free energy landscape (FEL) analysis revealed deep energy minima, supporting complex stability. MM/GBSA calculations showed that 122623 had the lowest binding free energy (-38.71 kcal/mol), followed by 638024 (-35.14 kcal/mol) and 197835 (-30.68 kcal/mol). Per-residue energy decomposition identified key residues (Phe137, Val155, Phe159, Phe160) involved in ligand binding. Network pharmacology analysis predicted additional gene targets for the selected compounds, providing insights into their broader therapeutic relevance. Collectively, compounds 122623 , 197835 , and 638024 derived from Hydrastis canadensis and Berberis vulgaris demonstrated promising inhibitory interactions with the Na⁺-NQR enzyme, suggesting their potential as anti-cholera agents.