Molecular docking of medicinal plants: Towards metabolic disease management

The rising global burden of metabolic disease, together with the high cost and limited accessibility of conventional antidiabetic therapeutics in sub-Saharan Africa, has renewed interest in agriculturally cultivated medicinal plants as sustainable sources of bioactive phytochemicals with well-documented ethnopharmacological relevance. This study aimed to identify and prioritize phytochemicals with therapeutic potential against validated metabolic disease targets through an integrated computational drug discovery workflow. 14 bioactive phytochemicals from six agricultural medicinal plants- Moringa oleifera, Vernonia amygdalina, Azadirachta indica, Curcuma longa, Hibiscus sabdariffa, and Zingiber officinale were retrieved from PubChem and docked against five protein targets implicated in type 2 diabetes mellitus (T2DM) and metabolic syndrome: α-glucosidase (PDB: 3W37), dipeptidyl peptidase-4 (DPP-4; PDB: 2I78), peroxisome proliferator-activated receptor gamma (PPARγ; PDB: 2PRG), α-amylase (PDB: 1B2Y), and glucose transporter type 2 (GLUT2; PDB: 5EQG). Docking was performed in AutoDock Vina via PyRx 0.8, ADMET properties were predicted using SwissADME and ADMETlab 3.0, and a random forest classifier trained on ChEMBL 33 bioactivity data was used for artificial intelligence-assisted lead prioritization. Nimbolide from A. indica showed the strongest binding affinity, docking against PPARγ at - 8.4 kcal/mol, while hibiscetin from H. sabdariffa bound GLUT2 at - 7.6 kcal/mol. Quercetin and 6-gingerol inhibited α-glucosidase more strongly than the reference drug acarbose (- 6.5 kcal/mol). All priority leads satisfied Lipinski’s Rule of Five and displayed favorable, low-risk ADMET profiles. The integration of ethnopharmacological plant selection, molecular docking, ADMET prediction, and AI-assisted prioritization identified nimbolide, hibiscetin, quercetin, and curcumin as promising, sustainable phytochemical leads for metabolic disease management in resource-limited settings.