Integrated computational discovery of AMPK-activating flavonoids reveals curcumin, resveratrol, and tangeritin as promising neuroprotective candidates

Flavonoid activators of AMP-activated protein kinase (AMPK) are emerging as promising modulators of neurodegenerative disease–relevant pathways, yet a systematic, multi-level in silico comparison of candidate molecules is lacking. This study presents the first integrated computational pipeline that combines pharmacokinetic screening, multi-site molecular docking, 200-ns molecular dynamics simulations, and MMPBSA binding free energy analysis to prioritize six flavonoids as AMPK-targeting agents (resveratrol, tangeritin, paeonol, eugenol, curcumin, and the reference inhibitor compound C). Bioactive flavonoids retrieved from literature and public chemical databases were filtered using SwissADME for Lipinski compliance, oral bioavailability, lipophilicity, solubility, gastrointestinal absorption, and blood–brain barrier permeability via the BOILED-Egg model to enrich for CNS-relevant neuroprotective candidates. AutoDock Vina–based docking to the AMPK C-chain revealed that resveratrol, tangeritin, and curcumin exhibit the most favourable binding at catalytically important sites, stabilised by extensive hydrogen-bonding and hydrophobic contacts. Subsequent 200 NS GROMACS simulations of all six AMPK–flavonoid complexes, followed by RMSD/RMSF analyses and MMPBSA calculations, indicate that resveratrol, tangeritin, curcumin, and compound C form the most stable and energetically favourable complexes, with binding free energies consistent with sustained target engagement. Collectively, these findings identify Resveratrol, Tangeritin, and Curcumin as potential AMPK-binding flavonoids predicted to exhibit favorable stability and brain permeability in silico support for neuroprotective efficacy, highlighting them as prioritised natural scaffolds for experimental validation and future development as disease-modifying neuroprotective agents.