Supporting data for: Putative phytochemical constituents of Caesalpinia bonduc (L.) Roxb. seeds as candidate ligands of enzymes associated with Alzheimer's disease and type 2 diabetes mellitus: an integrated LC-QTOF-MS/MS and in silico study

This deposit contains the input files, trajectories and analysis scripts needed to reproduce the molecular docking and molecular dynamics results reported in the manuscript. Docking: prepared receptor structures for six enzymes associated with Alzheimer's disease and type 2 diabetes mellitus (BACE1, acetylcholinesterase, GSK-3 beta, alpha-amylase, alpha-glucosidase and DPP-4), prepared ligand structures, AutoDock Vina configuration and log files, docked poses, native-ligand redocking results, and the SMILES strings of the three putatively annotated constituents. Molecular dynamics: for the hokbusine B-BACE1 and naringenin-4',7-dimethyl ether-BACE1 complexes, GROMACS topologies and ligand parameters, run-parameter (.mdp) files, portable run input (.tpr), mdrun logs, PBC-corrected starting structures, and the 100 ns protein-ligand trajectories written every 50 ps (2001 frames each) that were used for all reported analyses. ADMET: input SMILES and the raw outputs returned by pkCSM and ADMETlab 2.0. Analysis scripts: Python scripts (MDAnalysis 2.11, NumPy, Matplotlib, PyMOL 3.2) that produce the reported RMSD, RMSF, radius of gyration, ligand-dyad distance, hydrogen-bond and contact occupancies with block-averaged uncertainties, and the published figures. Simulations were run with GROMACS 2024.5 (CUDA) on an NVIDIA T4 GPU using the AMBER99SB-ILDN protein force field, GAFF2 ligand parameters with AM1-BCC charges generated by ACPYPE, TIP3P water and 0.15 M NaCl, for 100 ns at 310 K and 1 bar.