Computational supporting data for the article "Molecular docking and molecular dynamics dataset: differential accommodation of ivermectin by the p53-binding pockets of MDM2 and MDMX"

Complete computational dataset supporting a study of how the homologousp53-binding pockets of MDM2 and MDMX accommodate ivermectin, a large flexiblemacrocyclic ligand unlike the compact inhibitors developed against these targets. Three candidate sites were screened by triplicate AutoDock Vina docking afterredocking each receptor's native co-crystallised ligand validated the protocol;the MDM2-MDMX RING interface failed a reproducibility criterion and was excludedbefore any simulation. The two p53-binding pockets (PDB 4HG7 and PDB 3U15) werecarried into three independent 100 ns GROMACS trajectories each, with matchedligand-free (apo) controls run in triplicate under an identical protocol, andfour literature reference systems - Nutlin-3 and RO-5963 at MDM2, lithocholicacid and RO-5963 at MDMX - also run in triplicate through the same pipeline. The deposit contains docking inputs and poses, receptor and ligand preparationfiles, GROMACS topology/index/mdp files, per-replicate trajectory analyses(RMSD, RMSF, radius of gyration, hydrogen bonds, contacts, ProLIF interactionfingerprints, SASA, clustering, PCA/RMSIP, free-energy landscapes), MM-GBSA andPoisson-Boltzmann inputs and per-replicate outputs under two nonpolar treatments,pocket-geometry analyses under an a priori pocket definition taken from the p53peptide contacts of the reference complexes, ensemble-redocking robustness data,and every script that generates the figures and tables from those outputs. Several results in this deposit are negative and are reported as such: thesize-normalised ranking of ivermectin against a reference compound invertsbetween generalised Born and Poisson-Boltzmann solvation applied to identicalframes; the dispersion-decomposed nonpolar model fails its own positive control;the published admissibility criterion for both entropy estimators tested fails inevery run; and one replicate of a published reference inhibitor left its startingpose. The dataset supports a structural hypothesis about differentialaccommodation and does not establish experimental affinity, target engagement,inhibition or therapeutic activity. Raw trajectory files (.xtc/.trr) are not included because of their size. Theprocessed per-frame and per-replicate outputs included here are sufficient toreproduce every figure and statistic reported. Raw trajectories are availablefrom the corresponding author on reasonable request.