RojhanGenomics/tp53-prima1-md-simulation: v2.0.0 - Post-review revision: topology fix, additional analyses (IE, decomposition, DSSP, Zn), negative control, 4th replicate, cross-engine docking validation

Abstract

[Post-submission revision] - 2026-09 Following a desk rejection from JBSD (concerns: depth of analysis, novelty, rationale clarity), the following additional analyses and fixes were made. Fixed Protein topology (Leu289 truncation): topology/topol_Protein_chain_A.itp was found to be missing 11 atoms (CD1/CD2 side chain + backbone C/O) of the C-terminal residue Leu289, apparently truncated during unrelated covalent- adduct exploration. Rebuilt and atom-count-verified (194 residues, 3010 atoms) directly from structures/2OCJ_chainA_clean.pdb (see scripts/topology_rebuild/). This truncation affected only auxiliary MM-PBSA/decomposition post-processing on Run1's already-completed trajectory, not the trajectories themselves (Run1-3 production used a correctly-built system, as independently confirmed from md.tpr). Added Interaction Entropy (Duan et al. 2016) computed for Run1's MM-PBSA trajectory: -T*dS = -3.87 kcal/mol; dG_eff (dH + IE) = -14.90 kcal/mol (see scripts/mmpbsa_extra/interaction_entropy_from_csv.py). Per-residue MM-PBSA decomposition for Run1, identifying Thr102, Gln104, Asn131, Asn268, Tyr126, and Leu111 as the principal favorable contributors, consistent with the blind-docking pocket (see scripts/mmpbsa_extra/parse_decomp_from_mdout.py). DSSP secondary-structure analysis across all three original PRIMA-1 replicates: stable fold composition (~58-60% coil, ~33-35% beta-sheet, ~6.5-7% helix) independent of ligand retention/dissociation. Structural Zn2+ coordination analysis: the nonbonded Zn2+ ion is found to leave its crystallographic coordination site (Cys176/His179/Cys238/ Cys242) within ~10 ps of NVT equilibration in all three replicates, settling into a standard 6-coordinate aqua complex by production (see scripts/zn_analysis/). This is a known limitation of nonbonded, non-polarizable Zn2+ force-field models and is documented as such in the revised manuscript's Limitations section; it is independent of and does not affect the PRIMA-1 replicate-dependence finding. Negative-control simulation (quinuclidine, the PRIMA-1 scaffold without its hydroxymethyl/ketone substituents): full docking-to-20ns-MD pipeline. Result: quinuclidine also dissociated from the L1/S3 pocket (final distance 2.12 nm from an initial 0.28 nm), indicating that pocket dissociation is not specific to PRIMA-1 and that the functional substituents (not the bicyclic scaffold alone) likely contribute to PRIMA-1's retention in replicates where it remains bound (see scripts/negative_control/). Run4: a fourth independent 20-ns PRIMA-1 replicate, run to strengthen the replicate-dependence statistics. Result: ligand remained closely associated with the pocket throughout (mean distance 0.41 +/- 0.09 nm), bringing the overall tally to 2 of 4 replicates retained and 2 of 4 dissociated. Cross-engine docking validation: blind docking repeated with an independent AutoDock Vina version (1.1.2, 2011 scoring function) on a separate machine, cross-validating the L1/S3 pocket identified with Vina 1.2.5. Top pose affinity matched exactly (-4.50 kcal/mol); 9/9 primary pocket residues independently recovered (see scripts/cross_validation/). Reorganized Scripts and structures from an abandoned covalent MQ-Cys124 MD attempt (not part of the reported results) moved to archive/covalent_exploration/ for transparency, kept out of the main reproducible pipeline.