Computational data for peptide SPAFESTWDILK/target MDM2: molecular dynamics trajectories, AF3 benchmark, multi-method binding evaluation

README — SPAFESTWDILK / MDM2 computational data deposit Supporting data for the SPAFESTWDILK / MDM2 manuscript (Sherman Tree Nutraceuticals, in preparation). Files listed in the manuscript Section 6 (Data and Code Availability). Letter codes in brackets correspond to the data references [a]-[z] used throughout the manuscript. GROMACS CHARMM36m trajectories 5NS.rar — 17 systems, 5 ns CHARMM36m triage Plant-derived candidates (F-x(1-3)-W pharmacophore): SPAFESTWDILK Zingiber officinale (ginger) [LEAD] FESTWDILK Zingiber officinale (ginger) HAFPELWNIEK Zingiber officinale (ginger) DDFLNSWK Camellia sinensis (green tea) EFSDLWDNK Vanilla planifolia (vanilla) DNEFLQDWSK Artemisia SDLTSFMEEWR Artemisia FQSWEDLSK Cordyceps QNSFVDLWK Grifola frondosa (maitake) ISTAFLNDWDLAK Grifola frondosa (maitake) EHFETLWSSVK plant proteome candidate Hard negatives: EWSLDQSKF Cordyceps (scrambled) FSNLDKWDE Vanilla (scrambled) WSADITKFESPL Zingiber officinale (scrambled) FWELDSTLKLPNEQS scrambled Reference negative controls: AAAAAAAAAAAAAAAAA poly-Ala 17-mer SQETFSDLAKLLPEN p53(15-29) W23A mutant 100NS.rar — SPAFESTWDILK production simulation, 100 ns CHARMM36m. No dissociation. Each archive contains: .pdb, .top/.itp, .mdp, .xtc, .tpr, .gro, .xvg analysis outputs. Run on Google Colab GPU. NutrAI pipeline outputs peptides_hydrolysis.csv — Step 1 hydrolysis library, 8,645,509 unique fragments. Standalone file at deposit root.26_NutrAI_candidates.csv — 26 pharmacophore-passing candidates submitted to AF3.benchmark_3+3_panel.csv — 3+3 AF3 calibration panel. AlphaFold 3 Server raw outputs AF3_29_hard_negatives.zip [e] — Expanded benchmark, 29 hard-negative peptides × 5 seeds.AF3_8_calibration_peptides.zip [d] — 3+3 calibration panel and SPAFESTWDILK runs, 8 peptides × 5 seeds (placed in deposit root; manifest.tsv included inside). Neurosnap independent multi-method evaluation outputs Protenix_structure_predictions.zip [f] — Protenix co-folding, 5 models, best iPTM 0.923. https://neurosnap.ai/job/69df686cc7a74174eaf1a524?share=69e22150eec1918b946621fe Chai-1_structure_predictions.zip [g] — Chai-1 co-folding, 5 models, best iPTM 0.891. https://neurosnap.ai/job/69df6680c7a74174eaf1a51c?share=69e22150eec1918b9466220e EvoEF2_Protenix_structure.zip [h] — EvoEF2 on Protenix rank_1 structure, −55.57 EEU (manuscript reference value). https://neurosnap.ai/job/69df74bfc7a74174eaf1a585?share=69e22151eec1918b94662212 EvoEF2_AF3_structure.json [i] — EvoEF2 on AF3 model_0 structure, −47.27 EEU (cross-structure comparison). MMPBSA_AMBER_50ns.zip [o] — gmx_MMPBSA on AMBER99SB-ILDN trajectory, ΔG = −75.30 ± 4.92 kcal/mol. https://neurosnap.ai/job/69dfc05fc7a74174eaf1a983?share=69e22152eec1918b94662222 MMPBSA_CHARMM27_50ns.zip [p] — gmx_MMPBSA on CHARMM27 trajectory, ΔG = −55.07 ± 2.86 kcal/mol. https://neurosnap.ai/job/69e12743c7a74174eaf1bc52?share=69e22152eec1918b94662224 License: CC BY 4.0