Reproducibility archive for "Assessing Rank Stability and Protocol Sensitivity in Residue-Level MM/PBSA and MM/GBSA Decomposition"

Reproducibility archive accompanying the manuscript “Assessing Rank Stability and Protocol Sensitivity in Residue-Level MM/PBSA and MM/GBSA Decomposition”. This version corresponds to DecompStat v1.0.0 (software DOI: 10.5281/zenodo.21935187). It contains the canonical per-frame inputs, metadata sidecars, fixed residue mappings, manuscript and Supporting Information source tables, publication figures and their source data, analysis and verification scripts, structural-context files, software-environment provenance, and SHA-256 checksums used for the PC1/p17 and Diclofenac–COX2 case studies. For the PC1/p17 case study, the corrected partition treats residues 1–853 as receptor and residues 854–870 as the 17-residue p17 peptide. GB2, GB5, GB8, and PB calculations use the same 250 physical snapshots. The archive contains the corrected canonical table used for all reported PC1/p17 analyses. The archive supports reproduction of the manuscript-level statistical summaries, protocol comparisons, rank-agreement analyses, rank-stability diagnostics, tables, figures, and Supporting Information results. Large external molecular-dynamics trajectories are not duplicated and remain available from the cited public Figshare records. DecompStat operates downstream of endpoint MM/PBSA and MM/GBSA calculations. The software does not perform trajectory preparation, topology construction, or endpoint energy evaluation