Supplementary data for: Omission of the FAD cofactor during receptor preparation reorders natural-product rankings in virtual screening against protoporphyrinogen oxidase: a single-target case study

Supplementary data and code for the manuscript "Omission of the FAD cofactor during receptor preparation reorders natural-product rankings in virtual screening against protoporphyrinogen oxidase: a single-target case study" (submitted to In Silico Research in Biomedicine, manuscript INSILI-D-26-00849). The study docked sixteen compounds (fourteen plant allelochemicals and two reference herbicides) against protoporphyrinogen oxidase (PDB 1SEZ, chain A) with AutoDock Vina 1.2.7, using two receptors that differ only in whether the FAD cofactor is retained. The archive contains: the unmodified RCSB structure and both prepared receptors (FAD-free and FAD-containing, PDBQT); all ligand structure files (PDB and PDBQT) and the structure-verification audit; every docking run (exhaustiveness 32, seeds 1-3) with logs and all output poses; redocking validation files; and the analysis scripts. Revision-1 additions: a docking-box scan (18-26 A), Vinardo rescoring, carboxylic acids docked as anions, five commercial PPO inhibitors, FAD atom-typing variants, hydrogen-bond and residue-distance geometry, raw-score statistics (Steiger test, bootstrap, leave-one-out), Tables S3-S9 as CSV, Figure S1, and a pipeline script that regenerates the revision analyses. See README.txt inside the archive for the full manifest, the mapping between archive and manuscript figure numbers, software versions and Windows run instructions. Software: AutoDock Vina 1.2.7 (1.2.5 for the version-1 control), Meeko 0.7.1, Open Babel, RDKit, NumPy, SciPy, PyMOL, Matplotlib. Note: the archive README corrects a statement in an earlier version: replicate-seed standard deviation was at most 0.02 kcal/mol for the FAD-containing receptor and at most 0.11 kcal/mol for the FAD-free receptor.