This repository does not report docking results. It provides a defensible handoff checklist for candidates that pass model review.
- Select an experimentally appropriate protein structure and document organism, construct, resolution, ligands and missing residues.
- Prepare the receptor consistently: protonation assumptions, waters, cofactors, metals and alternate conformations must be recorded.
- Generate ligand protonation, tautomer and stereochemical states at the intended pH range.
- Define the search box from a co-crystallized ligand or justified binding-site evidence.
- Redock the reference ligand and record pose RMSD before screening candidates.
- Retain poses for expert review; do not interpret a docking score as binding affinity.
- Test sensitivity to structure choice and preparation settings.
- Link each computational decision to an experimental follow-up hypothesis.
The output should include tool versions, configuration, random seeds, receptor/ligand files, pose-review criteria and known failure modes.