drug-docking-vina

Virtual screening with AutoDock Vina generates ranked pose sets by docking prepared small-molecule ligands into a protein receptor. It outputs pose PDBQT files and a JSON report that can be aggregated into a ranked CSV for analysis.

Protein-ligand docking with AutoDock Vina requires prepared PDBQT files for both the receptor and ligands. You must also define an explicit docking box center and size to initiate the pose generation and scoring process.

Molecular docking supports batch virtual screening across a ligand directory. The skill processes multiple ligands consistently and aggregates their docking scores into a machine-readable JSON report for downstream ranking and enrichment.

AutoDock Vina Python API captures docking parameters and configuration metadata to ensure reproducible virtual screening outputs. It writes per-ligand docked pose files alongside a JSON report containing scores and runtimes for consistent result collation.

Collecting and ranking molecular docking results is done using the collect_results.py script, which aggregates per-ligand JSON reports into a ranked CSV. This output supports downstream analysis and allows joining library metadata such as identifiers and labels.

Specifying an explicit docking box center and size is required for virtual screening with AutoDock Vina to constrain the pose generation search space. This ensures the generated ligand poses are accurately scored within the targeted protein receptor binding pocket.