Source: https://github.com/aipoch/medical-research-skills

DiffDock Molecular Docking

When to Use

• Blind docking when you have a protein structure (PDB) and a ligand (SMILES) but no known binding site.
• Pose prediction to generate multiple plausible 3D binding conformations and rank them.
• Virtual screening support to quickly evaluate candidate ligands by predicted binding poses and confidence.
• Drug discovery workflows where you need automated docking outputs (SDF poses + scores) for downstream analysis.
• Batch/advanced docking when running many ligand–protein pairs or using alternative inputs (e.g., sequence-based workflows; see references/workflows_examples.md).

Key Features

• Diffusion generative sampling to produce diverse ligand binding poses.
• Confidence model scoring to rank predicted poses.
• Simple CLI inference for single protein–ligand docking.
• Batch/advanced workflows documented in references/workflows_examples.md.
• Structured outputs including ranked SDF pose files and a confidence score report.