chem-docking-void

What problem does it solve?

Finding plausible binding poses for small molecules inside porous host materials is slow and error-prone when done manually, especially when you need multiple sampled conformers and spatially diverse placements.

Core Features & Use Cases

• Conformer generation and energy ranking: Generates 3D RDKit conformers, optimizes them with MMFF94, and selects the lowest-energy candidates.
• Voronoi-based porous sampling: Places guests throughout a CIF host using VOID’s Voronoi clustering to cover accessible pore space.
• Collision-aware pose acceptance: Applies physics-informed minimum-distance fitness functions to filter clashes between host and guest.
• Practical outputs for downstream research: Exports ranked docked complexes as CIF files plus a docking_results.json summary for traceability and later DFT/MLIP steps.
• Use Case: Prepare candidate zeolite/MOF-inclusion docking poses for follow-up energy evaluation, screening, or catalyst/adsorbate discovery.

Quick Start

Ask the agent to run chem-docking-void with your ligand SMILES and host CIF, requesting it to generate ranked docked CIF poses and a docking_results.json summary.