Overview
RDKit is a collection of cheminformatics and machine-learning software written in C++ and Python. It provides the core algorithms for representing chemical molecules, generating 2D and 3D coordinates, building massive molecular graphs, and preparing molecular structures for AI models.
Features
- Molecular Formatting: Supports SMARTS, SMILES, SDF, TDT, and MolBlock reading and writing out of the box.
- Fingerprinting: World-class generation of molecular fingerprints for machine learning similarities.
- Reaction Handling: Chemical reaction manipulation using SMARTS-based reaction templates.
Use Cases
- High throughput combinatorial drug discovery screening.
- AI-based toxicity prediction algorithms.
- Chemical similarity search engines.