CDK Pywrapper
Enables Large Language Models to interact with cheminformatics functionalities, performing tasks like compound identification, structure conversion, and property calculation.
About
CDK Pywrapper provides a robust Python interface for the Chemistry Development Kit (CDK), a powerful Java-based cheminformatics library. Its MCP server component extends this capability by integrating with external chemical databases like UNII, ChEMBL, and Guide to Pharmacology data. Designed for tool-use capable Large Language Models, it facilitates a wide range of cheminformatics tasks, from generating and inter-converting chemical identifiers (SMILES, InChI) to searching for compounds by name, calculating basic molecular properties, and even generating SVG representations of chemical structures.
Key Features
- Convert between SMILES, InChI, and InChI Key identifiers.
- Calculate basic compound properties, such as molecular mass.
- Generate SVG images of compound structures.
- Retrieve compound names and synonyms by structure via database lookup.
- Search for chemical compounds by name using an LLM.
- 33 GitHub stars
Use Cases
- Automating chemical structure identification and property retrieval with LLMs.
- Facilitating inter-conversion of chemical identifiers for data processing.
- Enabling LLMs to search external chemical databases for compound details.