Pubchem Mcp Server

What is Pubchem Mcp Server?

Pubchem Mcp Server is an unofficial Model Context Protocol (MCP) server for querying the PubChem chemical database. It provides access to over 110 million compounds, molecular properties, bioassay data, and chemical informatics tools, serving researchers and developers in drug discovery, toxicology, and computational chemistry.

How to use Pubchem Mcp Server?

Install via npm, build with npm run build, then start with npm start. Invoke tools such as search_compounds, get_compound_info, or search_similar_compounds by sending JSON‑RPC arguments. Resource templates like pubchem://compound/{cid} allow direct data retrieval.

Key features of Pubchem Mcp Server

• Search compounds by name, CAS, SMILES, or InChI
• Similarity and substructure searches with Tanimoto threshold
• Molecular descriptors and ADMET predictions
• Drug‑likeness analysis (Lipinski Rule of Five)
• Bioassay and activity data across 1.5 million assays
• Safety, toxicity, and regulatory information
• Batch processing of up to 200 compounds
• Cross‑references to ChEMBL, DrugBank, and 500+ databases

Use cases of Pubchem Mcp Server

• Lead discovery: search anti‑inflammatory compounds and assess drug‑likeness
• Toxicity screening: retrieve GHS classifications and LD50 data
• Structure‑activity relationship (SAR) analysis: compare bioactivity profiles
• Patent and literature searching: find chemical patents and PubMed references
• Integrate with other MCP servers (UniProt, STRING) for multi‑omics workflows

FAQ from Pubchem Mcp Server

How many requests can I send per second?

PubChem API guidelines limit to 5 requests per second and 400 requests per minute. No API key is required, but respectful usage is encouraged.

What programming language and dependencies are needed?

The server is built with TypeScript and requires Node.js. Dependencies include Axios for HTTP requests and the MCP SDK for protocol compliance.

Where does the data come from?

All data is fetched from the PubChem REST API, which aggregates over 110 million compounds, 1.5 million bioassays, and cross‑references from 500+ external databases.

Is authentication required to use the server?

No. The PubChem API is open and does not require an API key or any authentication tokens.

How many compounds can I process in a single batch?

The batch_compound_lookup tool supports up to 200 compounds per operation, enabling efficient retrieval of properties, safety data, or other descriptors.