jessicalh/biorest-open-mcp

Protein MCP Server

A comprehensive Model Context Protocol (MCP) server providing access to 12 major protein databases with 29 analysis tools.

Overview

Extract everything about a protein from authoritative sources:

• 🔬 Sequence & evolutionary data
• 🧬 Structure (predicted & experimental)
• ⚙️ Function & pathways
• 🔗 Protein interactions
• 📊 Domains & families

🎓 NEW in v1.1.0: Automatic Citations

Every response includes proper scientific citations for all data sources:

• ✓ Database citations with versions
• ✓ Primary publications with DOIs
• ✓ BibTeX entries
• ✓ "How to cite" instructions
• ✓ Ready for scientific publication

Databases Included

✅ Fully Tested & Working (100%)

• UniProt - Protein sequences & annotations
• AlphaFold - AI-predicted structures (200M+ proteins)
• PDB (RCSB) - Experimental 3D structures
• STRING - Protein-protein interactions (v12.5 with regulatory networks)
• InterPro/Pfam - Protein families & domains
• KEGG - Metabolic pathways
• Ensembl - Comparative genomics
• Gene Ontology - Functional annotations
• DisProt - Intrinsically disordered proteins
• PROSITE - Domain patterns
• NCBI/BLAST - Sequence search & alignment

⚠️ Partially Working

• Reactome - Pathways (75% functional)

Quick Start

For Claude Desktop Users

Easiest method (after package is published to npm):

1. Add to claude_desktop_config.json:

   {
     "mcpServers": {
       "protein": {
         "command": "npx",
         "args": ["-y", "biorest-open-mcp"]
       }
     }
   }
   

2. Restart Claude Desktop

3. Start analyzing proteins!

For Developers

# Clone repository
git clone https://github.com/jessicalh/biorest-open-mcp.git
cd biorest-open-mcp

# Install dependencies
npm install

# Build
npm run build

# Run tests (optional)
npm test

# Start server
npm start

Example Queries

Once connected to Claude Desktop, try:

Search UniProt for human p53
Get AlphaFold structure prediction for P04637
Find protein interactions for BRCA1 and TP53 in humans
Get GO annotations for insulin
Scan sequence MTEYKLVVVG... against PROSITE
Analyze pathway enrichment for BRCA1, BRCA2, TP53

Available Tools (29 total)

See for complete documentation of all tools.

Quick Reference:

• uniprot_search, uniprot_get_entry, uniprot_map_ids
• alphafold_get_prediction, alphafold_get_pdb, alphafold_get_confidence
• pdb_search_text, pdb_search_sequence, pdb_get_entry, pdb_download_file
• string_get_network, string_get_enrichment
• interpro_search, interpro_scan_sequence
• reactome_search, reactome_analyze
• kegg_get_pathway, kegg_search_genes
• ensembl_lookup, ensembl_get_sequence, ensembl_get_homology
• go_get_term, go_get_annotations
• disprot_search
• prosite_scan
• ncbi_search, ncbi_fetch, ncbi_blast

Configuration

Optional: NCBI API Key

For higher rate limits (10 req/sec instead of 3 req/sec):

1. Get free API key: https://www.ncbi.nlm.nih.gov/account/
2. Add to config:

   "env": {
     "NCBI_API_KEY": "your_key_here",
     "NCBI_EMAIL": "your_email@example.com"
   }
   

Testing

Run the comprehensive test suite:

npm test

Test Results: 54/59 tests passing (91.5%) - See

Documentation

• - Complete reference for all 31 tools
• - Test coverage and results
• - Publishing and distribution guide
• - Version history

Performance

• Fast: Most queries < 2 seconds
• Moderate: AlphaFold, Reactome (2-5 seconds)
• Slow: DisProt (13-14 seconds) - consider caching

Requirements

• Node.js >= 18
• No external dependencies required
• Optional: NCBI API key for higher rate limits

Authentication

Good news: Most databases require no authentication!

• ✅ No auth: UniProt, AlphaFold, PDB, STRING, InterPro, Ensembl, GO, DisProt, PROSITE, Reactome
• 🔑 Optional: NCBI API key (for higher rate limits)
• 📚 Academic: KEGG (academic use only)

Contributing

Contributions welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Add tests for new features
4. Submit a pull request

Support

• 📖
• 🐛 Report Issues
• 💬 Discussions

License

MIT License - see file for details

Citation

If you use this tool in your research, please cite:

@software{biorest_open_mcp,
  title = {BioREST Open MCP: Comprehensive Protein Analysis via Model Context Protocol},
  author = {Jessica H},
  year = {2025},
  url = {https://github.com/jessicalh/biorest-open-mcp}
}

Acknowledgments

This server integrates data from:

• UniProt Consortium
• AlphaFold Database (DeepMind/EMBL-EBI)
• RCSB Protein Data Bank
• STRING Database
• InterPro/Pfam
• Reactome
• KEGG
• Ensembl
• Gene Ontology Consortium
• DisProt
• PROSITE (ExPASy)
• NCBI

Built with the Model Context Protocol SDK