Protein Mcp Server
Let your agent query RCSB PDB, PDBe, and UniProt for protein structures and metadata while you build bioinformatics or ML tooling.
Overview
Protein MCP Server is a Build-phase MCP server that retrieves and analyzes 3D protein structural information from RCSB PDB, PDBe, and UniProt for agent-driven bioinformatics workflows.
What is this MCP server?
- Retrieves 3D protein structural data and related metadata from RCSB PDB, PDBe, and UniProt
- npm package protein-mcp-server v1.0.3 with stdio and streamable-http (localhost:3010/mcp)
- Bun-based local run via start:stdio or start:http
- MCP_LOG_LEVEL env for tuning agent-session verbosity
- Suited to structure-aware analysis prompts without manual PDB site navigation
- Package version 1.0.3
- npm identifier protein-mcp-server
- Upstream sources: RCSB PDB, PDBe, UniProt
Community signal: 4 GitHub stars.
What problem does it solve?
Coding agents confidently make up PDB IDs and chain metadata when you are wiring real structural biology features.
Who is it for?
Solo builders creating research copilots, structural biology dashboards, or ML feature pipelines that need agent access to authoritative protein records.
Skip if: Clinical decision support without human expert review, or workflows that need full molecular dynamics simulation inside the agent.
What do I get? / Deliverables
Your agent can call MCP tools backed by PDB, PDBe, and UniProt so structure retrieval steps in your repo match live archive data.
- Agent-callable protein structure and metadata retrieval from PDB, PDBe, and UniProt
- Configured stdio or local HTTP MCP endpoint for structural biology queries
Recommended MCP Servers
Journey fit
Builders add this during implementation when connecting structural biology data sources to agents or apps, which maps to the Build phase shelf. It is a specialized external-data MCP connector, so integrations is the canonical subphase rather than generic backend CRUD.
How it compares
Database MCP bridge for structural biology archives, not a molecular visualization app or a wet-lab LIMS.
Common Questions / FAQ
Who is io.github.cyanheads/protein-mcp-server for?
Developers and researchers building agent-assisted tools that must read protein structures and identifiers from RCSB PDB, PDBe, or UniProt.
When should I use io.github.cyanheads/protein-mcp-server?
During Build when you are implementing integrations that fetch or compare 3D protein data and metadata from standard public structural databases.
How do I add io.github.cyanheads/protein-mcp-server to my agent?
Add the npm package protein-mcp-server to your MCP config, run with Bun using start:stdio for stdio or start:http for http://localhost:3010/mcp.