Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT and other tools is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AI Fire in da houseCheck it out free

bio-mcp

MCP.Pizza Chef: acashmoney

BioMCP is an MCP server that enhances large language models with advanced protein structure analysis and disease-related protein search capabilities. It integrates with the RCSB Protein Data Bank to provide detailed active site examination and functional residue insights, facilitating biomedical research and drug discovery workflows. BioMCP supports agent-based biomedical R&D by enabling real-time, structured protein data access and analysis.

Use This MCP server To

Analyze protein active sites using PDB IDs for drug target validation Search for proteins linked to specific diseases or medical conditions Integrate protein structure data into biomedical AI workflows Enable agents to perform protein function and binding site analysis Support research in protein-disease associations for therapeutic discovery

README

BioMCP: Enabling agent-based biomedical R&D

smithery badge BioMCP

Overview

BioMCP is a Model Context Protocol (MCP) server designed to enhance large language models with protein structure analysis capabilities. It provides tools for analyzing protein active sites and searching for disease-related proteins by interfacing with established protein databases.

Future work will be centered around enabling agents to utilize the BioMCP.

Features

  • Active Site Analysis: Examine the binding sites and functional residues of proteins using PDB IDs
  • Disease-Protein Search: Find protein structures associated with specific diseases or medical conditions
  • Integrated Data Access: Connect seamlessly with RCSB Protein Data Bank (PDB)

Technical Details

BioMCP implements the Model Context Protocol, allowing language models to access specialized protein structure knowledge without requiring this information to be part of their training data. The server handles API connections, data formatting, and error handling to provide reliable protein structure insights.

API Endpoints

BioMCP exposes two primary tools:

  1. analyze-active-site: Provides detailed information about protein binding sites using a PDB ID
  2. search-disease-proteins: Returns proteins related to specified diseases or medical conditions

Getting Started

Installing via Smithery

To install BioMCP for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @acashmoney/bio-mcp --client claude

Manual Installation

# Clone the repository
git clone https://github.com/acashmoney/bio-mcp.git

# Install dependencies
npm install

# Start the server
npm start

Setup Instructions

Running the MCP Inspector

  1. Start the BioMCP server:

    npm start
  2. In a separate terminal, install the MCP Inspector globally (if not already installed):

    npm install -g @anthropic-ai/mcp-inspector
  3. Launch the MCP Inspector and connect to your local BioMCP server:

    npx @modelcontextprotocol/inspector node build/index.js
  4. Use the inspector interface to test tools and view responses.

Using with Claude Desktop

  1. Build the BioMCP server:

    npm run build
  2. Configure Claude Desktop to launch the MCP server:

    a. Locate your Claude Desktop config.json file (typically in your user directory)

    b. Edit the config.json to include the BioMCP server build path. Example configuration:

    {
      "globalShortcut": "",
      "mcpServers": {
        "bio-mcp": {
          "command": "node",
          "args": [
            "/path/to/your/build/index.js"
          ]
        }
      }
    }

    c. Replace /path/to/your/build with your actual path to the project directory.

  3. Restart Claude Desktop for the changes to take effect.

  4. You can now ask Claude questions that utilize the BioMCP tools:

    • "What are the key residues in the active site of PDB structure 6LU7?"
    • "Find proteins related to Alzheimer's disease"

Example Usage

When integrated with a compatible language model, Bio-MCP enables queries like:

  • "What are the key residues in the active site of PDB structure 6LU7?"
  • "Find proteins related to Alzheimer's disease"

Requirements

  • Node.js 20.0.0 or higher
  • TypeScript 5.0+
  • Compatible MCP client implementation

Testing

BioMCP includes a comprehensive testing suite with unit, integration, and end-to-end tests.

Running Tests

Run all tests:

npm test

Run specific test suites:

# Unit tests only
npm run test:unit

# Integration tests only (API interactions)
npm run test:integration

# End-to-end tests only
npm run test:e2e

Linting

Check code quality:

npm run lint

Fix linting issues automatically:

npm run lint:fix

Roadmap

  • Expand level of detail for active site descriptions
  • Leverage 3-D coordinates
  • Tools for interfacing with literature
  • Tools for interfacing with computational biology models:
    • RFdiffusion
    • ProteinMPNN
    • ColabFold
    • Additional protein design and structure prediction tools
  • Agent-based research pipelines
  • Introduce client with protein visualization tools

bio-mcp FAQ

How does BioMCP connect to protein databases?
BioMCP integrates directly with the RCSB Protein Data Bank to access protein structure data.
Can BioMCP analyze specific protein binding sites?
Yes, it provides detailed active site and functional residue analysis using PDB IDs.
Is BioMCP designed for real-time agent use?
Future development aims to enable agents to utilize BioMCP for dynamic biomedical research tasks.
What types of diseases can BioMCP search proteins for?
It can search for proteins associated with a wide range of diseases and medical conditions.
How does BioMCP enhance large language models?
By providing structured protein data and analysis tools, it enriches LLMs with biomedical context.
What technical standards does BioMCP follow?
BioMCP implements the Model Context Protocol for secure, scoped, and observable model interactions.
Can BioMCP be integrated into existing biomedical workflows?
Yes, it is designed to seamlessly connect with AI workflows requiring protein structure insights.