What is Augmented Nature Pdb Mcp Server

Augmented-Nature-PDB-MCP-Server is a Model Context Protocol (MCP) server that provides access to the Protein Data Bank (PDB), which is a global repository for information regarding the 3D structures of proteins, nucleic acids, and complex assemblies.

Use cases

Use cases include searching for specific protein structures (e.g., insulin), retrieving detailed information about a PDB structure, downloading structural coordinates for analysis, and finding PDB structures associated with specific UniProt accessions.

How to use

To use the Augmented-Nature-PDB-MCP-Server, install it via npm, configure it in your ‘claude_desktop_config.json’, and utilize various tools like ‘search_structures’, ‘get_structure_info’, and ‘download_structure’ for querying and retrieving data.

Key features

Key features include searching for protein structures by keywords, obtaining detailed structure information, downloading structure coordinates in multiple formats, and accessing structure quality metrics and validation data.

Where to use

The Augmented-Nature-PDB-MCP-Server is applicable in fields such as bioinformatics, structural biology, and pharmaceutical research, where understanding protein structures is crucial.

PDB MCP Server

A Model Context Protocol (MCP) server that provides access to the Protein Data Bank (PDB) - the worldwide repository of information about the 3D structures of proteins, nucleic acids, and complex assemblies.

Features

Tools

• search_structures: Search PDB database for protein structures by keyword, protein name, or PDB ID
• get_structure_info: Get detailed information for a specific PDB structure
• download_structure: Download structure coordinates in various formats (PDB, mmCIF, mmTF, XML)
• search_by_uniprot: Find PDB structures associated with a UniProt accession
• get_structure_quality: Get structure quality metrics and validation data

Resources

• pdb://structure/{pdb_id}: Complete structure information for a PDB ID
• pdb://coordinates/{pdb_id}: Structure coordinates in PDB format
• pdb://mmcif/{pdb_id}: Structure data in mmCIF format
• pdb://validation/{pdb_id}: Structure validation data and quality metrics
• pdb://ligands/{pdb_id}: Ligand and binding site information
• pdb://search/{query}: Search results for structures matching the query

Installation

npm install
npm run build

Usage

With Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "pdb-server": {
      "command": "node",
      "args": [
        "/path/to/pdb-server/build/index.js"
      ]
    }
  }
}

Example Queries

1. Search for insulin structures:

   Use the search_structures tool to find insulin protein structures
   

2. Get information about a specific structure:

   Get detailed information about PDB structure 1ABC using get_structure_info
   

3. Download structure coordinates:

   Download the PDB file for structure 1ABC using download_structure
   

4. Find structures for a UniProt protein:

   Find PDB structures for UniProt accession P01308 using search_by_uniprot
   

5. Check structure quality:

   Get validation metrics for PDB structure 1ABC using get_structure_quality
   

6. Access resources directly:

   Show me the resource pdb://structure/1abc
   

API Integration

This server integrates with:

• RCSB PDB REST API (https://data.rcsb.org/): For structure metadata and information
• RCSB PDB Search API (https://search.rcsb.org/): For advanced search capabilities
• RCSB PDB File Download (https://files.rcsb.org/): For structure file downloads

Structure Data Formats

• PDB: Original Protein Data Bank format
• mmCIF: Macromolecular Crystallographic Information File (newer standard)
• mmTF: Macromolecular Transmission Format (binary, compressed)
• XML: Extensible Markup Language format

Quality Metrics

The server provides various structure quality indicators:

• Resolution (for X-ray crystallography)
• R-work and R-free values
• Ramachandran plot statistics
• Clash scores
• Rotamer outliers
• Model validation scores

Complementary Servers

This PDB server works well with:

• UniProt MCP Server: For protein sequence and functional information
• AlphaFold MCP Server: For predicted protein structures

Together, these provide comprehensive access to protein structure and function data.

Error Handling

The server includes robust error handling for:

• Invalid PDB IDs
• Network connectivity issues
• API rate limiting
• Missing structure files
• Malformed requests

Development

# Install dependencies
npm install

# Build the server
npm run build

# Run in development mode
npm run dev

Attribution

This project is developed by Augmented Nature
🌐 Website: augmentednature.ai