PubMed MCP Server

Enhanced Python-based PubMed MCP Server with API key support for faster and unlimited downloads.

Features

• Search PubMed for biomedical literature and research papers
• Find similar articles, cited articles, and citing articles
• Retrieve abstracts and full text of open access articles
• Batch search and advanced search capabilities
• API key support for faster and unlimited downloads

API Key and Email Setup

The NCBI E-utilities API recommends using an API key and email address with your requests. This brings several benefits:

• Higher rate limits: 10 requests/second instead of 3
• More results per query: Up to 200 results per request instead of 100
• Priority service: Better queue position for your requests

How to Get an API Key

1. Create an NCBI account at https://www.ncbi.nlm.nih.gov/account/
2. Go to the API Keys Management page
3. Generate a new API key

Setting Up API Key and Email

The easiest way to set up your API key and email is using the setup script:

python setup_api.py

This interactive script will guide you through the process and offer multiple setup options.

Alternatively, you can set up your API key and email manually using one of these three methods:

1. Environment Variables

export NCBI_API_KEY=your_api_key_here
export NCBI_EMAIL=your_email@example.com

2. Command Line Arguments

python python-pubmed-mcp-enhanced.py --api-key your_api_key_here --email your_email@example.com

3. Configuration File

Create a config.json file based on the example:

cp config.json.example config.json

Edit config.json to include your API key and email:

{
  "api_key": "your_api_key_here",
  "email": "your_email@example.com"
}

Then run the server with the config file:

python python-pubmed-mcp-enhanced.py --config config.json

Usage

Installing via Smithery

To install pubmed-mcp-easy for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @t0mst0ne/pubmed-mcp-easy --client claude

Standard Usage

Run the server:

python python-pubmed-mcp-enhanced.py

Claude Desktop Integration

To integrate with Claude Desktop, add the following to your claude_desktop_config.json file:

"pubmed-easy": {
    "command": "/opt/anaconda3/bin/python",
    "args": [
        "/GITHUB_cloned_dir/pubmed-mcp-easy/python-pubmed-mcp-enhanced.py", "--config", "config.json"
    ]
}

Make sure to:

1. Replace /opt/anaconda3/bin/python with the path to your Python executable
2. Replace /GITHUB_cloned_dir with the actual path to your GitHub directory
3. Create a config.json file with your API key and email as described above

After adding this configuration, you can access PubMed tools directly from Claude Desktop.

Available Tools

The server provides the following MCP tools:

• pubmed_search: Search for articles by keyword or query
• pubmed_similar: Find similar articles
• pubmed_cites: Find articles cited by a specific paper
• pubmed_cited_by: Find articles that cite a specific paper
• pubmed_abstract: Retrieve the abstract of an article
• pubmed_open_access: Check if an article is open access
• pubmed_full_text: Retrieve the full text of an open access article
• pubmed_batch_search: Perform multiple searches in parallel
• pubmed_author_search: Search for papers by a specific author
• pubmed_advanced_search: Perform advanced field-based searches
• pubmed_journal_search: Search for papers in a specific journal

Important Notes

1. Including an email address is recommended by NCBI as it allows them to contact you if there are issues with your requests.
2. If you make heavy use of the E-utilities, NCBI recommends that you limit large jobs to either weekends or between 9 pm and 5 am Eastern Time weekdays.
3. Always be considerate in your usage and follow NCBI's usage guidelines.