NCBI Literature Search

NCBI Literature Search is an MCP server that connects AI assistants to NCBI databases such as PubMed for scientific literature exploration. It is built for researchers in life sciences, evolutionary biology, bioinformatics, and computational biology who need direct access to academic papers from their development and chat environments. The server enables users to perform keyword queries with boolean logic and PubMed field tags, retrieve article abstracts, fetch author lists, extract DOIs, and access Medical Subject Headings (MeSH). It also supports multi-criteria filtering by journal, author, and publication type, as well as discovery of related articles using NCBI relationship algorithms. Users can run local Python or Poetry installations, deploy via Docker with Redis caching, and monitor operational analytics like cache hit rates and query frequencies.

Category: Notes & Knowledge Management

Tags: biology, ncbi, pubmed, research, scientific-literature

Visit NCBI Literature Search

How to install and configure NCBI Literature Search

1. Initialize the project with dependencies using Poetry: bash mkdir ncbi-mcp-server && cd ncbi-mcp-server poetry init Add mcp, httpx, and typing-extensions during initialization. 2. Create the source file directory: bash mkdir -p src/ncbi_mcp_server Save the server implementation into src/ncbi_mcp_server/server.py and run poetry install. 3. Configure Claude Desktop by editing claude_desktop_config.json: json { "mcpServers": { "ncbi-literature": { "command": "poetry", "args": ["run", "python", "src/ncbi_mcp_server/server.py"], "cwd": "/FULL/PATH/TO/YOUR/ncbi-mcp-server" } } } 4. Restart Claude Desktop to begin querying.

What you can do with NCBI Literature Search

• Running targeted PubMed queries using boolean operators and title tags to find phylogenetic studies or genomic methods. - Fetching complete article metadata including abstracts, authors, MeSH terms, and DOIs via PubMed IDs. - Conducting multi-criteria literature reviews filtered by date ranges, specific journals like Nature, and publication types. - Identifying related research papers algorithmically through NCBI's citation graph based on a single reference PMID. - Searching standardized MeSH terms to improve biomedical search accuracy and vocabulary consistency across queries. - Tracking research analytics and request cache performance across historical PubMed operational metrics.

Key facts

• https://github.com/vitorpavinato/ncbi-mcp-server
• Notes & Knowledge Management, Web Search & Research
• biology, ncbi, pubmed, research, scientific-literature

Part of MCP Servers

What is NCBI Literature Search?

NCBI Literature Search is a Model Context Protocol server that connects AI assistants to NCBI databases like PubMed. It enables users to execute structured scientific literature searches, pull abstracts, locate related papers, and inspect MeSH terms directly through conversational queries.

How do I install NCBI Literature Search?

You can install it by setting up a Python environment with Poetry, installing dependencies like mcp and httpx, and placing the server script in src/ncbi_mcp_server. Then, register the server command and path in your Claude Desktop configuration file and restart the client.

What tools are included in NCBI Literature Search?

The server includes tools such as search_pubmed, get_article_details, search_mesh_terms, get_related_articles, and advanced_search. It also features performance tracking tools like get_analytics_summary and get_detailed_metrics for monitoring search patterns.

Do I need an NCBI API key to use this server?

An NCBI API key is optional. Without an API key, NCBI limits requests to 3 per second. Supplying an API key and registered email in your configuration increases the rate limit to 10 requests per second and facilitates bulk retrieval.

Can I run NCBI Literature Search with Docker?

Yes, Docker deployment is supported. You can launch a standalone container using docker-compose.simple.yml or use docker-compose up with Redis to enable persistent caching and the Redis Commander monitoring UI.