Awesome MCP › Databases

genomoncology/biomcp

⭐ 630 Python added to this list on 2025-05-06 repository created 2025-04-01

BioMCP is an open-source toolkit designed to implement the Model Context Protocol (MCP) specifically for the biomedical domain. It serves as a bridge connecting AI systems with authoritative biomedical data sources, enabling these systems to provide precise and in-depth answers related to clinical trials, scientific literature, and genomic variants. The toolkit is certified by MCP Review, ensuring adherence to best practices for MCP implementation and reliable access to biomedical data. The core value of BioMCP lies in its ability to provide structured access to a wide range of biomedical data sources, including peer-reviewed literature from PubTator3/PubMed, preprints from bioRxiv and medRxiv, open science content from Europe PMC, clinical trial data from ClinicalTrials.gov, genetic variant annotations from MyVariant.info, cancer genomics data from TCGA/GDC and cBioPortal, and population frequency data from the 1000 Genomes project via Ensembl. This integration allows AI assistants to perform natural language queries across these diverse datasets without requiring users to know the specific query syntax of each source. BioMCP offers a set of 13 specialized tools to support biomedical research workflows. The three core tools include the Think Tool, which facilitates systematic breakdown and planning of complex biomedical queries; the Search Tool, which supports both a unified query language and domain-specific searches across articles, trials, and variants; and the Fetch Tool, which retrieves detailed information for specific articles, trials, or variants. Additionally, there are 10 individual tools that provide direct access to specific functionalities such as article searching, trial searching, and variant searching, often enhanced with cancer genomics data from cBioPortal. The toolkit supports integration with AI assistants and agents by functioning as an MCP server, and it provides a command-line interface for direct interaction with the underlying databases. It also supports environment variables for optional authentication with external APIs to improve rate limits. BioMCP is designed to empower AI systems with specialized biomedical knowledge, addressing the limitations of large language models in accessing up-to-date and domain-specific biomedical information.

https://github.com/genomoncology/biomcp

1000-genomesaiai-assistantsai-integrationai-platformsai-systemsarticle-searchbioinformaticsbiomcpbiomedical-databiomedical-knowledgebiomedical-model-context-protocolbiomedical-researchbiorxivcbioportalclaude-desktopclinical-trialsclinicaltrials.govcommand-line-interfaceensembleurope-pmcfetch-toolgdcgenomic-variantsgenomicshealthcare-integrationhipaa-compliantllmmcpmcp-inspectormcp-servermedicalmedrxivmit-licensemodel-context-protocolmyvariant.infonatural-language-queriesoncomcpopen-sourceprecision-oncologypubmedpubmed-centralpubtator3scientific-literaturesearch-tooltcgathink-tooltrial-searchvariant-search