内容摘录
BioMCP
BioMCP is a single-binary CLI and MCP server for querying biomedical databases.
One command grammar, compact markdown output, 12 entities across 15+ data sources.
Install
PyPI tool install
This installs the biomcp binary on your PATH.
Binary install
Install skills
Install guided investigation workflows into your agent directory:
MCP clients
From source
Quick start
Command grammar
Entities and sources
| Entity | Sources | Example |
|--------|---------|---------|
| gene | MyGene.info, UniProt, Reactome, QuickGO, STRING, CIViC | biomcp get gene BRAF pathways |
| variant | MyVariant.info, ClinVar, gnomAD, CIViC, OncoKB, cBioPortal, GWAS Catalog, AlphaGenome | biomcp get variant "BRAF V600E" clinvar |
| article | PubMed, PubTator3, Europe PMC | biomcp search article -g BRAF --limit 5 |
| trial | ClinicalTrials.gov, NCI CTS API | biomcp search trial -c melanoma -s recruiting |
| drug | MyChem.info, ChEMBL, OpenTargets, Drugs\@FDA, CIViC | biomcp get drug pembrolizumab targets |
| disease | Monarch Initiative, MONDO, CIViC, OpenTargets | biomcp get disease "Lynch syndrome" genes |
| pathway | Reactome, g:Profiler | biomcp get pathway R-HSA-5673001 genes |
| protein | UniProt, InterPro, STRING, PDB/AlphaFold | biomcp get protein P15056 domains |
| adverse-event | OpenFDA (FAERS, MAUDE, Recalls) | biomcp search adverse-event -d pembrolizumab |
| pgx | CPIC, PharmGKB | biomcp get pgx CYP2D6 recommendations |
| gwas | GWAS Catalog | biomcp search gwas --trait "type 2 diabetes" |
| phenotype | Monarch Initiative (HPO) | biomcp search phenotype "HP:0001250" |
Cross-entity helpers
Pivot between related entities without rebuilding filters:
Gene-set enrichment
Sections and progressive disclosure
Every get command supports selectable sections for focused output:
API keys
Most commands work without credentials. Optional keys improve rate limits:
Multi-worker deployment
BioMCP rate limiting is process-local. For many concurrent workers, run one shared
biomcp serve-http endpoint so all workers share a single limiter budget:
Skills
14 guided investigation workflows are built in:
| # | Skill | Focus |
|---|-------|-------|
| 01 | variant-to-treatment | Variant annotation to treatment options |
| 02 | drug-investigation | Drug mechanism, safety, alternatives |
| 03 | trial-searching | Trial discovery and patient matching |
| 04 | rare-disease | Rare disease evidence and trial strategy |
| 05 | drug-shortages | Shortage monitoring and alternatives |
| 06 | advanced-therapies | CAR-T and checkpoint therapy workflows |
| 07 | hereditary-cancer | Hereditary cancer syndrome workup |
| 08 | resistance | Resistance mechanisms and next-line options |
| 09 | gene-function-lookup | Gene-centric function and context |
| 10 | gene-set-analysis | Enrichment, pathway, and interaction synthesis |
| 11 | literature-synthesis | Evidence synthesis with cross-entity checks |
| 12 | pharmacogenomics | PGx gene-drug interactions and dosing |
| 13 | phenotype-triage | Symptom-first rare disease workup |
| 14 | protein-pathway | Protein structure and pathway deep dive |
Ops
Documentation
Full documentation at biomcp.org.
Getting Started
Data Sources
Quick Reference
Troubleshooting
License
MIT