Biomedical Data Sources for AI Agents¶
BioMCP's User Guide is organized around entities such as genes, variants, articles, trials, and drugs. This Sources section flips the lens: it shows what each upstream database is good at, what BioMCP exposes from it, and where the boundary sits when a workflow is mixed-source.
Use these pages when you already know the provider you trust, the keyword you are targeting, or the provenance you need to explain to a reviewer, teammate, or downstream agent.
Source guides¶
| Source | Best when you want | Guide |
|---|---|---|
| PubMed | Article search, PubTator annotations, and PMC full-text handoff | PubMed |
| ClinicalTrials.gov | Recruiting-study search, eligibility text, and site details | ClinicalTrials.gov |
| ClinVar | Clinical significance and review-status context for variants | ClinVar |
| OpenFDA | FAERS, recalls, device events, labels, and U.S. approval context | OpenFDA |
| UniProt | Canonical protein cards and structure-linked context | UniProt |
| gnomAD | Population frequency and gene constraint context | gnomAD |
| Reactome | Pathway records, pathway genes, and contained events | Reactome |
| Semantic Scholar | TLDRs, citation graphs, references, and recommendations | Semantic Scholar |
| ChEMBL | Drug-target activity, mechanism context, and indication enrichment | ChEMBL |
| OpenTargets | Target-disease scores, druggability, and disease-gene evidence | OpenTargets |
| CIViC | Clinical variant evidence, therapy context, and disease-associated variants | CIViC |
| OncoKB | Oncology actionability tiers and treatment implications for actionable variants | OncoKB |
| cBioPortal | Cancer cohort frequencies and local study analytics workflows | cBioPortal |
| EMA | EU regulatory, safety, and shortage context for medicines | EMA |
| KEGG | KEGG pathway IDs, summary cards, and pathway genes | KEGG |
| PharmGKB / CPIC | Pharmacogenomic recommendations, frequencies, and clinical annotations | PharmGKB / CPIC |
| Human Protein Atlas | Tissue expression, localization, and cancer-expression context | Human Protein Atlas |
| Monarch Initiative | Phenotype-to-disease matching, disease genes, and model evidence | Monarch Initiative |
Reference and setup¶
- Data Sources explains runtime behavior, endpoints, auth mode, and operational caveats.
- Source Licensing and Terms explains direct vs indirect provenance, redistribution limits, and provider terms.
- API Keys shows the optional or required environment variables that upgrade selected source paths.