14 Biomedical Databases, One Question — Drug Discovery in 10 Minutes

The Problem

You're researching a rare disease. Or evaluating a drug repurposing opportunity. Or writing a grant proposal. Or doing due diligence on a biotech company's pipeline.

In any of these scenarios, you need data from multiple biomedical databases:

• ChEMBL for compound bioactivity and drug-indication mappings
• OpenTargets for disease-target associations with evidence scores
• PubMed for literature and clinical evidence
• ClinicalTrials.gov for active and completed trials
• OpenFDA for adverse event reports and safety signals
• OMIM for genetic disease associations
• Reactome/KEGG for pathway analysis
• UniProt for protein function data

Each database has its own interface, its own query syntax, its own output format. A researcher manually querying all of them for one disease spends hours — and that's if they know all the databases exist.

Traditional workflow:

1. Search ChEMBL. Export results to spreadsheet.
2. Search OpenTargets. Cross-reference with ChEMBL hits.
3. Search PubMed for supporting literature. Read 30 abstracts.
4. Check ClinicalTrials.gov for ongoing studies.
5. Check OpenFDA for safety signals on candidate drugs.
6. Map pathways in Reactome.
7. Synthesize everything manually into a coherent picture.

Time: 2-3 days for a thorough analysis. Reality: Most people check 2-3 databases and call it done.

The Solution

The Medical Research Toolkit skill gives your AI agent a unified interface to 14+ biomedical databases. One question triggers coordinated queries across all relevant databases, with results synthesized into a structured analysis.

Your agent doesn't just search — it cross-references, validates, and connects findings across databases to build a complete picture.

The Process

You message your agent:

What drugs exist for myasthenia gravis — approved and
investigational? What are the main targets, and are
there any active clinical trials?

The agent orchestrates queries across multiple databases:

Query 1 — ChEMBL: Drugs by Indication

curl -X POST https://mcp.cloud.curiloo.com/tools/unified/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"tools/call",
       "params":{"name":"chembl_find_drugs_by_indication",
       "arguments":{"indication":"myasthenia gravis",
       "max_results":20}},"id":1}'

Returns: pyridostigmine (approved, Phase 4), eculizumab (approved, Phase 4), efgartigimod (approved, Phase 3), rituximab (Phase 3, off-label), ravulizumab (Phase 3), zilucoplan (Phase 3), rozanolixizumab (Phase 3)...

Query 2 — OpenTargets: Disease Targets

# First: resolve disease ID
→ myasthenia gravis = EFO_0004991

# Then: get target associations
→ Top targets by evidence score:
   CHRNA1 (0.89) — acetylcholine receptor
   MUSK (0.82) — muscle-specific kinase
   C5 (0.78) — complement component 5
   FCGRT (0.71) — neonatal Fc receptor
   LRP4 (0.65) — LDL receptor related protein 4

Query 3 — ClinicalTrials.gov: Recruiting Studies

→ 47 active/recruiting trials found
→ Top highlights:
   - NCT05456789: Nipocalimab (anti-FcRn), Phase 3, 340 patients
   - NCT05234567: CAR-T for refractory MG, Phase 1/2
   - NCT05678901: Complement C2 inhibitor, Phase 2
   - NCT05345678: Subcutaneous efgartigimod, Phase 3

Query 4 — OpenFDA: Safety Signals for Lead Compounds

→ Eculizumab adverse events (top 5):
   - Headache (23% of reports)
   - Nasopharyngitis (18%)
   - Upper respiratory infection (15%)
   - Meningococcal infection (4% — boxed warning)
   - Serious infections requiring hospitalization (8%)

Query 5 — PubMed: Recent Literature

→ 12 systematic reviews published 2024-2026
→ Key finding: FcRn inhibitors showing superior efficacy
   to complement inhibitors in generalized MG
→ 3 head-to-head comparison studies underway

Synthesized Report The agent assembles everything into a structured analysis:

## Myasthenia Gravis — Drug & Target Landscape

### Approved Therapies (4)
| Drug | Mechanism | Target | Year |
|------|-----------|--------|------|
| Pyridostigmine | AChE inhibitor | Acetylcholinesterase | 1955 |
| Eculizumab | Anti-C5 | Complement C5 | 2017 |
| Efgartigimod | Anti-FcRn | Neonatal Fc receptor | 2021 |
| Ravulizumab | Anti-C5 (long-acting) | Complement C5 | 2022 |

### Investigational (Phase 2-3): 8 compounds
### Key Targets: 5 validated (CHRNA1, MUSK, C5, FCGRT, LRP4)
### Active Trials: 47 (12 Phase 3, 23 Phase 2)
### Safety Flag: Meningococcal risk with complement inhibitors
### Emerging: FcRn inhibitors overtaking complement pathway

### Repurposing Opportunity
Rituximab (approved for lymphoma/RA) showing Phase 3
efficacy data in anti-MuSK MG — label expansion likely.

The Results

Use Cases

• Pharma R&D: Drug repurposing screening across indications
• Biotech due diligence: Evaluate pipeline credibility against public data
• Grant writing: Comprehensive background for research proposals
• Clinical research: Find active trials for patient referral
• Drug safety: Post-marketing surveillance signal detection
• Academic research: Literature + data integration for publications

Setup on MrChief

skills:
  - medical-research-toolkit

No API keys needed — the toolkit queries public databases through a unified MCP endpoint. Just install and ask.