ToxMCP Suite

ToxMCP is the public suite hub for guardrailed, auditable MCP servers across computational toxicology, exposure science, mechanistic reasoning, QSAR workflows, ADMET utilities, and downstream kinetic modeling.

If you want the single best place to follow the public suite, module launches, and onboarding path, star this repo.

• Preprint (bioRxiv): https://doi.org/10.64898/2026.02.06.703989
• Citation metadata: CITATION.cff

Why Star This Repo?

• It is the canonical public index for the ToxMCP suite.
• It is the best place to track which modules are public now versus still in progress.
• It is the suite-level home for docs, architecture, quickstart guidance, and contribution links.

Architecture

What Is MCP?

MCP (Model Context Protocol) is a standard way for LLM clients and orchestrators to call external tools over a structured interface. See the official MCP docs/spec: https://modelcontextprotocol.io/

Public Suite Today

Module	What it covers	Upstream dependency	Requirements	Status	Repo
CompTox MCP	Identity, hazard, exposure, bioactivity, screening prioritization	EPA CompTox APIs + packaged metadata bundles	API key required (CTX_API_KEY)	Available now	Repo
ADMETlab MCP	Molecule washing, SVG rendering, ADMET prediction, CSV retrieval	ADMETlab 3.0 API	No key by default; upstream instability/rate limits may cause intermittent failures	Available now	Repo
AOP MCP	Mechanistic pathways, scientific review, assay discovery, authoring workflows	AOP-Wiki / AOP-DB / CompTox federation	Internet access recommended; fixture fallback available for offline development	Available now	Repo
O-QT MCP	OECD QSAR workflows, grouping/read-across dossiers, PDF reports	OECD QSAR Toolbox WebAPI	Requires a running QSAR Toolbox WebAPI instance (typically Windows-hosted)	Available now	Repo
PBPK MCP	PBPK simulation, qualification, verification, dossier export	Open Systems Pharmacology Suite + rxode2 runtime	Local worker/runtime setup required (see repo deploy scripts)	Available now	Repo
Direct-Use Exposure MCP	Auditable deterministic external-dose scenario construction, jurisdictional comparison, PBPK-ready handoff packaging	Versioned defaults, packaged archetypes, bounded worker/exchange surfaces	Python 3.12+; local install from repo release assets	Available now	Repo
Dietary Exposure MCP	Coming soon	Coming soon	Coming soon	Coming soon	Repo
Environmental Fate MCP	Coming soon	Coming soon	Coming soon	Coming soon	Repo
Bioactivity-PoD MCP	Coming soon	Coming soon	Coming soon	Coming soon	Repo

Getting Started

The fastest way to get productive with ToxMCP is:

1. Pick an entry point. Start with CompTox MCP for the most recognizable public toxicology data workflow, or use Direct-Use Exposure MCP when you need an auditable exposure object from the start.
2. Follow that module repo’s Quickstart TL;DR.
3. Run its Verification (smoke test) so you know the local setup is working before adding more modules.

Common Local Ports

• CompTox MCP: http://localhost:8000
• O-QT MCP: http://localhost:8001
• PBPK MCP: http://localhost:8002
• AOP MCP: http://localhost:8003
• ADMETlab MCP: http://localhost:8200

Which one should I try first?

• Choose CompTox MCP if you want the best first-run experience and the most recognizable toxicology dataset integration today.
• Choose Direct-Use Exposure MCP if you need reviewable exposure scenarios with explicit assumptions, provenance, limitations, and fit-for-purpose framing.
• Add AOP MCP, O-QT MCP, or PBPK MCP next depending on whether the workflow needs mechanistic context, QSAR Toolbox automation, or toxicokinetic modeling.

Exposure-side stack expansion is still in progress, but the external-exposure layer is now public. Direct-Use Exposure MCP is the current suite module for deterministic exposure screening and governed external-dose handoff generation.

Docs

Start with docs/README.md for the suite-level documentation map.

Notes

• Each MCP server is versioned and released independently.
• Some modules depend on proprietary or rate-limited upstream services. Check each module README for exact setup details.

Acknowledgements / Origins

ToxMCP was developed, in part, in the context of the VHP4Safety project (see: https://github.com/VHP4Safety) and related research and engineering efforts.

Funding: Dutch Research Council (NWO) — NWA.1292.19.272 (NWA programme)

This suite integrates with third-party data sources and services (e.g., EPA CompTox, ADMETlab, AOP resources, OECD QSAR Toolbox, Open Systems Pharmacology). Those upstream resources are owned and governed by their respective providers; users are responsible for meeting any access, API key, rate limit, and license/EULA requirements described in each module.

License

Apache-2.0

✅ Citation

Djidrovski, I. ToxMCP: Guardrailed, Auditable Agentic Workflows for Computational Toxicology via the Model Context Protocol. bioRxiv (2026). https://doi.org/10.64898/2026.02.06.703989

@article{djidrovski2026toxmcp,
  title   = {ToxMCP: Guardrailed, Auditable Agentic Workflows for Computational Toxicology via the Model Context Protocol},
  author  = {Djidrovski, Ivo},
  journal = {bioRxiv},
  year    = {2026},
  doi     = {10.64898/2026.02.06.703989},
  url     = {https://doi.org/10.64898/2026.02.06.703989}
}

Citation metadata: CITATION.cff

Contributing

Please see CONTRIBUTING.md. If you’re not sure which repository to use, open an issue here: https://github.com/ToxMCP/toxmcp/issues

Support

See SUPPORT.md.