Skip to content

DeepLife TwinCell

PyPI version Python 3.12+ License: MIT

Alpha

TwinCell is in alpha. The model and API are still evolving and outputs may change between releases. Feedback is very welcome — reach out.

What is TwinCell?

TwinCell is a virtual cell for target identification and simulation. It learns signalling pathways from large-scale cell-line data and generalises them to new cell types and tissues, modelling how protein targets propagate their influence onto downstream differentially expressed genes (DEGs).

It supports hypothesis generation and validation:

  • Target validation: Will this drug work on this disease?
  • Simulation: What happens if I stimulate this target? (coming soon)

The API is built to drop into any workflow. **Have a workflow idea? Reach out — we'd love to hear it.

Community & access

We work closely with the research community and are onboarding early collaborators:

  • Free trials to evaluate TwinCell on your own data.
  • Academic credits for non-commercial research.
  • Higher throughput for hypothesis generation and on-premise applications on request.

See Access & Support to get started or request credits.

Get started

  • Quick Start

    Get credentials, install, and run your first target validation in minutes.

    Quick Start

  • Tutorials

    End-to-end notebook: validate Deucravacitinib's target (TYK2) against a psoriasis signature.

    Tutorials

  • API Reference

    Parameter-focused reference for the TwinCell study, preprocessing, validation, and DE.

    API Reference

  • MCP server (coming soon)

    Drive TwinCell from MCP-compatible agents and assistants.

    MCP

Citation & methodology

For full methodological detail, see the TwinCell preprint.

Source & support