Galen provides a virtual‑cell platform that uses causal, computational models of cellular biology to predict the outcomes of genetic interventions before they are tested in the lab. By integrating high‑resolution measurements with iterative learning, the system turns cellular states into testable claims, enabling researchers to design more focused experiments and reduce costly trial‑and‑error in therapy development, particularly for engineered T‑cell therapies.
Funding
Funding not disclosed
Founders
Product
Problem
Scientists designing engineered cell therapies, such as CAR‑T cells, must decide which genetic modifications to test, but current biological models are descriptive only and cannot predict the effects of specific gene edits. This uncertainty leads to costly, time‑consuming experiments that often explore low‑yield hypotheses.
Solution
Galen offers a virtual‑cell platform that combines high‑resolution cellular measurements with causal computational models to forecast the outcomes of genetic interventions. By generating testable claims about how a cell will respond to a specific gene edit, the system enables researchers to prioritize the most promising hypotheses before conducting wet‑lab experiments. The platform iteratively refines its predictions using experimental data, maintaining alignment with real biology while reducing the number of unnecessary assays. This computational‑first approach accelerates the design‑build‑test cycle for engineered therapies and improves resource allocation in early‑stage research.
Target Audience
Primary users are researchers and development teams working on engineered cell therapies, including CAR‑T, TCR, and other adoptive immunotherapies, as well as academic labs focused on functional genomics and synthetic biology.
Features
- Integration of single‑cell and multi‑omics datasets to create detailed baseline cellular state representations
- Causal modeling framework that predicts the impact of specific gene perturbations on cell behavior
- Automated claim generation that translates model outputs into measurable experimental hypotheses
- Closed‑loop workflow that updates models with new experimental results to continuously improve prediction accuracy
- Compatibility with existing laboratory data pipelines and standard bioinformatics tools