Mapping how chemistry perturbs biology to build a virtual model of the human cell.
We are building AI models of the human cell, trained on our gigascale single-cell maps that measure how drug molecules interact with cells from heterogeneous patients. We are using them to design better drugs, starting from cancer.
Gigascale, perturbative, single-cell maps
A new data modality that enables frontier AI models and drives novel discoveries
Beyond proteins towards virtual cell models
We are building virtual cell models; foundation models that are pre-trained to learn how genes function in diverse biological contexts.
Single-cell data from diverse biological contexts
Text data on the internet enabled the emergence of language models. To bring about virtual cell models, we need perturbational single cell data. We generate this data at a scale that dwarfs other technologies.
A new data modality to design drugs
Our gigascale single-cell maps and the AI models trained on them have discovered novel therapeutics in cancer. We are taking them to the clinic with the hope to treat a large population of patients.
We believe in shooting for the moon and building openly to cure disease.
Meet our founders
A team that has discovered drugs for undraggable targets, trained frontier AI models on chemical and biological data, and invented new methods in single cell and cancer biology
