Research
Our mission is to understand life by simulating it, from individual cells to complex organisms, and by simulating it we hope to improve it.
Animating tissue
Cell dynamics is a product of its environment. Learning how cells behave in their native tissue context is crucial for identifying effective therapeutic targets
Designing cells
Cells reprogrammed to treat diseases display highly variable behavior inside patients. We need to engineer more consistent behavior in inconsistent environments.
Interpreting models
We mostly treat AI models as a black box: something goes in and a response comes out. This is not acceptable for models that will be used to treat human diseases.
Progress
Below are selected publications describing our research progress.
Building Foundation Models to Characterize Cellular Interactions via Geometric Self-Supervised Learning on Spatial Genomics
A foundation model for cell interactions
Identifying perturbations that boost T-cell infiltration into tumors via counterfactual learning of their spatial proteomic profiles
Find drug targets from tumor images
Localization of signaling receptors maximizes cellular information acquisition in spatially structured natural environments
How cells navigate complex environments
Inferring gene regulation dynamics from static snapshots of gene expression variability
Learn dynamical systems from non-dynamical data