Morphodynamics in cells and tissues

 

Can biological shape reveal the hidden states, identities, and developmental trajectories of cells? Living systems exhibit remarkable morphological diversity, from the spatial organization of intracellular proteins to the collective architecture of tissues and embryos (Sem. in Cell. Dev. Bio. 2022). We investigate whether these complex structures contain morphological fingerprints that encode cell states and cell types. Combining advanced imaging, machine learning, and quantitative shape analysis, we aim to identify interpretable morphological features that distinguish cellular identities and capture how cells change during development. By tracking these fingerprints in space and time, we seek to understand when and where cells undergo state transitions and acquire distinct identities.

​

At larger scales, we investigate how the complex shapes of embryos and organoids can be organized into low-dimensional landscapes of morphology. We combine generative machine learning with symmetry-based analysis to identify fundamental modes of shape variation and the symmetry-breaking processes underlying development, such as polarization and elongation. By mapping developmental trajectories onto these landscapes, we aim to reveal the dynamical principles governing how biological forms emerge and transform. Ultimately, we envision using these learned representations to reconstruct virtual developmental trajectories and predict how cells and tissues evolve toward specific morphological states, connecting biological form to function and developmental fate.​