Data driven modeling of cell shapes and movements in multicellular systems.

Unraveling how cells move and self organize in a multicellular setting

is crucial for understanding a variety of biological processes like

embryogenesis, tissue formation and many diseases. This is a highly

dynamic process, where cells move and change shape guided by

biochemical as well as physical cues. A precise reconstruction of cell

shape over time provides detailed information on force generation in

a cell and force transmission between neighboring cells. However

extracting 3D cell shapes from microscopy remains a challenge,

especially when cells are irregularly shaped. In this project a novel

approach is proposed that leverages a biophysical model of cell

shape to automatically reconstruct cell shapes from microscopy

images of stained cell membranes. The developing C. elegans

embryo will be used as a model. The reconstructed cell shapes are

subsequently used to quantitatively assess the effect of a genetic

perturbation on cell geometry. Next, cell shapes will be analyzed to

infer biomechanical parameters like cortical tension and cell-cell

adhesion, together with the active forces that drive shape change

and cell movement. Finally, the approach will be extended to offer a

novel framework for data driven mechanical modeling of multi-cellular

movement. As a proof of concept, an explanatory model will be made

of early gastrulation in the C. elegans embryo to quantitatively

evaluate competing hypotheses on force generation.