Dynamic Morphoskeleton

During embryonic development, cells undergo large-scale motion generating tissues rearrangement, which ultimately defines the final shape of the embryo. While developmental biology has identified several genes driving local cellular processes, the interplay between cell-intrinsic and external stresses is fairly less understood because several local mechanisms are still unknown or hard to measure. By contrast, with the significant advances in live imaging techniques, it is now possible to fully track cell trajectories. Using ideas from nonlinear dynamics, we propose a rigorous objective kinematic framework for analyzing cell motion, which uncovers the underlying dynamic morphoskeleton, i.e. the centerpieces of cell trajectory patterns in space and time. The dynamic morhposkeleton provides a quantitative tool for comparing different morphogenetic phenotypes, quantifying the impact of genetic and physical manipulations, studying cell fate, and overall bridging the gap between bottom-up and top-down modeling approaches. We illustrate our results on a Drosophila gastrulation dataset obtained by in toto light-sheet microscopy.