Abstract
Developmental biology studies embryos develoment by studying model organisms. Ascidians, marine animals closely related to vertebrates like humans, are an appealing model as homologous cells can be identified across embryos allowing to study the development at a cellular level [Conklin, 1905]. When looking at an embryo population, it can be observed that all embryos maintain roughly the same cell number through time [Malandain, 2024], suggesting that the cell division rate is conserved across embryos, while the cellular composition differs between embryos, indicating cell division heterochro- nies.Here, we investigate the mechanisms that may lead to this conservation of cell number during the em- bryo development by the means of simulation. More precisely, we have tested different hypotheses and compared the simulation results to measures issued from an observed population.The material from [Guignard, 2020] includes seven 3D time-series images of Phallusia mammillata embryos with segmented and tracked cells, providing quantitative information as cell division times, cell lifetimes, etc. We initially simulated the evolution of cell numbers in the ascidian embryo, assuming cell division time as an independent random variable. While this approach accurately reproduced the average cell count over time compared to the ground truth population, it failed to capture the heteroge- neity in cell numbers over time. Then we simulated the cells lifetimes as random variables which also did not reproduce exactly the heterogenity in the ground truth population. To determine whether this is due to insufficient data, we simulated populations of simulated embryos in which we can control the number of individuals to evaluate how population reproduction varies with reference sample size. Our results indicated that we need more embryos acquisitions to conclude whether this heterogeneity not well reproduced because of the small number of individuals in the dataset or be- cause of a missing assumption not considered in the simulations.