Ben Steventon (@bensteventon.bsky.social)

A toolkit for mapping cell identities in relation to neighbors reveals conserved patterning of neuromesodermal progenitor populations by Matthew French, Rosa P. Migueles, Alexandra Neaverson, Aishani Chakraborty, Tom Pettini, Benjamin Steventon, Erik Clark, J. Kim Dale, Guillaume Blin, Valerie Wilson, Sally Lowell Patterning of cell fates is central to embryonic development, tissue homeostasis, and disease. Quantitative analysis of patterning reveals the logic by which cell-cell interactions orchestrate changes in cell fate. However, it is challenging to quantify patterning when graded changes in identity occur over complex 4D trajectories, or where different cell states are intermingled. Furthermore, comparing patterns across multiple individual embryos, tissues, or organoids is difficult because these often vary in shape and size. This problem is further exacerbated when comparing patterning between species. Here we present a toolkit of computational approaches to tackle these problems. These strategies are based on measuring properties of each cell in relation to the properties of its neighbors to quantify patterning, and on using embryonic landmarks in order to compare these patterns between embryos. We perform detailed neighbor-analysis of the caudal lateral epiblast of E8.5 mouse embryos, revealing local patterning in emergence of early mesoderm cells that is sensitive to inhibition of Notch activity. We extend this toolkit to compare mouse and chick embryos, revealing conserved 3D patterning of the caudal-lateral epiblast that scales across an order of magnitude difference in size between these two species. We also examine 3D patterning of gene expression boundaries across the length of Drosophila embryos. We present a flexible approach to examine the reproducibility of patterning between individuals, to measure phenotypic changes in patterning after experimental manipulation, and to compare of patterning across different scales and tissue architectures. http://dlvr.it/TM2L2Q