Motivation: The embryonic mouse brain develops in 19 regions, where stem cells change from radial glia to neuroblasts to neurons. Which cell populations develop from which remains unclear: each stage is measured in a different embryo, so the same cells cannot be followed over time, and only a few relationships have been identified previously.

Results: We analyzed over 4 million cells across six stages (E9.5–E15.5), from 14 embryos and 121 brain sections. We developed a pipeline to identify possible connections between cell populations, filter out unlikely ones, and combine different types of evidence to select the strongest candidates. We found that cells develop at different speeds across the brain and justified these differences with anatomical observations. Our recovered all 11 previously reported relationships, and identified 8 new possible developmental trees.

Outlook: Adding more embryos and more information from each cell should improve the detail and reliability of these trees. The framework can then help us understand how brain region and cell position influence development, and identify where and when cell growth or development goes wrong in neurodevelopmental disorders.