Time Machine Learning The earliest stages of life are…

Time Machine Learning

The earliest stages of life are delicate, challenging scientists to come up with new ways observe without disturbing. Here they’ve turned to a form of time travel. Seen from above (top row) or the side (bottom), gentle light sheet microscopy captures individual cells (grey) in a mouse embryo as they grow and develop. Eventually the cells separate into stripes (right) destined to form different tissues. Following these developments backwards in time, (from right to left), a form of machine learning uses multi-coloured blobs to trace each cell’s ancestry, highlighting, for example, that the green-coloured cells on the left share the same lineage as the green cells in the early brain – the neural tube – on the right. Understanding the timings of this separation into layers of cells – a process called gastrulation – will help to improve the picture of early development as well as problems that may lead to miscarriage.

Written by John Ankers

• Image from work by Katie McDole and colleagues


• Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, VA, USA


• Image copyright Elsevier 2018. Reproduced with permission.


• Published in Cell, October 2018

You can also follow BPoD on Instagram, Twitter and Facebook

Archive link