A genomic systems biologist muses on how shared DNA mistakes reveal shared cellular ancestry.

The way an organism or cancer develops often depends on how cells relate to each other. This is because ancestral cells make molecular decisions that affect the regulation of genes in their offspring. Revealing complex lineage relationships usually requires sophisticated methods of molecular embryology that are available for only a few 'model' species and not for humans. I recently stumbled upon a string of creative papers from two competing labs that prepare the ground for change.

When cells double their genomes, they make very rare mistakes that are passed on to their daughter cells, such that two closely related cells share more identical DNA mistakes than they each do with a third, less closely related cell. If one could quickly and accurately sequence the genomic DNA of single cells, phylogenetic algorithms could reconstruct the underlying lineage relationships by analysing the patterns of shared mistakes in the various cells.

Marshall Horwitz at the University of Washington in Seattle and his colleagues have accomplished this feat by cataloguing mutations in 300 cells from a single mouse (S. J. Salipante et al. Evol. Dev. 12, 84–94; 2010). Ehud Shapiro at the Weizmann Institute of Science in Rehovot, Israel, and his co-workers had previously conceived this strategy but focused on different cell types (A. Wasserstrom et al. PLoS ONE 3, e1939; 2008).

With these approaches, along with new, faster and cheaper high-throughput sequencing methods, we will one day be able to pick apart any group of cells from any interesting organism and establish full lineage trees. This would revolutionize not just comparative embryology and the study of evolution, but also medicine. Tumours could be dissected in this way and drugs designed against their specific lineage compositions, as part of truly personalized genomic therapies. Who would have predicted that human ingenuity could turn our common genomic rubble into such intellectual gems?

View the archive at http://blogs.nature.com/nature/journalclub