Ludwig, L. S. et al. Cell 176, 1325–1339 (2019).

We all start life as a zygote, and if we could follow each cell’s divisions and fate decisions, we would understand a lot more about the development of complex tissues in vertebrates. One option is to genetically tag cells early in development, but this is not feasible in humans. Ludwig et al. instead use sequence variations in the transcriptome of mitochondrial DNA as natural barcodes to infer clonal relationships and draw lineage trees. In combination with single-cell ATAC-seq data, the researchers can use this approach to correlate clonal dynamics and gene expression with chromatin accessibility. These natural barcodes allowed them to define the subclonal structure in colorectal tumors and to follow clones of chronic myelogenous leukemia in patients from the time of diagnosis to the sixth month of therapy. The approach holds promise for elucidating the clonal evolution of cancer and the development of drug resistance.