The study by Navin et al. demonstrates the use of single nucleus sequencing (SNS) to investigate tumor population structure and evolution in breast cancer cases. By flow-sorting nuclei, whole genome amplification (WGA), and next-generation sequencing, the authors accurately quantify genomic copy number within individual nuclei. They apply SNS to two breast cancer cases: a polygenomic tumor (T10) and a monogenomic primary tumor (T16P) with its liver metastasis (T16M). In T10, 100 single cells revealed three distinct clonal subpopulations likely representing sequential clonal expansions. In T16P, 100 single cells showed a single clonal expansion forming the primary tumor and seeding the metastasis. Both primary tumors also contained an unexpectedly abundant subpopulation of genetically diverse 'pseudodiploid' cells that did not travel to the metastatic site. The data indicate that tumors grow through punctuated clonal expansions with few persistent intermediates, contrasting gradual models of tumor progression. The study highlights the utility of SNS in identifying previously undetectable cell types and inferring tumor evolution.The study by Navin et al. demonstrates the use of single nucleus sequencing (SNS) to investigate tumor population structure and evolution in breast cancer cases. By flow-sorting nuclei, whole genome amplification (WGA), and next-generation sequencing, the authors accurately quantify genomic copy number within individual nuclei. They apply SNS to two breast cancer cases: a polygenomic tumor (T10) and a monogenomic primary tumor (T16P) with its liver metastasis (T16M). In T10, 100 single cells revealed three distinct clonal subpopulations likely representing sequential clonal expansions. In T16P, 100 single cells showed a single clonal expansion forming the primary tumor and seeding the metastasis. Both primary tumors also contained an unexpectedly abundant subpopulation of genetically diverse 'pseudodiploid' cells that did not travel to the metastatic site. The data indicate that tumors grow through punctuated clonal expansions with few persistent intermediates, contrasting gradual models of tumor progression. The study highlights the utility of SNS in identifying previously undetectable cell types and inferring tumor evolution.