I will discuss our efforts to develop and apply single cell genome sequencing methods to study breast cancer evolution in the context of invasion in premalignancies and chemoresistance in triple-negative breast cancer patients, as well as our efforts to build an atlas of cell types in normal breast tissues. To study invasion in 10 patients with ductal carcinomas in situ (DCIS) premalignancies we developed a Topographic Single Cell Sequencing (TSCS) method that has spatial resolution to measure genomic copy number profiles in patients with synchronous DCIS-IDC tissues. Our data revealed a multi-clonal invasion model, in which genome evolution occurred within the ducts leading to the generation of multiple clones that co-migrated across the basement membrane to establish the invasive carcinomas. By combining scDNA-seq and scRNA-seq to study longitudinal samples from 20 triple-negative breast cancer patients, we identified a model of chemoresistance in which the adaptive selection of resistant genotypes was followed by additional transcriptional reprograming to achieve a fully resistant state. These data implicated gene signatures including EMT, hypoxia, angiogenesis, CDH1 signaling and AKT1 signaling in chemoresistance. Our group has also led efforts to establish a human breast atlas of normal tissues using single cell RNA sequencing of 20 women. These data identified major cell types including epithelial cells (luminal HR+, luminal secretory, basal), fibroblasts, adipocytes and endothelial cells (vascular, lymphatic) as well as immune cells (lymphocytes, myeloid). Additionally, we have identified minor cell types include pericytes, apocrine cells, mast cells and others. For each cell type, we identified 2-5 cell expression states that correspond to different biological functions that have been validated using spatial transcriptomics. Collectively these data provide a ‘human genome’ reference of normal cell types and cell states in the human breast tissues.