In this talk I will discuss the role that evolution of the structural genome plays in tumor heterogeneity and drug resistance.  In particular I will focus on whole genome doubling (WGD) as a mutational process that is ongoing throughout cancer evolution. We found through novel approaches combining single cell whole genome sequencing and tumor phylogenetic modeling that WGD is pervasive in ovarian cancer and that nearly all tumors exist on a distribution of mixed populations of diploid and WGD cells. This implies that WGD is continually active as a mutational process in ovarian cancer. In comparing WGD cells to diploid cells, we find that WGD cells have higher rates of chromosomal instability and overall higher degrees of genomic diversification.  Moreover cell-specific WGD has phenotypic consequences that impact proliferation and immunosuppression.  In the  second part of the talk I will discuss how genomic diversification in ovarian cancer can be exploited through studying evolution in patients through timeseries sampling of cell free DNA before, during and after therapy over a period of several years.  We found through modeling clone-specific structural variants that diverse cellular populations are often pruned into simpler compositions in relapsed disease. Moreover, clones that were  more 'fit' by population genetics modeling often harbored clone-specific high level amplifications of oncogenes, thereby revealing previously unappreciated drivers of drug resistance.  Our work therefore demonstrates the discovery potential of single cell whole genome sequencing and the role of structural genome evolution as a determinant of therapeutic response in ovarian cancer.