Humans are complex organisms, comprised of many tissues, which are comprised of different cell types. Single cell sequencing technologies now allow scientist to interrogate epigenetic, genetic and transcriptional profiles of single cells, but have also challenged the definition of “what is a cell type?” In this thesis, I define cell types as cells with unique transcriptional signatures, and go on to characterize these cells in normal and tumor tissues.

First, I optimize an analysis workflow for single nuclear RNA-seq. To assess this method, I analyze deeply characterized human and mouse brains, and move on to apply this method to a less characterized tissue—human heart ventricles. I describe the ability of this method to capture rare and meaningful biological cell sub-types, and describe differences between human right and left ventricles at the single cell level.

Next I use this method to analyze flash frozen tumor sections from five glioblastoma patients. This data is analyzed in conjunction with single cell open chromatin profiling data (snATAC-seq). By combining these two single cell technologies, we can first confidently identify cells carrying copy number variation in tumor samples, ask which of these copy number changes are represented at the transcriptomic level, and ultimately characterize the inter and intra patient heterogeneity that we observe.