To explore differences in DNA accessibility landscapes between cell types in the PBMC dataset, we identified ATAC-seq peaks open in CD8+ effector T cells relative to CD8+ naive T cells, revealing many regions of open chromatin that were specific to the CD8+ effector T cells. We assessed the rate of false positive results in our differential accessibility (DA) testing by drawing two populations of 100 cells at random from the CD4+ central memory T (TCM) cell population and comparing peak accessibility between these two cell populations, revealing no DA peaks. To further assess how the fraction of cells in the population with a true difference affected our ability to identify DA peaks, we gradually increased the fraction of cells in the comparison group drawn from a separate natural killer (NK) cell population. When drawing >40% of cells from a truly different population, we were able to detect DA peaks between the groups, with the fraction of true DA peaks able to be recovered increasing as a higher fraction of cells in the comparison group originated from a distinct population (Supplementary Fig. 4).
