Data set preprocessing and alignment
We align single-cell data from two different experimental conditions and recover shared subpopulations, thus removing batch effects. We use the data from Kang et al. (2018), where authors stimulate the peripheral blood mononuclear cells (PMBCs) with interferon-alpha and compare them to a control sample. Dataset alignment reduces treatment-specific variation that may occlude the differences among cell subpopulations shared between the two data sets.