Big data approach towards the characterization of normal peripheral immune cells with data from ImmPort (TECH1P.860)

To date, our understanding of a normal immune system is far behind that of other healthy organ systems. One reason for this is the lack of a widespread adoption of standardization in the lab techniques, especially flow cytometry. Here we demonstrate the use of the Immunology Database and Analysis Portal (immport.niaid.nih.gov) as a source of publicly available flow cytometry files from hundreds of participants in several clinical trials, to study immune cells from the blood of healthy individuals. To characterize well-defined cells in a normal immune system, we used an unbiased, gate-free method to compare data from different cytometers and antibody staining panels, based on the distribution of fluorescence intensity of the markers. In order to perform the meta-analyses, the marker information from each file was obtained in an automated fashion and their nomenclature was made consistent. Various transformation and normalization strategies were applied and evaluated to make the flow cytometry data comparable across studies. Initial promising results were observed for the cell-surface markers used to define T and B cells in general, and some specific subsets including regulatory T cells as well as plasmablasts. We found high concordance in normal individuals from two studies performed at different times measured within the same facility. The extension of this approach to studies of healthy individuals across facilities promises to give insights towards the normal immune system.