Fig. 6

SIC pipeline applied to human peripheral blood flow-cytometry data. The results of multidimensional QFMatch alignment between user-guided and fully automated clustering outcomes for one of the samples (~200k live singlets). The following sets of measured parameters were used for user-guided clustering, EPP clustering, cluster matching and data visualization of pregated live singlets: Side Scatter, Dump (CD3, CD14, CD16), CD19, CD20, CD38, CD27, IgM, IgD. Unmatched cell subsets are indicated as red squares. These unmatched subsets are cell populations that were not identified by the user in the manual gating strategy. User’s gating strategy was not exhaustive, i.e., it did not aim to identify all of the subsets present in the sample, and was limited to identification of the cell populations listed on the left panel. In contrast, EPP is an exhaustive subset identification technique, i.e., all of the subsets present in the sample were identified. These unmatched subsets are cell subsets that were not identified by the user in the conventional gating strategy, but they can now be readily explored looking at: a the expression level in each channel via pathfinder tool (see Supplementary Fig. 7); b the gating strategy that EPP built (see Supplementary Fig. 11b); c backgating with the highlighter tool (see Supplementary Fig. 7). This toolkit (a–c) was designed to interpret the fully automated clustering outcomes and assign cell subset names to identified clusters. Also, this toolkit can help reveal the presence of a false cluster created by the EPP approach. Essentially, this is a strategy that can be applied to identify and characterize new cell subsets using the SIC pipeline