Fig. 2 | Communications Biology

Fig. 2

From: Automated subset identification and characterization pipeline for multidimensional flow and mass cytometry data clustering and visualization

Fig. 2

Exhaustive Projection Pursuit (EPP) applied to simulated data set. a We simulated two three-dimensional sets of data (Sample A and Sample B). b To identify clusters in these data sets, we applied EPP approach that, recursively, (1) projects the data in a collection of 2D linear projections, (2) characterizes every given 2D projection by a numerical index that indicates the smallest misclassification error across a decision boundary (green line) between identified clusters, (3) separates the data across the top-ranked decision boundary to produce two subsets, (4) repeats on both subsets until no splits are found. Subsets that have no further splits (groups# 1–4) are final clusters identified by the EPP approach

Back to article page