Fig. 3
From: A comparison framework and guideline of clustering methods for mass cytometry data
Number of clusters detected by each tool. a, b Number of clusters detected by default settings of Accense, DEPECHE, Xshift, PhenoGraph, and flowMeans. Translucent points represent results in 20,000 cell subsampling tests. c, d Number of clusters (c) and precision (d) of highlighted tools and settings were impacted by sample size in the Levine32dim dataset. Dotted lines represent performance of alternative settings of different tools (Xshift: Elbow Plot Determination; FlowSOM: automatic estimation of number of clusters). Since the precision of the default settings of Xshift and FlowSOM have been shown in Fig. 2, they are not presented here