Correlation of cell type–specific enhancers to gene expression and potential regulators of T-cell subsets. (A) Bubble plot representations of CAGE TSS activity around eRA+Treg vs eRA+Tconv enhancer candidate regions showing at least twofold different H3K27ac or H3K4me1 signals. The bubble plots encode 3 quantitative parameters per CAGE cluster: distance from the putative enhancer, log10 fold change in CAGE cluster tag count between eRA+Treg and eRA+Tconv (y-axis), and the absolute CAGE cluster tag count of the T-cell subset with the highest expression level (bubble diameter). There is a clear bias for the putative enhancer elements to associate with CAGE clusters upregulated in the corresponding cell type (P < .001, Wilcoxon signed rank test). (B) A nonredundant combined set of de novo motifs from CAGE clusters and candidate enhancers in eRA+Treg vs eRA+Tconv. Shown are motifs with high similarity to known TF motifs. Absolute tag counts (log2 transformed) of differentially expressed TFs matching a de novo motif are presented as colored boxes with yellow, blue, and red representing low, intermediate, and high expression, respectively. (C) STRING-based networks (confidence views) of the top TFs predicted to have the strongest regulatory input in human Treg subsets compared with their Tconv counterparts (additional networks are shown in supplemental Figure 8). Only connected edges are included. Complete TF lists are provided in supplemental Table 5. Direct interaction partners or targets of FOXP3 are marked as indicated.