Fig. 2
Multifactorial ChIP-seq data integration. a Venn diagram depicting the overlap of ERα with steroid hormone receptors (left) and overlap of ERα with FOXA1 and GATA3 (right). b Unsupervised clustering correlation matrix of bound regions among all factors. Top bar indicates the factors: ERα (blue), AR (red), PR (purple), GR (black), FOXA1 (light green), and GATA3 (dark green/orange border). Pearson correlation is depicted from 0 (white) to 1 (dark green). Sites used to construct the matrix are the union of all consensus binding sites of the factors. c Individual Venn diagrams depicting the pairwise overlap of ERα with each factor. Consensus binding sites for each factor were used (Fig. 1d), and overlap with ERα sites from the same tumors was assessed, taking the same threshold. d Heatmaps indicating the binding peak intensity for each combination in (c). In each panel, ERα (left) and factor (e.g., FOXA1, right) ChIP-seq data are depicted. Shown are peaks selectively observed for ERα (top), selectively observed for other factors (bottom), or shared between them (middle). e Correlation heatmap for ChIP-seq data sets between ERα and the other factors. Colors in top bar indicate patient (top) and factor (bottom)
