Figure 1

The Regulatory Build process. In a first step we run segmentation software across multiple cell types. For each cell type and at each base pair, the genome is assigned a state, identified by an arbitrary number assigned by the segmentation software. We assign to each state a non-unique functional label, represented by its color on the browser, as shown at the top. For each state at each base pair, we compute the number of cell types sharing that state at that position, as shown in the center of the figure. Having selected relevant states and set some thresholds, we define regions of interest, which are the foundation of the regulatory build. These regions are then complemented with unannotated ChIP-Seq transcription factor binding site peaks and unannotated DNase1 hypersensitivity sites.