Skip to main content
. 2019 Aug 28;20:180. doi: 10.1186/s13059-019-1784-2

Fig. 1.

Fig. 1

Schematic of annotation pipeline. All available data for a given cell type is input to Segway, which produces an annotation with integer state labels. A machine learning classifier then assigns an interpretation to each state, using features derived from all segments with that state in the genome