Fig. 2
Enhancer identification methods vary in the number and length of predicted enhancers. (a) The number of K562 and liver enhancers identified by each method varies over two orders of magnitude. There is considerable variation even among methods defined based on similar input data, e.g., histone modifications. (b) The length of K562 and liver enhancers identified by different methods shows similar variation. Enhancer lengths are plotted on a log10 scale on the y-axis. Data for other contexts are available in Table 1 and Additional file 1: Figure S1