This plot has two columns, one for diseased (plotted in magenta) and one for control (plotted in purple) samples. There are two sets of peaks: Gain peaks (higher in diseased), represented by dark blue, and Loss peaks (higher in control), represented in cyan.
Each column has three plots showing the signal (read density) in windows centered on each binding site and expanding 1500bp up- and down-stream. By default, the read density is computed by taking the average normalized read counts across all the samples in each sample group (diseased and control).
The top plots each contain curves showing the mean read density over all the samples in each group across all the binding sites; there is a curve for signal across the Gain sites (dark blue curve) and for the Loss sites (cyan curve). In the first column (diseased), the dark blue curve is higher than the cyan curve, indicating the higher binding levels in diseased samples, while in the second (control) column, the cyan (Loss) sites show higher signal (with the degree of Loss in the second column being greater than the degree of Gain in the first column).
The remaining two plots in each column show the read density for windows around each binding site. The top plots show signal in and around Gain binding sites (averaged across the samples in the diseased group on the left, in magenta, and the control samples on the right, in purple). The signal is higher in the diseased samples (as expected for Gain sites). The bottom plots show mean read count levels in windows in and around the Loss binding sites (with a greater signal in the right-hand control sites).
The help page for
dba.plotProfile() offer some more details, as does a workbook available here: