Too small values for Peak Count Frequency

[amizeranschi]

How are the y values (Peak Count Frequency) computed for the various peak profile plots? The shape of the curves looks like what I would expect for my samples, but the y values seem much too small and this seems quite confusing.

Are these values normalized somehow? How should they be interpreted?

For example, for a y value of 0.0004, does this mean that only 0.04 from the genes in the genome had peaks at their TSS?