Depths at loci where one or both BAMs have zero coverage

[ryan-williams]

I'm imagining the left half of the x-axis being BAM 1's coverage at loci where BAM 2 has 0 coverage, and vice versa for the right half, the y-axis in both cases being the number of loci with coverage (0,x) or (x,0).

As a special case, x=0 should be the number of loci where neither had any coverage; that bit is probably contingent on #2, depending on a reference track being incorporated. v1 of the plot that is undefined at x=0 would still be useful.