Saturation curve in duplication module

[s-andrews]

We've had a few comments about the presentation of the duplication information. One issue which was raised was that it's hard to compare total duplication levels between samples of different sizes.

One fix for this could be constructing a sequencing saturation curve. This would use the same code we currently use to calculate total duplication, but running it every million sequences and drawing out the curve. This would give way points which would be more comparable between samples and might be an easier read-out of the overall saturation of the library.