I am very interested in using your CopywriteR technique to analyze gains and losses from exome data.  I went through the entire pipeline and have a very messy output.

It does not look nearly as clean as the examples in the copywriter paper.  Am I supposed to interpret the green line at the top and red line at the bottom as gains and losses, or am I only supposed to interpret the red lines in the middle of the image as indicating gains and losses.  Does this look reasonable as output or is there a way to improve it?

My plot does not look nearly as clean as what you show in your publication:

Any suggestions?  Thanks so much for your help.

Thomas Kuilman's response:

You seem to suffer from a limited number of off-target reads (your MAD-value is quite high, your profile is rather vague and stripy (this comes from low count bins, which give rise to only a limited number of actual log2-values)). There are two things you could try: either you could reanalyse with increased bin size (you get more counts per bin giving better results (at the expense of resolution of course)) or you can try and see whether there is other reasons why you could have low off-target numbers (please refer to here https://github.com/PeeperLab/CopywriteR#number-of-off-target-reads). If you send my your CopywriteR.log file and tell me what type of data this is (WES or smaller panels) I could give you an indication of whether your off-target read numbers are in the expected range.

With regards to the plots, the red and green triangles at -2 and +5 indicate that data points are outside of the plotting area. The red lines in the middle are the segmentation values, just like the ones in the example we provided. I hope that makes things a little clearer.

Thanks for posting this one on the Bioconductor support site! Thomas