Hi,

I am having a similar issue to #37 with the purity/ploidy estimation for one of my samples. I am running wakhan as part of lumos, using this container: docker://gokcekeskus/wakhan:55692a6

The purity/ploidy heatmap and optimised peak look like this:

MM17347_optimized_peak.html
MM17347_heatmap_ploidy_purity.html

While the copy number output files report a ploidy of 7.81, purity of 1:

MM17347_7.81_1.0_0.92_genome_copynumbers_breakpoints.html.zip

This is the coverage/variation plot:
CN_VARIATION_INDEX.html

This sample has very high purity, so I tried limiting the purity range to 0.8-1.0, but I didn't see any change in the heatmap (the copy number output was also unaltered). The ploidy range I had not altered from the default, but I believe a ploidy solution of 7.81 is out of this range.

When ran on other samples, the heatmap covers a much larger range of values, do you know why it is so limited in this case? Secondly, does the specified purity range affect the purity/ploidy estimation step or just the copy number calling?

Finally, how does wakhan determine whether a change in copy number is subclonal? For this sample I believe that coverage pattern is better explained by some subclonal variation rather than with a ploidy of 7.81 but would be interested to know how this works.

I am assuming the container in the lumos pipeline is up to date with v0.2.0 but if not please let me know and I will rerun.

Many thanks for your help!