CN from log2 is inaccurate at times

[ashbig]

In brief, there seems to be issues calling copy number from log2. This is from a no template control run in our lab. The depths and log2 are both accurate, but the CN call shows a gain of 8. I've also observed CN being off by a few copies than what is expected from the log2 and I can dig up more examples if helpful.

code for 0.9.6

cnvkit.py call ${cns} -o ${segs} cnvkit.py genemetrics ${segs} --stdev --ci -o ${s_gainloss}

gene chromosome start end log2 cn depth weight n_bins DDR2 chr1 162688843 162750046 -4.06365 8 1.49628 10.3477 16

[tetedange13]

Hi @ashbig ,

Not an author of CNVkit, but would it be possible for you to upgrade to latest release of CNVkit? => A lot of things were fixed/improved since CNVkit v0.9.6 => So if you can, tell us if you still got same inconsistant CN values with CNVkit v0.9.9

Have nice day. Kind regards. Felix.

[tskir]

Hi @ashbig, I was wondering if you had the chance to try this with CNVkit v0.9.9?

[ashbig]

@tskir @tetedange13

I did upgrade and test in CNVkit v0.9.9 and I no longer see the inconsistency. However in upgrading I re-read the documentation and observed that my call command was executed incorrectly. The gene-metrics command also requires the .cnr file which I was failing to provide.