How to find genes with altered CNV from CNVkit output?

Question

[ahmad mousavi]

Hi

I have performed the CNVkit on WGS deduplicated bam files ( normal and tumor) using the "batch" command, I have two questions :

1- How can I find the genes with CNV ? ( I mean on which files the information could be fetch?)

2- Is the batch command for such CNV analysis ( normal bam file and tumor bam file) ?

thanks.

Answer 1

[DGTool]

CNVkit can give out various tables/reports: https://cnvkit.readthedocs.io/en/stable/reports.html. For seeing which genes have CNV's and their copy-number value, you can use CNVkit's genemetrics command (https://cnvkit.readthedocs.io/en/stable/reports.html#genemetrics) which outputs a list of genes who's log2 threshold has passed a certain value (by default this is 0.2; from the page: "A threshold of .2 (the default) will report single-copy gains and losses in a completely pure tumor sample (or germline CNVs)"). If you want copy-numbers in the output, make sure cnvkit.py call has already been performed on the necessary files, otherwise only log2 (and some other columns) will be reported.

[ahmad mousavi]

Thanks DGTool for answer. For running call you mean I need to run the call part on .cnn/.cns files again? Becuase the batch run "call" as a part of pipeline.

[DGTool]

Ah, if it already has run call, then that should be fine, you can just run genemetrics on it and it should give you the copy-number in the output. (To check if call had run successfully on the cns/cnr files, the .cnr and/or .cns file should have a cn column in them; if yes then all good, otherwise one can run call on them separately)

[ahmad mousavi]

Thanks a lot, How can I find the CN-LOH from those files?