Download .vcf.cgh i balsamic

[AleksandraKrstic]

Hi,
We would appreciate if we could get the ikon to download .vcf.cgh files for all the samples analysed with balsamic in cust087.

Regards, Aleksandra

[dnil]

Hi Alex! Afaik Scout would show the icons, if the files existed. I had a quick look at a few cust087 cases on the file system, but couldn't find any cgh files to add links for. I think what you want is to include vcf2cytosure in Balsamic, or alternatively rerun your wgs samples with MIP where it is enabled by default.

See issue Clinical-Genomics/BALSAMIC#605 in Balsamic, and note your support for the idea! 😄

[dnil]

We are also somewhat close to start using GENS (already available in scout, nearly done in MIP) for the RD samples. This might be of interest for your hematology samples as well? https://github.com/Clinical-Genomics-Lund/gens

[AleksandraKrstic]

Hi Daniel, Yes we want to include vcf2cytosure in Balsamic, as our samples are panel seq not wgs. Is that possible?

[dnil]

For wgs, it should be rather straight forward, though subject to prioritisation by the Balsamic developers.

For panels, it would be technically possible, but how useful it is depends on your panel. Do you have sufficient "backbone" coverage from over the chromosomes to call SVs and/or ploidy?

If I'm missing the point, what exactly is the result or readout you would be looking for, and what in your data would allow you to analyse it?

[dnil]

I had a quick look at cust087. It seems the panel is a small one (like a Mb) for deep sequencing? Perhaps what you would like to do is avoid having to run a microarray? A suggestion would then be to run a shallow depth wgs in parallel with the panel. But this is more of a project discussion. Perhaps we can set up a small work group at the clinic for it? I think I will close the Scout issue for now. Please fill in on the Balsamic issue if I missed something and you believe vcf2cytosure would produce something useful on the data you already have!

[vwirta]

Quick comment to this. There is a backbone with approx 2 Mb spacing in the myeloid panel. We are going to make this more dense (1 Mb) in future. Will be the same backbone as in the lymphoid panel.

CNVkit is being run on all these samples and that allows picking up some aberrations.
Let us know what analysis you would like to do and we can see how that can be solved. I'll add @khurrammaqbool and @ashwini06 here to make them aware of this discussion.

[dnil]

Quick fix from @J35P312! It should now be possible to use the coverage estimate from CNVkit to render a .cgh with vcf2cytosure v0.6.1. https://github.com/NBISweden/vcf2cytosure

[J35P312]

Yep, vcf2cytosure works well with cnvkit output!
I run it like this:

vcf2cytosure --vcf CNV.somatic.setperch.cnvkit.vcf.gz --cn 740694/F0041870/tumor.merged.cnr --bins 1

where --vcf is the cnvkit annotated vcf, and cnr is cnvkit cnr file

good luck!

[ashwini06]

@northwestwitch @dnil : Now BALSAMIC can generate '.CGH' files for panel cases. Fixed in PR. It will be available in production from the next release version. However, there is no '.CGH' download icon for panels as in WGS. Can you please look into it
@ivadym : Can you please check if the required file .cnvkit.vcf2cytosure.cgh is being stored in HK and available for Scout uploads

[dnil]

Super! As long as the file name is passed on the key "vcf2cytosure" in the case load config as with the RD cases, they are supposed to show in Scout without any further changes! But many things that are supposed to be in theory actually turn out not to be upon encounter with practical trials. If you have a demo case on stage or so we can help you check! 😸

[ivadym]

I just checked and it seems to only support MIP workflow, but I can make the respective changes in CG. Thank you @dnil.