Feature Request - support CNV calls (and ideally SV calls)

[adbeggs]

HI all

Self explanatory hopefully. We have got a germline adaptive sampling panel working on Nanopore, and are generating copy number variants and SVs (including fusions) - do you think there is any chance of support for CNA and SV. Ideally from VCF (compliant ones obviously) but I can lift over calls if easier?

A

[sigven]

Hi Andrew,

Thanks for reaching out. I am very much interested in looking into this, as I think you point to a gap among available tools. Although your request is somewhat self-explanatory, I would definitely need some example input files to get going, and also some thoughts on the minimum functionality you'd like to see wrt annotation of these variant types. I think VCFs would be the best.

Furthermore, some ideas on what type of information from the Nanopore calling (pulled from the VCF) that would be useful to show in the final interpretation report (given that the call properties related to SVs/CNAs are different from SNVs/InDels), would be valuable.

kind regards,
Sigve

[sigven]

BTW: here is an example report preview of the forthcoming release, including reference data updates etc:

https://www.dropbox.com/scl/fi/93j6o03megarubkrnglio/IMPR-T006_GL_0.cpsr.grch38.html?rlkey=rw7dif1vqqoe0l2hn6za5yy3m&dl=0

[sigven]

Hi Andrew,
Quick follow-up on this: you have a dummy VCF with CNVs/SVs I can play with? Realize it may be difficult with patient-sensitive stuff, but you might have your hands on non-sensitive stuff too?

best,
Sigve

[adbeggs]

Hi Sigve
I have got a NA12878 genome sequencer somewhere so let me run the pipeline on it and I can attached the SVs, CNVs etc on it. Will take a few hours

[adbeggs]

GM12878.CNV.vcf.gz
GM12878.SV.vcf.gz

Hi Sigve

Try these two. CNV is called with Spectre2 and SV with Sniffles2 but they are both VCF format compliant