Is it possible to infer duplications/deletions/translocations/insertions/inversions from SvABA output?

[Giuseppe1995]

Hi, I have a series of SvABA vcfs, and I wanted to know if it was possible to tell somehow the type of a variant, since the field "TYPE" has only one observation=BND. I am sure that this is something easily doable, but I cannot figure out how.
Thanks,

Giuseppe

[walaj]

It depends. If this is for cancer genomes then in general the answer is no, since this is a very complex problem that has only recently been partially solved. For germline, where the variants are smaller and not as complex, you just have to infer it from the BND annotation per VCF 4.2. See issue #4 for more details. I've not been excited about maintaining that conversion tool since the "del/dup" etc names are only based on relative orientations and not on copy-number calls (as is the case in DELLY, etc), and so calling a <-> a duplication is only sometimes true whereas specifying separate breakend annotations is always true.

[Giuseppe1995]

Ok, thanks, I got it... Since I have also copy number data (I'm working on tumor-normal matched samples) I can try to integrate them with the output from SvABA . Thank you!