Need to support complex rearrangements in SV (VCF)

[yfarjoun]

The VCF parser is currently unable to parse VCFs with complex rearragnements and this is showing up as Picard issues: https://github.com/broadinstitute/picard/issues/1406

Need to prioritize and fix.

[cwhelan]

Just wanted to alert whoever looks at this that @vruano started to work on this and ended up wanting to refactor the Allele class. A lot of that work is in this large PR:

https://github.com/samtools/htsjdk/pull/1370

There may be a shorter-term fix to prevent errors, but this was his attempt to restructure the allele class so that it could support https://github.com/samtools/htsjdk/issues/18

[jrobinso]

@yfarjoun I also need to support this in IGV, I have no idea what it will do with a complex SV but probably nothing useful. I see this is still open here, when fixed could you provide me with your test cases (VCFs)?

[vruano]

@cwhelan is there still and interest in this?

[cwhelan]

This is definitely something we still need to implement at some point.

[vruano]

In this case I can undust the old large PR code with respect SV allele support and put you as a reviewer, @cwhelan. Do you have some use cases in gatk or gatk-sv that have their own custom Java solutions?