Handle CTX_INV subtype in SVAnnotate

[epiercehoffman]

Updates

A record with SVTYPE=CTX and CPX_TYPE=CTX_INV was added to a recent GATK-SV VCF after manual curation. The following changes were made to be able to properly annotate this type of event.

• CPX_TYPE will be checked for CTX records, and if it is CTX_INV, the INV interval from CPX_INTERVALS will be added to the annotation segments.
• Additionally, instead of annotating two breakpoint intervals CHROM:POS-END and CHR2:END2-END2+1 for CTX events, we will now annotate 4 individual breakpoints to cover the case where END != POS+1. Those 4 breakpoints are CHROM:POS-POS, CHROM:END-END, CHR2:END2-END2, and CHR2:END2+1-END2+1.
• In the future, to be able to represent intervals on CHR2, POS2 may be added. SVAnnotate will need to be updated accordingly at that time.

Testing

• Unit tests for CTX_INV added
• Unit tests for other CTX updated
• A one-line VCF was created to test the real-life example CTX_INV event that was curated, and it was annotated correctly

[mwalker174]

Thanks @epiercehoffman looks fine to me. As we discussed offline, I'm concerned about special cases like this creeping into the VCF, but it's not he first and this should be a larger discussion. One minor suggestion below.

[mwalker174]

It would be good to either have getComplexAnnotationIntervals() return an ArrayList or wrap this call in new ArrayList(). The reason being that the returned List could be immutable (it's not right now, but possible it could change in the future), in which case the calls to .add() below would throw an error.