Seqr SV's male X chromosome incorrectly displayed

[cvlvxi]

Hi team,
We have an inquiry about how X chromosome male SV variant’s genotypes are being displayed in seqr. We have noticed that unlike in SNVs where an X chromosome male hemizygous variant is displayed as G/-, in SVs a hemizygous variant can be displayed as either ref/alt or alt/alt.

As you can see in the screenshots below there are three different genotypes displayed for X chromosome male SV variants (ref/ref, ref/alt and alt/alt). As there is only one X chromosome in a male, and therefore variants are hemizygous, there should really only be ref/- or alt/- and so we are wondering if there was a decision made for SVs to be displayed as ref/alt for a hemizygous state.

As far as we can tell the ref/alt is acting as a hemizygous variant in seqr and appearing in recessive searches correctly, so it’s not a major performance issue, we are just wondering why they are being displayed this way and whether there was a conscious decision to display them differently to SNVs.

Any information about why they are being displayed differently to SNVs would be much appreciated.
Thanks

[hanars]

Hello, unfortunately Structural Variants are not supported in seqr for local installations at this time. There is therefore no support or documentation available, and trying to load and use this data is not something we can guarantee will work at all or be accurate

[bellmartin]

Thanks for your response.

We do see this behaviour of hemizygous SV's being displayed as het's in the broad seqr instance.
We are interested to know if this was a conscious decision to display them differently to SNVs or not ?

We may work on changing how the hemizygous SV variants are displayed in our seqr instance , but before we do that, wanted to check if there was a reason behind this display choice that is not immediately obvious ?

Thanks for your help

[hanars]

We are still developing our SV loading and search pipelines and have not made any final decisions on how SVs will behave in seqr and if and how they will be different from other variant types. The lack of clarity around expected behavior is one of the reasons we have not released this functionality yet