Clinical-Genomics/genmod

Compound is missed if one parent is genotyped for each variant

Opened this issue · 5 comments

bartcharbon commented

We have a probably rather rare situation:
After quality checks we end up with two variant in one gene:

  • variant one is called (hetrozygote) for a child and its mother(hetrozygote), but has a "./." for the father
  • variant two is called (hetrozygote) for a child and its father(hetrozygote), but has a "./." for the mother

This is possibly a comp_het, however it is missed by Genmod.

I changed the missing parent GT's in a testfile to 0/0 to confirm it works as expected in that case, and indeed: Genmod then marks the variants a compounds.

moonso commented

Ok I will create a test case to replicate the situation

dennishendriksen commented

@moonso we have been running into this issue a couple of times. are you still planning to look into this issue?

moonso commented

Hi, sorry for taking so long. I will try to find time and look into this as soon as possible. Get back when there is a time estimate

dennishendriksen commented

No problem. I'm not proficient in Python (yet), but if there is anything I can do to help please let me know :)

dnil commented

A naive replication in the example files supports this. Pinning it as a bug, and prioritising it there after! 😊