Entering edit mode
5.2 years ago
star
▴
350
I have multiple VCF files (they are not multiple-sample VCF) and I just looked the genotypes and I saw there are 4 genotypes for each variant like: 0/0, 0/1, 1/1, ./. . The last three genotypes are clear to me, but I don`t understand why we have 'Hom_ref genotype' in the VCF files? And shall I ignore them?
Was this VCF generated from just a single sample or was it created by a multi-sample VCF being split? If it's the latter, the presence of
0/0
has a more obvious reason. If it's the former, the variant calling algorithm needs to be looked into to understand the criteria for a hom-ref call to be made.I guess it is only one sample.
here is the header of VCF:
I provided an answer, here: C: Why is 0/0 homozygous reference call in VCF useful information?
Thanks but my VCF is a single sample VCF file.
How does one account for single sample hom-ref calls though?
The occurrence of 0/0 in a single sample VCF may be caused by low-quality mapping.
Can you expand on this a little? I'm curious how low quality mapping can result in hom-ref calls.
Actually, I came across a freebayes issue with Homo ref call, forgot to reference it. https://github.com/ekg/freebayes/issues/517
That issue does not reach the conclusion that a
0/0
is caused by low quality mapping. Low quality would result in./.
if anything, IMO.We don`t get ./. genotype, if there is low Q mapping?