Difference between . and ./. for missing genotype in VCF

Question

[William]

What is the difference between . and ./. for a missing genotype in a VCF file? For example in one VCF record I have these two sample genotypes.

GT:AD:DP:GQ:MMQ:PGT:PID:PL    .:0,0:.:.:.:.:.:.       ./.:0,0:0:.:.:.:.:0,0,0

There is also is a difference in which genotype attributes are 0 or .

Answer 1

[Santosh Anand]

Each DOT represents one missing allele.

. => Haploid organism

./. => Diploid organism

See VCF specs whenever in doubt http://www.internationalgenome.org/wiki/Analysis/vcf4.0/

I guess that your VCF is old (and hence doesn't follow the convention) as usually you will not find haploids and diploids in the same VCF.

[Emily]

Could be X chromosome, in which case males and females can be represented in the same VCF.

[William]

The single record is from the output of a recent variant calling of a diploid organism with GATK4. GATK4 variant called multiple samples via GenotypeGVCFs (i.e. there was intermediate GVCF output that was later merged).

[Santosh Anand]

Oh thanks @Emily, I completely forgot about the Sex-chromosomes!

[vanuhi.hambardzumyan]

I came across this question, and I see that it is an old discussion, but I will try. SO I am getting vcf file from freebayes for 1 autosome diploid chromosome for 51individuals, but for some missing genotypes I am getting "." and for others with "./." in the same vcf file. While calling variants with freebayes I mentioned ploidy 2. Does anybody have an idea why this happens and how can I fix it?

[Emily]

Are they male? Are the variants on the sex chromosomes?