Meaning of single GT value

[hdashnow]

Hi,

We're seeing many loci with GT values like "0" or "1" in a diploid sample on autosomes. Should we interpret these as "0/." and "1/."?

@GARRISON-2 could you please add some example output?

Warm regards,
Harriet

[GARRISON-2]

Example output:

chr1 10627 . A CNV:TR . PASS END=10800;RUL_REF=29;SVLEN=173;RN=1;RUL=29;RB=263;CIRB=-14,11 GT:DP:AS 0/1:46:25,18
chr1 103260 . G CNV:TR . PASS END=103280;RUS_REF=TAAAA;SVLEN=20;RN=1;RUS=TAAAA;RB=33;CIRB=-9,0 GT:DP:AS 1:14:14
chr1 668334 . G . PASS END=668486;RUL_REF=76 GT:DP:AS 0:30:30
chr1 862076 . T . PASS END=862088;RUS_REF=GATG GT:DP:AS 0:31:28
chr1 905012 . T . PASS END=905114;RUL_REF=28 GT:DP:AS 0:26:26
chr1 912714 . A . PASS END=912732;RUL_REF=9 GT:DP:AS 0:34:32
chr1 930081 . A . PASS END=930152;RUL_REF=21 GT:DP:AS 0:30:29
chr1 969958 . G . PASS END=970156;RUL_REF=55 GT:DP:AS 0:25:25
chr1 1091706 . T . PASS END=1091782;RUL_REF=38 GT:DP:AS 0:30:29
chr1 1108170 . T . PASS END=1108275;RUL_REF=32 GT:DP:AS 0:26:25
chr1 1141113 . A . PASS END=1141137;RUL_REF=12 GT:DP:AS 0:39:38
chr1 1158353 . A . PASS END=1158375;RUL_REF=7 GT:DP:AS 0:41:38
chr1 1168009 . G . PASS END=1168125;RUL_REF=35 GT:DP:AS 0:35:30
chr1 1221641 . C . PASS END=1221664;RUL_REF=7 GT:DP:AS 0:32:32

[hdashnow]

Hi, just following up:
Should we assume a GT of 0 means 0/0, 0/. or is this a bug?
We're currently benchmarking this algorithm so the answer could substantially change our results. These genotypes make up large proportion of the output.

I would really appreciate if you could let me know ASAP as we are submitting this paper in a few weeks.

Warm regards, Harriet

[readmanchiu]

sorry about the delay in reply, GT 0 should mean 0/0 (homozygous reference), and GT 1 should mean homozygous for the alternate allele. I seem to have "0/0" in my test files, so surprised to see a single number for homozygous allele. I will double check and make a fix to put the slash back in for homozygous allele if I can confirm on my end. Thanks for reporting.

[hdashnow]

Thanks @readmanchiu! We'll convert 0 -> 0/0 and 1 -> 1/1 in our analysis.

Since it sounds like that output is unexpected, here's what we're seeing.

We ran STRaglr with ~40,000 random TR loci on HG002 and see the following distribution of genotypes:

count	GT
30914	0
910	0/0
970	0/1
2658	1
192	1/2

Only ~670 are on the X chromosome; the rest are autosomes.

[readmanchiu]

Thanks for the info. I will sort out the single digit issue for homozygous alleles.