Dear GATK team,
I have an example of a deletion spanning another mutation resulting in a star allele and I need some help to understand the called genotypes.
I have 3 individuals.
Individual 1 has only a deletion (BP_RESOLUTION)
9 418268 . G . . . GT:AD:DP:GQ:PL 0/0:18,0:18:15:0,15,225
9 418269 . TTTTG T, 160.73 . . GT:AD:DP:GQ:PL:SB 0/1:4,5,0:9:99:198,0,165,210,181,391:0,4,1,4
9 418270 . T . . . GT:AD:DP:GQ:PL 0/0:13,4:17:0:0,0,333
9 418271 . T . . . GT:AD:DP:GQ:PL 0/0:13,4:17:0:0,0,326
9 418272 . T . . . GT:AD:DP:GQ:PL 0/0:13,4:17:0:0,0,328
9 418273 . G . . . GT:AD:DP:GQ:PL 0/0:13,4:17:0:0,0,355
9 418274 . T . . . GT:AD:DP:GQ:PL 0/0:12,4:16:0:0,0,281
Individual 2 has a deletion AND a mutation spanned by this deletion (GVCF)
9 418264 . T . . END=418268 GT:DP:GQ:MIN_DP:PL 0/0:35:51:33:0,51,765
9 418269 . TTTTG T, 366.73 . . GT:AD:DP:GQ:PGT:PID:PL:SB 0/1:21,12,0:33:99:0|1:418269_TTTTG_T:404,0,760,467,796,1263:10,11,5,7
9 418272 . T C, 509.77 . . GT:AD:DP:GQ:PGT:PID:PL:SB 0/1:14,21,0:35:99:1|0:418269_TTTTG_T:538,0,505,581,568,1149:7,7,10,11
9 418273 . G . . END=418274 GT:DP:GQ:MIN_DP:PL 0/0:34:0:33:0,0,384
Individual 3 has only the deletion (GVCF)
9 418262 . T . . END=418268 GT:DP:GQ:MIN_DP:PL 0/0:10:12:9:0,12,180
9 418269 . TTTTG T, 96.82 . . GT:AD:DP:GQ:PGT:PID:PL:SB 0/1:1,4,0:5:17:0|1:418269_TTTTG_T:134,0,17,137,30,167:0,1,0,4
9 418274 . T . . END=418274 GT:DP:GQ:MIN_DP:PL 0/0:5:0:5:0,0,105
The resulting genotypeGVCFs with these 3 individuals :
9 418269 . TTTTG T 705.98 . AC=3;AF=0.500;AN=6 GT:AD:DP:GQ:PGT:PID:PL 0/1:4,5:9:99:.:.:198,0,165 0/1:21,12:33:99:0|1:418269_TTTTG_T:404,0,760 0/1:1,4:5:17:0|1:418269_TTTTG_T:134,0,17
9 418272 . T C,* 646.23 . AC=1,1;AF=0.167,0.167;AN=6 GT:AD:DP:GQ:PGT:PID:PL 0/0:13,4,4:17:0:.:.:0,0,328,0,328,328 0/1:14,21,0:35:99:1|0:418269_TTTTG_T:538,0,505,581,568,1149 0/2:1,0,4:5:17:0|1:418269_TTTTG_T:134,137,167,0,30,17
As expected, everybody is 0/1 for the deletion at position 418269. What I don't understand is the genotypes at the position 418272 :
- individual 1 : for me he has to be 0/2 but he is called 0/0. Interestingly, if I transform its gvcf in GVCF mode (resulting in a band from 418270 to 418274), its genotype becomes 0/2
- individual 2 : this individual has the deletion and the mutation. For me he has to be called 1/2, but he's called 0/1.
- individual 3 : this individual is called 0/2 as expected.
So, I don't understand both called genotypes for individual 1 and 2. Morevover, it seems that BP_RESOLUTION and GVCF are processed differently.
Those GVCF were produced using HaplotypeCaller from GATK-3.3. The genotypes presented remain the same using GenotypeGVCFs from version GATK3.3, GATK3.4-46 or GATK-3.6
Best,
PS : Is this normal that no tag are displayed when writing a message ? I used the "code" tag for genotypes but I coudn't remember others. I use Firefox browser 49.0