Open fellen31 opened 1 month ago
For example: A variant is falsely assigned 0/0
with DP=0
by GLNexus, that should really be a ./.
call. Genmod treats this call as genotyped since it's not ./.
:
XR and XD patterns are then excluded, when in reality they should not be:
So a ./.
call should result in
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT father mother proband
X 302253 . CCCTCCTGCCCCT C 100 PASS MQ=1;Annotation=GTPBP6,PLCXD1;GeneticModels=1:XR|XD;ModelScore=1:55 GT:AD:GQ 0/0:10,10:60 0/1:10,10:60 ./.:10,10:60
but a 0/0
call results in
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT father mother proband
X 302253 . CCCTCCTGCCCCT C 100 PASS MQ=1;Annotation=GTPBP6,PLCXD1 GT:AD:GQ 0/0:10,10:60 0/1:10,10:60 0/0:10,10:60
This might also result in a variant being set as de novo, if the parents are falsely genotyped as 0/0
:
parents 0/0
:
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT father mother proband
X 302253 . CCCTCCTGCCCCT C 100 PASS MQ=1;Annotation=GTPBP6,PLCXD1;GeneticModels=1:XR_dn|XD_dn;ModelScore=1:57 GT:AD:GQ:DP 0/0:10,10:60:60 0/0:10,10:60:60 0/1:0,0:0:0
parents ./.
:
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT father mother proband
X 302253 . CCCTCCTGCCCCT C 100 PASS MQ=1;Annotation=GTPBP6,PLCXD1;GeneticModels=1:XR|XR_dn|XD|XD_dn;ModelScore=1:57 GT:AD:GQ:DP ./.:10,10:60:60 ./.:10,10:60:60 0/1:0,0:0:0
Not sure how often this would happen, and what the implications really are (if any), ping @jemten @dnil
Perhaps GLNexus should be replaced. No work has been done the last two years, with open issues like assigning
0/0
when it should be./.
(https://github.com/dnanexus-rnd/GLnexus/issues/286), which might have implications for genmod.