1/0 genotype generated by "bcftools norm" at multi-allele sites

[weishwu]

I ran "bcftools norm" to split multi-allele variants into individual rows. For some of them, 1/0 genotypes were generated.

For example, this variant

1	970549	.	TGGGG	TGG,TGGG,TG,T,GGGGG	1.53856e+06 . . GT:AD:DP:GQ 2/3:1,0,3,7,0,0:11:52

became

1	970549	.	TGGGG	TGGG	1.53856e+06 . . GT:AD:DP:GQ 1/0:1,3:11:52
1	970549	.	TGGGG	TG	1.53856e+06 . . GT:AD:DP:GQ 0/1:1,7:11:52

Why did the first one become 1/0 instead of 0/1? Thanks.

[pd3]

There is no special reason for this, it is just an implementation quirk. The genotypes are unphased, so hopefully it should not matter?

[weishwu]

So far it didn't cause any issue. Just hope it could keep the consistency unless there is a reason for it...