deletion in v3_consensus but not in majority of v3_haplotype present

[sschmutz]

Example for sample 1000392567 sequenced in the run 180321_M02081_0277_000000000-BN2F5-4:

There is a three nucleotide deletion in the consensus:

>v3_consensus
TGTACAAGACCCAACAACAATACAAGAAGAAGTATACATATAGGACCAGGGARARCATTT
TATGGAACA---AYARKAGGAGATATAAGACAAGCACATTGT

This deletion is only present in low frequency haplotypes:

>v3_haplotype_1-support_0.57
TGTACAAGACCCAACAACAATACAAGAAGAAGTATACATATAGGACCAGGGAGAGCATTT
TATGGAACAGACATAATAGGAGATATAAGACAAGCACATTGT
>v3_haplotype_2-support_0.20
TGTACAAGACCCAACAACAATACAAGAAGAAGTATACATATAGGACCAGGGAAAGCATTT
TATGGAACAGACATAATAGGAGATATAAGACAAGCACATTGT
>v3_haplotype_3-support_0.12
TGTACAAGACCCGGCAAGAAGAAAATAAGAATGATACATATAGGACCAGGGAGAACATTT
TATGGAACA---ACAGGAGGAGATATAAGACAAGCACATTGT

...

Is it maybe because that deletion is present at a frequency >15% and somehow "overwrites" the nucleotides at that position?