How Haplotypecaller could call a 17-bp duplication that is not seen in sorted bam (in IGV)?

0

Entering edit mode

2.1 years ago

pragnapcu ▴ 10

enter image description here

A heterozygous 17bp duplication variant is shown from GATK HaplotypeCaller however, IGV is not showing the same. I understand the region is not entirely covered. But, how HaplotypeCaller managed to call this variant?

BaseRecalibrator IGV GATK HaplotypeCaller frameshiftinsertion • 879 views

ADD COMMENT • link 2.1 years ago by pragnapcu ▴ 10

1

Entering edit mode

Hi! HaplotypeCaller calls SNPs and indels via local de-novo assembly of the reads. Therefore, the alignment you see in the input BAM files, could vary from the actual reassembly that the tool does "on-the-fly".

ADD REPLY • link 2.1 years ago by iraun 6.2k

0

Entering edit mode

HaplotypeCaller does the denovo assembly whenever it encounters a sign of variation however, in this case, variation is not seen. Then, how it might have done the denovo assembly. I am just trying to understand it better.

ADD REPLY • link 2.1 years ago by pragnapcu ▴ 10

1

Entering edit mode

Well, I don't think it is always easy to see what GATK sees just by looking at the BAM file. It looks like a GC enriched region, quite repetitive, so probably GATK, with its intrinsic parameters and thresholds to evaluate each region, it sees some kind of variation sign.

ADD REPLY • link 2.1 years ago by iraun 6.2k

0

Entering edit mode

Have you turned on 'show soft-clipped bases' in IGV? My guess is that there is soft-clipping here showing additional inserted sequence (or sequence from the other side of the deletion).