How does GATK IndelRealigner behave with indels in homopolymer stretches?
I would like to ask if someone else have never encountered the same problem I have by using the IndelRealigner to adjust the alignment of indels occurring within homopolymeric stretches.
I'm trying to realign bam files of mitochondrial aligned reads (obtained with a simulation) by using a custom vcf I made as reference for known mitochondrial indels.
Specifically, I know that in simulated reads there will be an insertion of a C in a six C-stretch. I also know that the aligner I use (GMAP) is prone to align such indels at the beginning of the stretch (i.e as 3565insC). However, I prefer such insertion to be aligned at the end of the C-stretch (i.e. as 3571insC), as (unfortunately) was always been annotated in literature. Thus I reported in my custom vcf, used with the -known option, this annotation:
chrRSRS 3571 . C CC . PASS AC=1;AN=2
..hoping that it would be considered as reference for realignment. However, IndelRealigner failed in realigning the C insertion.
I wonder if there is any parameter I can use to force this realignment..
I would be very grateful if someone could share with me his/her experience about this topic