Conflict between HaplotypeCaller variant and BAM
Hello, I'm using GATK version 3.6 to analyze a trio exome, and I followed the best practices (prinseq, bwa, markduplicates, baseRecal, printReads, HC and hard filtering, to resume the pipeline; with b37 as reference genome).
HC calls an 43 bases insertion (7:2983941, gene CARD11, C->CCTTGACCAGCTCGTCATTGTAGCTGTCCCGCTCTTCCTTCAT, Qual: 2218.73, Depth: 148) in both the proband (affected) and an affected parent with high quality and depth, and does not call it in the unaffected parent. But when I inspected the BAM to check it, I can't see the insertion, but there is an SNP in the same position (C->T), in both affected samples (please see attached image).
So, I started checking the data, I've found that the reported insertion actually match with the bases in the reference in the same position, and it is not duplicated in the reads. Also, I tried adding RealignerTargetCreator-IndelRealigner to the pipeline with the same results. Finally I tested the samples with Platypus and Samtool mpileup: the first one calls the SNP and the latter doesnt found anything in that position.
I would like to ask for your help to understand this result, and the improve my pipeline if there is a mistake.
Please let me know if you need more information.
Thanks in advance.
BIA (Argentine Bioinformatics Platform)