Dear GATK team,
I have analyzed samples sequenced with an Illumina custom amplicon sequencing kit.
I aligned samples with BWA mem algorithm and then I applied haplotypecaller following the suggestions for hard filtering:
http://gatkforums.broadinstitute.org/gatk/discussion/2806/howto-apply-hard-filters-to-a-call-set
I used two different strategies: with BQRS and without BQRS and I had different results.
In particular using BQRS in two different samples I have 2 variants that are not called by haplotypecaller, I saw them with other callers and they were also validated with Sanger sequencing.
Without the BQRS haplotyecaller is able to detect the variants in both cases.
My question is: Do you have any experience in this kind of problem?
Is there any concerns about using BQRS with amplicon sequencing? Is it better to perform the analysis without BQRS?
Thank you so much,
Stefania
