We've moved!
This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!

Why is HaplotypeCaller calling very few variants

Hello, I am currently working on a benchmark analysis using different variant calling methods, including your HaplotypeCaller algorithm, using version 3.7. However the results I am getting from your algorithm show very few variants in comparison with the other tools I am using. I am not sure if any default parameters are limiting the sensitivity of the algorithm. Everything is being done with simulated reads from dwgsim program (which besides generating the reads in fq format, it also generates the vcf catalog with the simulated variants). The data points generated by gatk are around half if not less than the ones generated with the other variant calling tools.

I am leaving everything by default in the HaplotypeCaller except for: genotyping_mode DISCOVERY; -stand_call_conf 0; -dt NONE.

Thank you for your help


  • SheilaSheila Broad InstituteMember, Broadie ✭✭✭✭✭


    Can you post some IGV screenshots of the BAM files and bamout files in the regions where you think variants should be called/are getting called by other callers?

    Can you also confirm the base qualities and mapping qualities are high in those regions? What are the base qualities like? Are they all set to 20?


Sign In or Register to comment.