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!

HaplotypeCaller gives out variants that are not present in the BAM file

mschuemschue BerlinMember

Dear Geraldine,

I am using the last version of HaplotypeCaller on human diploid BW alignments and sometimes I find variants in the VCF file that have been called with a high quality that are simply not present in the BAM (sometime missense somtimes deletions) file (that had been duplicate removed, realigned and recalibrated before). How could I find out at what step the mistake happens.

Thanks a lot for the software and your help,



  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    Hi Markus,

    Do you mean you don't see evidence for the variants in the reads?

    If so, try enabling display of soft-clipped reads in IGV. It may be that the evidence is there but is not displayed because soft-clips are turned off by default. If you still can't see evidence for these variants please post an example (variant record + screenshot of pileup).

  • KurtKurt Member ✭✭✭

    You can also try the --bamOutput option in HaplotypeCaller just in the region where you have a question, use the -L option to limit it to an interval, so that you can see what HaplotypeCaller is thinking.

Sign In or Register to comment.