Problem with allele-frequencies for deletions using HaplotypeCaller
I am using the last version of HaplotypeCaller on human diploid BWA alignments that have been duplicate removed, realigned, and recalibrated). In the case of heterozygous deletions it often happes the HapotypeCaller gives out a near 100% allele frequency (e.g. homozygosity) despite the fact that the deletion is only present in 50% of the reads. How can I correct that?
Thanks a lot for the software and your help,