Pindel understimates AF

[erfan1985]

I have tried pindel on several exome files with known indels. Pindel can detect all but the estimated allele fraction is way bellow what it should be. In most cases where I know the indel is fully heterozygous and the AF should be around 0.5, the estimates by pindel is less than 0.2. I m now wondering what cut-off i should use for AF to call a variant. Is there a way to fix the issue? I m using the default parameters.

Thanks