I am comparing the variant calls from samtools and GATK. For samtools, I have been using quality score cut-offs 100 for SNP and 1000 for INDEL (quite stringent) and as a result, many variants are excluded after filtering. In case of GATK, I have been using our default setting (99% sensitivity for SNPs and 95% sensitivity for INDEL) and included only the variants with FILTER field "PASS". I was wondering if there is any more stringent filters that I can apply and that could be equivalent to samtools QS thresholds since it does not look like this is a fair comparison. Any of your suggestions will be appreciated.