bcftools strand bias for strand specific libraries

[palb]

Hi all,

I am using samtools/bcftools to call variants using RNA-seq data. We have strand specific paired-end libraries. does it make sense to put some threshold for strand bias (in PV4) of samtools/bcftools to filter some positions considering that we have strand specific data? I assume it is gonna be the bias for the number of first in pair and second in pair reads between the two alleles. Is that correct?

Thanks