I have directional RNA-seq data and read count libraries derived from Bowtie2 alignments to gene-models (cDNAs) for both sense and antisense transcripts. I want to know how I can statistically test the differences in antisense-to-sense ratios between my test groups and my control groups.