The -pe,--paired setting results in big difference in exonic mapping

[Nina Hertoghs]

Hi all,

We have some RNAseq data generated by a vendor, which resulted in lower than expected percentages of exonic reads-mapping. The vendor argues that it is due to a setting in qualimap which they refer to as the "-pe,--paired setting", which is selected as a default since we have paired-end read sequencing data. This setting allows fragments to be counted vs. reads, but this makes a huge difference in the % of reads mapped to exonic regions. (50% vs >80%). We would like to understand how particular types of reads are counted, and which setting would be the best for our data to get a feel for the actual quality of the data. Would it be acceptable to deselect the "paired" setting in the case of paired-end sequencing, or is this not a valid approach? 

Thanks!

Nina