Are all antisense transcripts from polyA enriched library RNA-seq polyadenylated?

Question

[Anny]

Dear all, I have a confusion about antisense detection. From my RNA-seq result which is strand-specific and polyA enriched during library preparation, about one-third of genes showed antisense expression. Does it mean all the antisenses detected here have polyA tails/polyadenylated?

Thank you for your time!

[Satyajeet Khare]

Quite a few antisense transcripts are known to get polyadenylated so it is not very unusual. You can check if all the antisense transcripts you observe in your dataset are polyadenylated or not by doing further experiments.

Answer 1

[Phil Ewels]

One possible explanation for this is a poor efficiency of the poly-A enrichment, possibly combined with DNA contamination. No enrichment is ever 100% efficient, so other material can always get into the library. If you have antisense reads in exons, this can sometimes be due to DNA contamination. This is fairly easy to check - have a look at how many reads are aligning between genes. Also, I guess whether this is a worry or not depends on how _much_ antisense expression those genes showed ;)

See this QCFail article for more information on DNA contamination in RNA-seq libraries.