Do I need to explicitly remove rRNA from E. coli RNAseq data?

Question

[Lina F]

Hi all,

I am looking at E. coli RNAseq data (generated from the Kapa RNA Hyper prep kit). Someone recommended I specifically remove rRNA from the data before mapping against the genome to calculate differentially expressed genes.

I did some more reading online, and I found a few papers that don't specifically mention this step for microbial RNAseq experiments (i.e. Yung et al 2016, Tan et al 2015), and other posts that suggest specific rRNA removal is not needed (i.e. this, but for mouse, not microbes).

I'd love to hear if anyone has any definitive resources to share on this subject -- thanks in advance!

[GenoMax]

As long as you don't count/include reads that are aligning to rRNA in your DE analysis you should be ok.

If you do have a lot of reads mapping to rRNA then it is going to result in you omitting a lot of data which may not leave enough for the DE analysis (and if the amount of those rRNA mapping reads is uneven across the samples then you would have an additional issue).

[Lina F]

Thanks for the feedback!

I talked to the lab tech and they are expecting as much as 80% rRNA, so I will see how much actual usable data I will be left with.