quantifying ribosomal content in Illumina libraries

Question

[Richard]

Hi all,

Is there a standard tool people use to quantify the relative amounts of Ribosomal reads in RNA-seq libraries?

Answer 1

[Danielk]

Yes, you can use Picard CollectRnaseqmetrics.

[Richard]

Does everyone use the same set of Ribosomal intervals on hg19, or is there some variation between group?

[Danielk]

I use these: https://github.com/dakl/genome

I've also seen people get them from the ucsc mysql dB. I guess they give very similar results.

[Richard]

Hmmm. This list doesn't seem to include 28S or 18S which I've read are the great majority of detected rRNAs

[Danielk]

You're right. In GRCh37, it maps to an unplaced contig. However, using these regions does distinguish between good and bad ribodepletions. With a good depletion, we get 0-2% rRNA content, and I our failed experiments with a robotic protocol with a bad sometimes see as high as 30%. With TruSeq Stranded Total RNA with Ribo-Zero Gold, we consistently get 0-2%.

But you are right, these values should be interpreted as relative within a series, rather than estimates of absolute rRNA content.