RNA-Seq library size

[buthercup_ch]

Hello,

Does any have strong experience and can give good advice regarding the optimal concentration of libraries for RNA-Seq?

I normally have worked with 30-50 ng/ul, but in the last set of samples, libraries were too small, of about 2-10 ng/ul.

I am worried if I would lost any non-abundant transcript and the preparations are biased.

Thanks for the comments

Cristina

[jteeee2]

What type of RNA-Seq are you doing? Poly-A enriched? rRNA depleted?

How much Total RNA did you start with? Was it of high quality (RIN score, etc.)?

[buthercup_ch]

Hello jteeee2,

We are working with prokaryotes, and the protocol followed for library preparation was:
- Ribo-Zero Magnetic Gold
- TruSeq Stranded mRNA Sample Preparation

In a previous experiment we started from 1 ug of total RNA, with final yields of 30-50 ng/ul. But in the current experiment, where libraries are 2-10 ug/ul, we started from 500 ng.

[jteeee2]

There are a lot of possible reasons your yields may be lower than what you have seen previously. In a perfect world, we would see consistent recoveries from one RNA-Seq experiment to the next, but that doesn't always happen. Overall yield depends on many factors. How does the quality of your total RNA compare to what you've used previously? Are you quantifying RNA concentration using a fluorimetric assay like Qubit or Picogreen? Has anything changed within your RiboZero protocol? Equipment, reagents, etc? Are you using Ampure beads to cleanup the ribozero reactions or ethanol precipitation?

The real question is whether the libraries you generated are of sufficient quality/quantity to sequence. The yield range that you mention should be plenty of material to sequence. Have you run them out on a Bioanalyzer? Do the traces look like they have in the past? Are the traces smooth or do you see large spikes that may indicate over represented sequences?

[buthercup_ch]

Exactly, that is my main concern... As I don't have that much experience and never yet faced this scenario..., I am not sure if the libraries will be good for sequencing.

They look pretty good as for Bioanalyzer, although they still show a high peak of adaptor dimer. We are going to clean-up once more and see how they look.

[jteeee2]

Do you have any control RNA material that you've made into a library using this workflow? The only way to avoid a situation like this in the future is to always include some sort of batch control. Use a good quality total RNA sample that has some expected level of performance. This allows you to compare different batches of library prep. If you start seeing lower yields for this control sample, then you know something is up with the kit, or an error was made during the library prep.

It sounds like you just have a lower yielding set of samples. Frequently, this is due to a mis-quantification of the RNA up front or an inefficient recovery of the RNA post ribo-zero. As I said previously, if you think the traces look ok, then I would recommend sequencing them. You may consider doing qPCR to quantify the libraries because there will definitely be a portion of that library that does not have the appropriate illumina adapter.