Hi Arthur,

Thanks for your answer. Yes indeed it is a wild guess and I am quite uneased at the idea to proceed with the seuqnecing of all the samples. I think we will have to do some tests to see how the presence of the gut in the extraction influences the quality check of the RNA. We might also decide to just run a few samples and look at the library we obtain to take a decision.

Hi! Interesting problem.

Some time ago I would say something like Artur, until I saw the work of a colleague doing deep sequencing of plant roots+mycorrhiza - in which you cannot really clean out all the soil remnants during prep - where they had a significant amount of RNA-seq reads coming from a wide variety of organisms including... worms. Their interpretation was that they were getting reads from dejects and I couldn't come up with anything better...

So now I think it is quite likely that you may get reads from bacteria.

A pilot sequencing round (Miseq, for example) will be important for you to get an idea of your data, but if you have any bona fide Daphnia mRNA sequence you could try the old approach to get a general sense if you are having significant mRNA degradation: use and oligo-dT based RT-qPCR to obtain a 5' to 3' ratio (e.g doi:10.1093/nar/gkr065). 

We were going for the HiSeq because we would have a higher number of reads sequenced than with MiSeq (we usually reach 200 Millions easily with our facility) and would allow us to still have a good representation of the expressed genes for differential expression studies. It would be paired end stranded sequencing.

Margarida thanks for the input, it is a very good method to check quality indeed. However, sadly, I do not think we have such sequences for Daphnia, or only a handful which might not give a proper estimation of the degradation.

As for reads of bateria, we would be doing polyA enrichment before sequencing and bacterial RNA usually lack the polyA tail, so it would not be a concern for us. However, we expect to see a large amount of algae RNA since we included the gut in the extraction.

Well, I think so far our idea is to try to do the same extraction protocol on two different treatment (animals with algae in the gut, and animal treated with a bead solution to push out the algae from the gut). Then we will run again the RNA on the bioanalyzer to see the profiles.

If indeed the degraded RNA seem to come from the algae, we will go ahead and sequence by reducing the original multiplexing, and do the necessary post processing controls (normalization and 3' over-representation controls) to be sure that our RNA was not the degraded one.

In any case, it still looks like we could get a nice technical paper out of it, for all the folks out there trying to deal with whole animal RNA extraction for RNASeq! Indeed there is not that many papers (if any) dealing with mixed RNA content from whole animal extraction in the literature...

Hi Celine, afraid I can't offer and help to you specifically.

I actually have a question myself. I'm also doing whole animal RNA extraction using a coral (Acropora humilis).

I'd like to get some more information on necessary post processing controls and would appreciate if someone could point me towards some good resources. Thanks!

Hi MAthew,

Sorry for the delayed answer. For your post processing, I would do first FastQC to control for quality. Then it really depends of which resources are available to you... Do you have a genome or a transcriptome available for your organism?

Hi Celine,

I'm building this transcriptome de novo. I have similar related organisms I could compare to though before (coral species within same genus). Thanks for advice with FASTQC I've done a tutorial on this before and I was planning on it.

Does anyone have information/literature/blog posts on how to interpret electrophereograms? I'm trying to figure out my own at the moment as well

This maybe:

http://genomics.nchresearch.org/submitting_RNA.html

Basically what you are trying to look at is not the RIN number (since you seem to work on coral). RIN has been developed for mammal RNA. However, you want to look at the presence of smaller fragments than the 18S and 28S (the two peaks at the end). Then the ratio of 18S/28S should be 1:2 for good quality.

Hi Kay,

I know, I was doing this treatment along with an antibiotic treatment for DNA extraction when I did RADseq libraries.  However, for this experiment, I did not want to have stress expression, and if you feed them with beads, they really don't like it (they act stressed within the first 7 hours of the treatment), so I would get expression of stress mRNA that do not interest me... It is a tricky situation...

But today there is an update: I got the RNA sequenced, the sequences are good quality, the complexity of the library passes the checks so far, and the levels of duplication are minimal... I will map the reads to the genome next week and see what we get so far...

Did any of the people in your lab come across such RNA profiles? I talked to people in Australia and in Germany (Daphnia labs too) and some of them had bad RNA profiles but sequenced it anyway and got great results...

Cheers

Celine

Hi Kay,

Thanks for that!

I now mapped my reads to the Daphnia genome with Tophat (I am still fighting with the indexing of the transcriptome so I just did it without the gff file). Results are great, since I got 74.9% overall read mapping rate, and 65% concordant pair alignment! I did my estimations for the amount of reads needed per individuals based on 50% algae contamination in the library.  Seeing the RNA profiles, I was hoping not to have more than half of my library being algae and I see that I am completely ok!

Now I need to properly index the GFF and use it to map the reads to be able to do post-alignment quality checks and see if really all is good.

Say Hi to John from the daphnia lab in Montpellier!!

Cheers

Céline