opinions on rna-seq validation

[nsl]

Hi,

I am close to writing up an article which we hope to submit for publication in the coming months. I have 4 rna seq treatments with 3 biological reps each. I've chosen a fairly small list of genes (21) based on Fold change, FPKM , p val and of course biological interest.

What is the accepted protocol for such studies now? Are ppl doing qpcr validation assays? has anyone gone through the review process and been asked to show proof of DE genes. Any input would be greatly appreciated.

thank you.

nsl

[mbblack]

It will depend entirely on what question(s) you are actually trying to answer with these selected genes (and also somewhat related, on which journal you submit to).

If your study is of a very general exploratory nature, you may very well never have the issue of validation raised at all. If your study is attempting to definitively answer some question of biological function or relationship highly specific to those specific 21 genes, then yes, a reviewer may feel validation is necessary.

In my experience, there is no absolute answer to your question, and the specifics of your study and your presentation of the question and your data interpretation will determine if validation of those specific 21 genes is necessary or not. Editors and reviewers will base their criticism (or lack of it) for validation based on what you give them in the way of a manuscript.

While I cannot see anyone being criticized for including validation of such a small group of genes, whether it is necessary and adds to the value of the data interpretation is a contextual question, dependent on your study and your presentation of it. Some studies will clearly benefit from validation (or even depend on it), while for others, it will add very little if anything and thus will likely be seen as optional.

[nsl]

Micheal,

Thank you for being kind enough to answer. I understand and agree. This field seems like a moving target and I wanted to know whether the protocol has changed.

I view and am shaping this publication as a screen and am introducing potential candidates for the biological phenomenon we focus on. In my view, this step should not warrant validation, but my PI ( who is not very familiar with the field of rna-seq or computational biology) is worried about it.

In your opinion, should these need validation even though I am presenting them as potential candidates?

Thanks again

nsl