Should RNA-seq get conducted at the same time/batch?

Question

[wuyc]

I have sequenced 6 samples, however, one sample went wrong. So, do I have to re-sequence all the 6 samples, or I just need to sequence the bad one?

Thanks!

[grant.hovhannisyan]

What do you mean by went wrong?

[GouthamAtla]

I guess Its not practical to sequence all samples again. It depends on your experimental design and what do you want to do with the data.

Answer 1

[WouterDeCoster]

The entirely-correct answer would be yes, to be entirely sure you are not looking at a batch effect, you should repeat all 6.

But the pragmatic answer is: if you make sure the same kit is used, the same reagents, the same technician, the same machine, etc then everything should be quite okay. But you should keep in mind that this sample was treated "slightly" different.

[Kevin Blighe]

As per Wouter and others, it's more practical to just re-do the 'bad' sample. When you re-sequence it, it will most likely be [sequenced] to a much different read depth depending on the level of barcoding that you do; so, just include batch as a covariate in your downstream modelling (and in the DESeq2 / limma / EdgeR design models, should you use those).