I have a RNAseq dataset that may be amendable to a full factorial linear modeling experiment. I have 3 timepoints and 4 genotypes, 2 or 4 biological replicates for each sample. Can anyone point me in the right direction on how to do the factorial linear modeling from DESeq2? I am just a beginner and do not have much knowledge/background in statistics. Would you please give me a detailed instruction or script to do the factorial linear analysis from DESeq2?
A pointer in the right direction would be to mention that the authors of DESeq2, edgeR, and limma have put together a very comprehensive set of docs/user manuals that cover the topic of linear models applied to gene expression in great depth.
You should take the time to really read through those rather thoroughly. You would then be equipped to come back and ask a more focussed question regarding your analysis. As it stands now, your question is way too broad for anybody to help, as it sounds like you just need help with "everything."
In these cases (when analyzing multi-factorial experiments) I've found that it often makes things easier to think about when you setup the design by encoding the factors (and their levels) into a single factor, as described in section 9.5 of the limmaUsersGuide entitled "Interaction Models: 2 × 2 Factorial Designs". Specifically I'm talking about the "Analysing as for a single factor" (section 9.5.2).
The ball is in your court, now, as you've got a lot of homework to do ;-)
Hi Steve,
Thanks for your helpful comment.
Now it's time for me to start reading the manuals of DESeq2, edgeR and limma.
Emily