Hi !

I found rlog transformation method in DESeq2 package was very useful for clustering RNA-seq data.

Recently, I was dealing with some time-course data, and ready to run a co-expression network analysis via a R package, which recommends a variance-stabilizing transformation. I'm wondering how to do rlog (vst) for time-course data.

For example, if I have three time-points, Day0, Day1, and Day2, shall I just take the Day0 sample as untreated sample (Day1 and Day2 sample as treated samples), then simply call the rlogTransformation function ?

Thanks so much.

hi,

Yes you can just use vst() or rlog() functions in order to produce variance stabilized data. You don't need to provide sample information or define which one is the treated or untreated.

You can take a look at the plots we have in the DESeq2 vignette and see which transformation you prefer. If you need, the matrix of values from a DESeqTransform object can be extracted with assay() function.