Hi all,

if I have expression data from two time-points and I am interested in changes over time- Would it be possible to use the difference in expression between the time-points (delta values) as input for WGCNA (or for co-expression networks in general)?

I am not sure if there could be a problem with the assumptions (for example scale-free connectivity etc.). And I think the interpretation would become more complicated, and probably an unsigned network would make more sense than a signed one?

Thanks for any suggestions!

Hi Lin,

It is possible and it has been done before, for example see this this paper .

So you can log-transform your data (if your data is an RNA-Seq data) and then: countmat2 - countmat1. This will give you 'fold change' expression matrix relative to the baseline.