Confusion regarding WGCNA

Question

[ldyer2006]

Hi folks, I'm currently attempting some time-course analysis of a recent RNA-Seq run using WGCNA. I have 5 time points, with 4 biological replicates of 2 conditions at each point, so around 40 samples total. The main problem that I'm having is simply one of confusion. I'm not sure exactly how to include the time-course information and sample-matching to WGCNA. Reading through the tutorials, time-course data is simply not added to the experimental design frame.

Perhaps my problem is coming from not understanding exactly how the analysis progresses. Should I simply be treating time as though it were a clinical feature and ignore the sample matching?

Any help would be greatly appreciated.

Answer 1

[Jake Warner]

You do not provide 'trait data' (for you that'd be timepoint and condition) or even replicate info to WGCNA since the network construction is independent of these data. You can use this information in the downstream analysis by correlating traits (timepoints) to the module eigengene expression.

See This tutorial for a walk through.