Hi everyone!
I am working with WGCNA and GWENA packages in attempt to get consensus gene modules for several datasets (from about 5 to 500 samples). I would like to find best combination of parameters in net construction and modules detection that would give me reasonable results, and then to apply those parameters to all of my datasets. Now I am trying to select the best power value.
However, I found out that for some datasets (apparently, not really depending on the number of samples in them...) curve of Scale Free Topology R^2(power value) saturates before reaching the proposed threshold for R^2 of ~0.8. Should I choose the maximum power value in this case? And should I determine power value in each dataset separately, or choose one for all of them?
Thank you in advance!
