Question about module preservation / reproducibility when creating the modules

[DelongZHOU]

Hello,

Is there a way to evaluate how stable the modules are? In the projection section you discussed how modules created from one set may not apply to another set. Here I'm more interested in the stability within the same data set. Is there a way to measure how robust the modules are? Thanks!

[smorabit]

Currently hdWGCNA does not implement a method for determining the module stability. I am definitely interested in adding this as a new feature but I cannot promise when it will be done. In principle this would require performing something called "robust WGCNA" (rWGCNA), where you are basically performing WGCNA separately on bootstrapped samples of the dataset and then merging the results across different networks. You can see an example of rWGCNA in bulk RNA-seq in this paper.