Hi all,

I understand WGCNA created to assess gene expression data, however; I have noticed that this method has been applied to microbial communities in some studies (Duran-Pinedo et al., 2011; Aylward et al., 2015; Guidi et al., 2016; Wilson et al., 2018).

I am trying to do WGCNA for microbial data (ASV abundance matrix). After doing filtering and removed the low abundant taxa (ASV), I have done the analsyis and picked the soft threshold at 14. As you can see in the figure below-

and then -

modules.Y <- blockwiseModules(omics_data,
                                power = 14, 
                                networkType = "signed", 
                                TOMType = "signed",
                                corType = "bicor",
                                maxPOutliers = 0.05,
                                deepSplit = 4, # Default 2
                                minModuleSize = 10, # Default 30
                                minCoreKME = 0.5,      # Default 0.5
                                minCoreKMESize = 2,    # Default minModuleSize/3,
                                minKMEtoStay = 0.5,    # Default 0.3
                                reassignThreshold = 0, # Default 1e-6
                                mergeCutHeight = 0.2,  # Default 0.15
                                pamStage = FALSE, 
                                pamRespectsDendro = TRUE,
                                replaceMissingAdjacencies = TRUE,
                                numericLabels = TRUE,
                                saveTOMFileBase = "TOM",
                                verbose = 3,
                                nThreads = 10,
                                maxBlockSize=8000)

I observed 24 modules. So far good (I suppose) 😊

Because my data has batch effect, therefore I removed the batch effect from my data and done the same analysis as mentioned above and observe that my soft threshold is going beyond 30 -

For downstream analysis, I can not provide the power over 30, therefore I used 30 as a soft threshold to generate 11 modules.

So, what would you recommend in this case? Do you think it's okay to use 30 (soft threshold)?