WGCNA: minModuleSize is not responding correctly

Question

[bioinformaticssrm2011]

This is a cross question posted in bioconductor.

I am running the WGCNA, but getting an error. As I can see that the minimum module size that I am getting is 1, but by default it should not give anything less than 30.

modules <- blockwiseModules(transcriptomics,
                                    power = 9, 
                                    networkType = "signed", 
                                    TOMType = "signed",
                                    corType = "bicor",
                                    numericLabels = TRUE,
                                    saveTOMs = save_TOM,
                                    saveTOMFileBase = "TOM",
                                    verbose = 3,
                                    maxBlockSize=8000, 
                                    nThreads = 11)


rownames(modules$MEs) <- rownames(metatranscriptomics)
names(modules$colors) <- colnames(metatranscriptomics)
hubs_M <- chooseTopHubInEachModule(metatranscriptomics, modules$colors, power = 9, omitColors = "0")

Error in hubs[m] <- colnames(adj)[hub] : replacement has length zero

I checked, how many modules I got-

table(modules$colors)

    0     1     2     3     4     5     6     7     8     9    10    11 
60409  8629  3608  3204  1909   483   111    88    32    31     4     1

table(names(modules$colors) == colnames(metatranscriptomics))

 TRUE 
78509

I have looked at this post- https://bioinformatics.stackexchange.com/questions/17598/wgcna-hub-genes-error

But I can't solve this issue. Any suggestions?

[andres.firrincieli]

You can use intramodularConnectivity to calculate the connectivity of genes and identity the top hub in each module (gene with the highest connectiviy).

Regarding the error

Error in hubs[m] <- colnames(adj)[hub] : replacement has length zero

Can you try to convert the module names with colors?