Hi,

I am trying to analyze a set of particular genes that are all within a module in addition to the HUB genes of that module. I am finding that when I extract the list of genes with the highest module membership at varying lengths, the genes that populate that list slightly change in their ranking. 

For example - when I extract the top 10 HUB genes:

      yellow
 [1,] "LOC106529418"
 [2,] "LOC106513277"
 [3,] "LOC106523395"
 [4,] "LOC106522037"
 [5,] "LOC106524105"
 [6,] "LOC106516167"
 [7,] "LOC106525246"
 [8,] "LOC106516366"
 [9,] "LOC106531344"
[10,] "LOC106522808"

And when I extract the top 20, I don't only see new genes populate the bottom of the list, but rather, I see some new genes making up the top 10 positions, too.

      yellow        
 [1,] "LOC106529418"
 [2,] "LOC106516441"
 [3,] "LOC106528572"
 [4,] "LOC106513277"
 [5,] "LOC106523804"
 [6,] "LOC106523395"
 [7,] "LOC106522037"
 [8,] "LOC106524105"
 [9,] "LOC106516167"
[10,] "LOC106525555"
[11,] "LOC106525246"
[12,] "LOC106523471"
[13,] "LOC106516366"
[14,] "LOC106515286"
[15,] "LOC106517146"
[16,] "LOC106531344"
[17,] "LOC106522808"
[18,] "LOC106534245"
[19,] "LOC106527578"
[20,] "LOC106520168"

Can anyone explain why the ranking of gene module memberships is being adjusted with with each of my attempts?

Here is the code chunk I am following from the tutorial with the value '10' in the 6th line down as what I change between the results I see above:

topGenesKME = NULL
for (c in 1:length(colors20)){
 kMErank1    = rank(-geneModuleMembership1[,c])
 kMErank2    = rank(-geneModuleMembership2[,c])
 maxKMErank  = rank(apply(cbind(kMErank1,kMErank2+.00001),1,max))
 topGenesKME = cbind(topGenesKME,Gene[maxKMErank<=10])
}; colnames(topGenesKME) = colors20
topGenesKME

Thank you in advance for any insight on this!

The code you show doesn't order Gene by rank (only subsets based on maxKMErank). You could easily do it with order, however, which order should it follow: KMErank1, KMErank2, both, just maxKMErank? What happens when there are two maxKMErank which are the same? I would do something like Gene[order(kMErank1, kMErank2, decreasing = TRUE)][1:10] where 1:10 is the number of Genes you want to subset.