Dear all,

Using the Limma package, I am analyzing an Agilent single-color microarray experiment with two groups (Treated and untreated). When I run a decideTests function with "global" as a method, I obtain a certain number of up- and down-deregulated genes. How can I extract the correspondent list of DEGs with their expression values? If I understand well, topTable and decideTests functions give the same results only if the "separate" method is used with the default adjust method, and I cannot choose the "global" method as an option in the topTable function.

> DEGs = decideTests(fit2, coef="Treated_NGF_72hvsNon_Treated_NGF_72h", method="global",adjust.method="BH",p.value=0.05,sort.by="p", lcf=log2(2))
> 
> summary(DEGs)
       Non_Treated_NGF_72h Treated_NGF_72hvsNon_Treated_NGF_72h
Down                     0                                  107
NotSig                   0                                30285
Up                   30507                                  115

> topTable(fit2, coef=Treated_NGF_72hvsNon_Treated_NGF_72h, n=222, method="global", adjust="BH", p.value=0.05, sort.by="p")
Errore in topTable(fit2, coef = Treated_NGF_72hvsNon_Treated_NGF_72h, n = 222,  : 
  argomento non utilizzato (method = "global")

Thank you very much in advance for your help.

Giulia

The first coefficient in your linear model appears to be an intercept term. In this context, you should not be using method="global" because including an intercept amongst the tested contrasts doesn't make any sense and will give too many positives for the contrasts of real interest.

Also, the decideTests function does not have a coef argument, which will just get ignored.

You appear to have mistyped the lfc argument, but I do not recommend logFC cutoffs, as the help page for decideTests explains. In the code as you show it, the lcf argument will just be ignored.

You should simply be using

topTable(fit2, coef="Treated_NGF_72hvsNon_Treated_NGF_72h")

Thank you very much for your corrections and suggestions.

Giulia