search for: decidetest

Dear All, I am using the LIMMA package to create 2 contrasts for my data and then calculating the vennCounts of the decideTests from the contrast.fit to be able to create venn Diagrams. The code works fine but the summary(results) shows zeros for all i.e. no gene were up regulated or downregulated. This is not true for my data since toptable output shows Log fold change greater than > 2. I am certain it is a small gli...

...eau-PLATEAU PLATEAU-PLAIN plain 1 0 0 plateau 0 1 0 PLATEAU 0 -1 1 > fit2MA.pi<-contrasts.fit(fitMA.pi,contrast.matrix) > fit2MA.pi<-eBayes(fit2MA.pi) > resultsi.001.fc2=decideTests(fit2MA.pi,method="nestedF",adjust.method="BH",p.value=0.001,lfc=log2(2)) write.fit(fit2MA.pi, results=decideTests(fit2MA.pi,method="nestedF",adjust.method="BH",p.value=0.001,lfc=1), "fit2MA.pi.nestedF.adj_BH.P_001.FC_2.csv", digits=4, adjust="...

I have a very basic question about limma. Assume I have experiments from 3 or more RNA sources in a reference design. It is easy to define individual contrasts but I want to specify a contrast matrix that tests for significant differences among ALL the different RNA sources (i.e. the analogous thing to a simple One-Way ANOVA). How can I do that? Thanks! Max --