Hello everyone, I would like to ask for your help and appreciate any comment from you! Here is my question: I want to compare group A with group B using a reference design by cDNA arrays and find DE genes. Each group has 6 biological replicates and A and B always come with Cy5 in each clip. I used the following codes in Limma. Design<-cbind (A=c(1,1,1,1,1,1,0,0,0,0,0,0), B=c(0,0,0,0,0,0,1,1,1,1,11)) fit1<-lmFit(MA, design) Then I got two coefficients, the first one is the average fold change for group A and the second one is for group B. contrast.matrix<-cbind(A=c(1,0), B=c(0,1), AvsB=c(1,-1)) fit2<-contrasts.fit(fit,contrasts=contrasts.max) eb<-ebayes(fit2) ebt<-eb$t ebtp<-eb$p ebb<-eb$lods So, the last column in ebt, ebtp and ebb will be my interesting (group A compares with group B). What about the first two columns? For example, can I use the first column in ebt, ebtp and ebb to find DE genes for comparing group A with the reference group and second column to get DE genes for comparing group B with reference group? I also do not understand the relationship between eb$t with eb$p, since in my output, the larger eb$t does not come with smaller eb$p. Is the eb$p for adjusted p-value? If not, in order to use this modified t-test to get DE genes, do I need to find adjusted p-value from multtest package? Can I use other rules for finding DE genes, such as ranking abs(t-value) or B-statistc? Thanks in advance! Darwin --------------------------------- Yahoo! Small Business $15K Web Design Giveaway - Enter today [[alternative HTML version deleted]]