#### The support.bioconductor.org editor has been updated to markdown! Please see more info at: Tutorial: Updated Support Site Editor

Question: Comparing groups with paired samples by using limma
0
10 days ago by

Dear All,

I have the following experiment: There are 3 donors (D1, D2 and D3), 2 treatment types (A and B) and 2 growth conditions (B and M). Each donor contributed samples for both treatments and both growth conditions (AB, AM, BB, BM).

I am hoping to examine the difference between growth conditions for each of the treatments separately (AB vs. AM and BB vs. BM) but we are also interested in the comparison between AB and BB. If all samples were obtained from different donors, I would have done the analysis in the following way:

design<-model.matrix(~0+factor(c(rep(1,3),rep(2,3),rep(3,3),rep(4,3)))); colnames(design)<-c("AB","AM","BB","BM");

contrast.matrix<-makeContrasts(AB-AM, BB-BM, AB-BB,levels=design) fit<-lmFit(dataset,design); fit2<-contrasts.fit(fit,contrast.matrix); fit2<-eBayes(fit2);

However, since there is also pairing of the samples based on the donors (from each donor we have AB, AM, BB, and BM samples), I understand that I need to add that information into the analysis, and I am not sure how to do that.

I will greatly appreciate any advice.

Thank you very much for your help!

Sincerly, Anna

limma paired analysis • 61 views
modified 10 days ago by Gordon Smyth36k • written 10 days ago by annadv770
Answer: Comparing groups with paired samples by using limma
0
10 days ago by
Gordon Smyth36k
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
Gordon Smyth36k wrote:

Just add + Donor to the model formula.

Is it also necessary to do duplicateCorrelation in addition to the blocking?