Dear Bioconductor mailing-list,
I tried to analyze one-color experiments with limma, but I am really not sure about the way to define tech and biol replicates. Any answer very welcome for me to know of my results are correct or not ...
The experiment is as follow :
I have a mutant (mut) and a wild type (wt) strain x 2 biol. rep x 4 technical replicates
So I have :
mut1.1 ; mut1.2 ; mut1.3 ; mut1.4 ; mut2.1 ; mut2.2 ; mut2.3 ; mut2.4;
wt1.1 ; wt1.2 ; wt1.3 ; wt1.4 ; wt2.1 ; wt2.2 ; wt2.3 ; wt2.4
So I have :
- 2 (biol rep) mutants : mut1 and mut2, with 4 tech rep each => 4x2=8 arrays
- and, similarly : 2 (biol rep) wt : wt1 and wt2, with also 4 tech rep each
=> 4x2=8 arrays
==> 16 arrays, one color design.
I (very classically) want to know what is the most significantly different between the mutant and the wt.
My limma code is this one :
#1 and 2 are biol rep for the mutant, and 3 and 4 are biol rep for the wt # I am not sure this is the right way to code this properly ? biolrep<-c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4) design<-cbind(MU=c(1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0), WT=c(0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1)) corfit<-duplicateCorrelation(my.data, design, ndups=1, block=biolrep) #corfit$consensus = 0.7658993 fit<-lmFit(my.data, design, block=biolrep, cor=corfit$consensus) cont.matrix<-makeContrasts(MUvsWT=MU-WT, levels=design) fit2<-contrasts.fit(fit, cont.matrix) fit2<-eBayes(fit2) anadiff.limma<-topTable(fit2, adjust="BH")
Am I right ?
Again, thanks a lot for your help,
Florence
