Hi,

I have a question about my agilent microarray data.

It includes before & after infection data of the same individuals and also replicates of non-infected other individuals as a kind of control to these infected ones . There are 15 individuals and thus 30 microarray files. I am using "limma" package to read and normalize them. However, i have doubts if i should use the codes below to identifiy genes differently expressed, because when i cluster the datasets, the replicates do not group together:

RG_2= backgroundCorrect(z, method="minimum")

MA.exp= normalizeBetweenArrays(RG_2)

MA.exp_log= log(MA.exp$E)

SibShip = factor(targets_paired$SibShip)

Treat = factor(targets_paired$Treatment, levels=c("Not Inf.","Infected"))

design = model.matrix(~SibShip+Treat)

fit = lmFit(MA.exp_log, design)

fit = eBayes(fit)

topTable(fit, coef="TreatInfected")

In this case, should I analyze the data by just grouping as Infected & Non-infected?

Thank you in advance.

One more thing, Why are you doing the MA.exp_log step ?

Have you read that in the limma reference document?. I don't think you need to do it, and if a log is calculated, it should be in base 2