Hello, I create a matrix with my intensities datas and read it in R. I have a doubt: I transform my intensities to log2 before use normalize.quantiles or this is not necessary? If the transformation is not necessary, I have another two questions: I already use normalize.quantiles on the intensities matrix with out any transformation and I use limma to get differentially expressed genes. lev <- c("ConT1","ConT2","ConT3","TraT1","TraT2","TraT3") f <- factor(Tol.targets$Target, levels=lev) design <- model.matrix(~0+f) colnames(design) <- lev Tol.matrix <- read.matrix("Tolerante_m.txt") Tol.matrix.norm <- normalize.quantiles(Tol.matrix) Tol.matrix.fit <- lmFit(Tol.matrix.norm,design,ndups=2,spacing=1) cont.Tra <- makeContrasts("TraT2-TraT1","TraT3-TraT2","TraT3-TraT1",levels=design) Tol.matrix.fit2 <- contrasts.fit(Tol.matrix.fit, cont.Tra) Tol.matrix.fit3<-eBayes(Tol.matrix.fit2) UniqueGeneNames<-read.table("UniqueGeneNames.txt",sep="\t",header = TRUE, colClasses = "character") Tol.matrix.fit3$genes <- UniqueGeneNames Tol.selected.Tra <- p.adjust(Tol.matrix.fit3$F.p.value, method="fdr") < 0.05 Tol.DE.Tra.Fpvalue <- Tol.matrix.fit3$genes[Tol.selected.Tra,] write.table(Tol.DE.Tra.Fpvalue,"Tol.DE.Tra.Fpvalue.txt",sep="\t") Tol.DE.TraT2TraT1 <- topTable(Tol.matrix.fit3,number=100,coef=1,adjust="fdr") Tol.DE.TraT3TraT2 <- topTable(Tol.matrix.fit3,number=100,coef=2,adjust="fdr") Tol.DE.TraT3TraT1 <- topTable(Tol.matrix.fit3,number=100,coef=3,adjust="fdr") In the topTable result I get a bigger *M* value ten times great that the original value in intensities raw matrix. This is correct or this is because the intensities is not log transformed before? In the limma users guide I get the information about M-value: "The M-value (M) is the log2-fold change, or sometimes the log2-expression level, for that gene." In my situation I get the log2-expression level (I am not sure). Fold change is not estimable im my situation? Thanks Marcelo *I do not have great knowledge of the English language. I started to study English, more or less, one year ago. I write my messages and use, in some cases, a translator on line to assist me. I send my excuses in advanced if my messages are confused, or offensive or unpolite.* Rafael A. Irizarry escreveu: > i used normalize.quantiles on data from a similar platform and was > satisfied with the results. give it try.. > > -r > > > > On Tue, 4 Jan 2005, Marcelo Luiz de Laia wrote: > >> Hi Bioconductors, >> >> A happy new year to all! >> >> We have a raw data from naylon membrane (cDNA) and we would like to >> read and normalize it on affy package. >> >> Anyone do this (read and normalize) in affy package? Is it possible? >> Is a good idea? >> >> Thanks a lot >> >> Marcelo >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor@stat.math.ethz.ch >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> > -- No virus found in this outgoing message. Checked by AVG Anti-Virus.