Dear all, I would like to here your comments and suggestions regarding my analysis described below. Im working with genomewide screening of CpG methylation with agilent two color CpG arrays with a common reference design. I have got the data of the future extraction and did loess normalisation, applied limma to check for the differentially regulated genes. In one of the comparision i have superficial and invasive tumors, surprisingly i saw zero differentially regulated genes which is quiet unrealistic. Here is the R session: > design1<-cbind(SM=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0), I=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1)) > colnames(design1)<- c("SM", "I") > design1 > dim(dataset1) [1] 237220 34 > fit<- lmFit(dataset1, design1) > ngenes <- nrow (dataset1) > cont.matrix<- makeContrasts(SMvsI=SM-I, levels=design1) > cont.matrix Contrasts Levels SMvsI SM 1 I -1 > fit2<-contrasts.fit(fit, cont.matrix) > fit2 <-eBayes(fit2) > topTable(fit2, adjust="fdr") > SMvsI <- topTable(fit2, number=ngenes, adjust="fdr") > results <- decideTests(fit2,adjust.method="fdr",p.value=0.05) > summary(results) SMvsI -1 0 0 237202 1 0 Best regards, -- Raju Kandimalla, PhD student Erasmus MC Department of Pathology JNI,Room H Be-302 Dr. Molewaterplein 50 3015 GE Rotterdam-NL phone: +3110-7043093 fax: +3110-7044762 r.kandimalla at erasmusmc.nl http://www.erasmusmc.nl/pathologie

Dear all, I m not sure about what kind of normalization process can be used for methylation status analysis because of the potential distribution of these data... I would be curious to know what people from this group think about this point and if it could explain part of these results below... Thanks Philippe Guardiola -----E-mail d'origine----- De : r.kandimalla <r.kandimalla at="" erasmusmc.nl=""> A : bioconductor at stat.math.ethz.ch Envoy? le : Lundi, 10 Novembre 2008 16:42 Sujet : Re: [BioC] Analysis of differentially regulated genes Dear all,? ? I would like to here your comments and suggestions regarding my analysis described below.? Im working with genomewide screening of CpG methylation with agilent two color CpG arrays with a common reference design.? I have got the data of the future extraction and did loess normalisation, applied limma to check for the differentially regulated genes. In one of the comparision i have superficial and invasive tumors,? surprisingly i saw zero differentially regulated genes which is quiet unrealistic.? Here is the R session:? > design1<-cbind(SM=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0, 0,0 ,0,0,0,0,0,0,0), I=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1)) ? > colnames(design1)<- c("SM", "I")? > design1? > dim(dataset1)? [1] 237220 34? > fit<- lmFit(dataset1, design1)? > ngenes <- nrow (dataset1)? > cont.matrix<- makeContrasts(SMvsI=SM-I, levels=design1)? > cont.matrix? ? Contrasts? Levels SMvsI? ? SM 1? ? I -1? > fit2<-contrasts.fit(fit, cont.matrix)? > fit2 <-eBayes(fit2)? > topTable(fit2, adjust="fdr")? > SMvsI <- topTable(fit2, number=ngenes, adjust="fdr")? > results <- decideTests(fit2,adjust.method="fdr",p.value=0.05)? > summary(results)? ? SMvsI? -1 0? 0 237202? 1 0? ? Best regards,? ? -- Raju Kandimalla, PhD student? Erasmus MC? Department of Pathology? JNI,Room H Be-302? Dr. Molewaterplein 50? 3015 GE Rotterdam-NL? phone: +3110-7043093? fax: +3110-7044762? r.kandimalla at erasmusmc.nl? http://www.erasmusmc.nl/pathologie? ? _______________________________________________? Bioconductor mailing list? Bioconductor at stat.math.ethz.ch? https://stat.ethz.ch/mailman/listinfo/bioconductor? Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor?

On Mon, Nov 10, 2008 at 10:42 AM, r.kandimalla <r.kandimalla@erasmusmc.nl>wrote: > Dear all, > > I would like to here your comments and suggestions regarding my analysis > described below. > Im working with genomewide screening of CpG methylation with agilent two > color CpG arrays with a common reference design. > I have got the data of the future extraction and did loess normalisation, > applied limma to check for the differentially regulated genes. In one of the > comparision i have superficial and invasive tumors, > surprisingly i saw zero differentially regulated genes which is quiet > unrealistic. Hi, Raju. I didn't see any major problems with the code below, but someone with a better eye might spot something. What makes you say that no differentially-regulated genes is "quite unrealistic"? It is (unfortunately) quite possible for that to be the case. On the other hand, have you looked at basic quality control issues? Sample relationships using unsupervised methods? You mention that you are using methylation arrays? What have you done for normalization? These are all important questions that you should address thoroughly before you can make final conclusions about your results. Sean > > Here is the R session: > > > design1<-cbind(SM=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0), > I=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1)) > > colnames(design1)<- c("SM", "I") > > design1 > > dim(dataset1) > [1] 237220 34 > > fit<- lmFit(dataset1, design1) > > ngenes <- nrow (dataset1) > > cont.matrix<- makeContrasts(SMvsI=SM-I, levels=design1) > > cont.matrix > Contrasts > Levels SMvsI > SM 1 > I -1 > > fit2<-contrasts.fit(fit, cont.matrix) > > fit2 <-eBayes(fit2) > > topTable(fit2, adjust="fdr") > > SMvsI <- topTable(fit2, number=ngenes, adjust="fdr") > > results <- decideTests(fit2,adjust.method="fdr",p.value=0.05) > > summary(results) > SMvsI > -1 0 > 0 237202 > 1 0 > > Best regards, > > -- > Raju Kandimalla, PhD student > Erasmus MC > Department of Pathology > JNI,Room H Be-302 > Dr. Molewaterplein 50 > 3015 GE Rotterdam-NL > phone: +3110-7043093 > fax: +3110-7044762 > r.kandimalla@erasmusmc.nl > http://www.erasmusmc.nl/pathologie > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]