Hi Sean, Thanks for your advice - that worked. I ended up needing to convert my RGList object into an MAList object anyway so that I could apply the avereps function (I couldn't get it to work with the RGList - it seemed to lose its format). Although I expect the new EList and EListRaw objects will soon make this "hack" redundant, for anyone interested in using avereps on single channel data at the moment, you simply do: E <- new("MAList", list(targets=RG$targets, genes=RG$genes, source=RG$source, M=RG$Gb, A=RG$G)) E.avg <- avereps(E,ID=E$genes$ProbeName) avereps is particularly important for doing any analysis with custom agilent arrays designed with Earray since it will (optionally) fill all remaining spots with replicates. Since I've received a few emails after making my post on single channel analysis for more information, I've put my steps up on our wiki here: http://jsm-research.imb.uq.edu.au/groupwiki/index.php?title=Single_cha nnel_analysis_of_Agilent_microarray_data_with_Limma Any comments for improvement are of course appreciated. Best regards, Marcel. Sean Davis wrote: > > > On Thu, Jun 25, 2009 at 1:56 AM, Marcel Dinger <m.dinger@uq.edu.au> <mailto:m.dinger@uq.edu.au>> wrote: > > Hi, > > I've been trying to use LIMMA to analyze a set of single channel data > created with Agilent 105k arrays. I'm still learning R, so > apologies if > there is actually a very simple programmatic solution to my problem - > but despite a lot of searching and reading I haven't been able to > figure > it out. > > By reading through various posts on this list (and of course the limma > user guide) I have managed to get most of the way there, but am > struggling with getting my single channel dataset into lmFit. > > This is what I'm doing: > library(limma) > targets <- readTargets("neurogenesis.txt") > RG <- read.maimages(targets, path="neurogenesis", source="agilent") > RG <- backgroundCorrect(RG,method="minimum") > RG$R <- NULL > RG$Rb <- NULL > RG$G <- normalizeBetweenArrays(RG$G, method="quantile") > RG$G <- log2(RG$G) > f <- factor(targets$Condition, levels = unique(targets$Condition)) > design <- model.matrix(~0 + f) > colnames(design) <- levels(f) > fit <- lmFit(RG$G,design) > contrast.matrix <- > makeContrasts("Condition1-Condition2",levels=design) > fit2 <- contrasts.fit(fit,contrast.matrix) > fit2 <- eBayes(fit2) > topTable(fit2, adjust="BH", coef="Condition1-Condition2") > > This works, but the problem is that because I cannot put RG (as an > RGList object) into lmFit, I lose the Probe Names in topTable. I > think I > could theoretically match them up again with my original list, but > this > seems very clumsy. > > > Hi, Marcel. > > It is always a good idea to read the help pages for any new functions > you are using, particularly those that appear to be causing problems. > The topTable() function includes an argument for specifying a gene list. > > Sean > > > > One solution appears to be to shift my RG data into an ExpressionSet > object, but I'm not sure how to do this in R. > > I see that LIMMA has recently added single channel support through the > EListRaw and EList objects, but I didn't have any luck with these > either > as there (currently) doesn't appear to be a way to go from an EListRaw > to an EList object? I get the impression that in the future, I will > simply be able to go: > > E <- read.maimages(targets, path="neurogenesis", source="agilent", > channels=1) > ENorm <- normalizeBetweenArrays(E, method="quantile") > ... > fit <- lmFit(ENorm,design) > etc > > But unfortunately normalizeBetweenArrays does not yet take EListRaw as > input. > > So my question is - is there a way I can maintain my probe annotations > at the lmFit step? > or more specifically - how do I shift the data from an RGList object > into an ExpressionSet object? > or... how can I shift from an EListRaw object to an EList object? > > Thanks in advance for any help or advice. > > Regards, > Marcel. > > -- > Marcel Dinger > <http: jsm-="" research.imb.uq.edu.au="" groupwiki="" index.php?title="Marcel_Dinger">, > Ph.D., > Senior Research Officer, > Division for Genomics and Computational Biology, > Institute for Molecular Bioscience, > University of Queensland, > Brisbane, Australia. > > <mailto:m.dinger@uq.edu.au <mailto:m.dinger@uq.edu.au="">> > > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch <mailto: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]]