Hi there, I'm trying to analyze the data of 8 dual-color microarray hybridizations, but I' a bit stuck with the design matrix. I tried looking into the previous emails from the list but could not quite find the answer. It'd be great if you could help out. The experimental design is fairly straightforward in priciple: this is a time course with only two time points. For each time point I have duplicate samples and dye swaps for each sample. Something like this below: 1 Wt1 Mu1 Time point 1 2 Mu1 Wt1 Time point 1 3 Wt1 Mu1 Time point 2 4 Mu1 Wt1 Time point 2 5 Wt2 Mu2 Time point 1 6 Mu2 Wt2 Time point 1 7 Wt2 Mu2 Time point 2 8 Mu2 Wt2 Time point 2 Apart from the time effect, I would actually like to test if the "replicates" are actual replicates, since the biological data seem to indicate they are not quite so similar. If I assume that the replicates are actual replicates, how would the design matrix look like? and if I want to see if the replicates are behaving differently, do I have to design two different matrices and analyse each set of data independently? Sorry if the questions are too naive! Thanks in advance Jose Garcia-Bernardo University of Cambridge Downing Street CB2 3DY Cambridge UK

The package limma will allow you to address these issues. What's more, its user's guide is a wonderful overview of the analysis of microarray data accessible to non-mathematicians (like me). You can obtain the user's guide by installing the package, or directly here (for example): http://cmmg.biosci.wayne.edu/Education/mbg8680/limma_users_guide.pdf I hope that this journey leads you to the conclusion that your replicates are replicates, otherwise it's back to the laboratory as it is difficult to perform statistical analysis with N=1. Enjoy, Eric -----Original Message----- From: bioconductor-bounces@stat.math.ethz.ch on behalf of jg295@mole.bio.cam.ac.uk Sent: Wed 12/14/2005 10:12 AM To: bioconductor at stat.math.ethz.ch Subject: [BioC] (no subject) Hi there, I'm trying to analyze the data of 8 dual-color microarray hybridizations, but I' a bit stuck with the design matrix. I tried looking into the previous emails from the list but could not quite find the answer. It'd be great if you could help out. The experimental design is fairly straightforward in priciple: this is a time course with only two time points. For each time point I have duplicate samples and dye swaps for each sample. Something like this below: 1 Wt1 Mu1 Time point 1 2 Mu1 Wt1 Time point 1 3 Wt1 Mu1 Time point 2 4 Mu1 Wt1 Time point 2 5 Wt2 Mu2 Time point 1 6 Mu2 Wt2 Time point 1 7 Wt2 Mu2 Time point 2 8 Mu2 Wt2 Time point 2 Apart from the time effect, I would actually like to test if the "replicates" are actual replicates, since the biological data seem to indicate they are not quite so similar. If I assume that the replicates are actual replicates, how would the design matrix look like? and if I want to see if the replicates are behaving differently, do I have to design two different matrices and analyse each set of data independently? Sorry if the questions are too naive! Thanks in advance Jose Garcia-Bernardo University of Cambridge Downing Street CB2 3DY Cambridge UK _______________________________________________ Bioconductor mailing list Bioconductor at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor This email message, including any attachments, is for the so...{{dropped}}