Hi, I am using limma for a variety of projects. It all seems pretty straightforward for projects where I am not as concerned with the interaction of the independent variables, but I am a bit confused on how to create the contrasts so that I can focus the analysis on the interactions between the independent variables. I have read the users' guides and vignettes and the help pages, but haven't found a good explanation for setting up a more complex model.matrix and contrasts and how to make proper use of all the information embedded in the limma object. Does anyone have a recommendation for a more comprehensive discussion and tutorial? Thanks in advance, Sue Huse

Hi Sue, On 7/10/2013 10:34 AM, Susan Huse wrote: > Hi, > > I am using limma for a variety of projects. It all seems pretty straightforward for projects where I am not as concerned with the interaction of the independent variables, > but I am a bit confused on how to create the contrasts so that I can focus the analysis on the interactions between the independent variables. I have read the users' guides and vignettes and the help pages, but haven't found a good explanation for setting up a more complex model.matrix and contrasts and how to make proper use of all the information embedded in the limma object. Does anyone have a recommendation for a more comprehensive discussion and tutorial? It depends on what exactly you are looking for. Note that limma is just leveraging off the existing linear modeling infrastructure in R, so if you are looking for a more comprehensive discussion of how to fit linear models in R, then you could look at MASS (Modern Applied Statistics with S), which has a pretty in depth discussion on fitting various linear models. Julian Faraway wrote a pretty nice book that he has contributed to CRAN (http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf), that covers linear modeling with R in even more depth than MASS. So if you are just looking for general information about how to fit a model using R, those two are pretty good resources (and the Faraway book is that much better being free). If I am missing your question, perhaps you can give an example of an experiment you are trying to analyze, the goals, and ideally what you have tried (and possibly failed) to do the analysis. Best, Jim > > Thanks in advance, > Sue Huse > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor -- James W. MacDonald, M.S. Biostatistician University of Washington Environmental and Occupational Health Sciences 4225 Roosevelt Way NE, # 100 Seattle WA 98105-6099