I think Limma (and many other microarray methods) considers log- transformed data because then it is easy to assume that the data is normally distributed (after log-transformation) for computing moderated t-test (Limma statistics). For a non log-transformed data, I am not sure if the method is valid if the data does not have normal (like) distribution. Say for instance if you have Poisson counts in the dataset, the data wont be normally distributed and hence Limma wont be fair choice in that case. Thanks, ~Nikhil On Fri, Jul 8, 2011 at 5:37 AM, Laurent Gautier <laurent@cbs.dtu.dk> wrote: > I do not think that there is any technical limitation that would prevent > you from using it with "non log-transformed data". > However, note that if working on microarray data homoscedasity might be > very far from respected and this place the post-fit part of the analysis > pipeline (eBayes or treat) on unknown ground. > > > L. > > > > > > On 2011-07-08 11:13, 張詩婷 wrote: > >> Hi, >> I want to ask some questions about limma package . >> if my data is not log-transform, whether also could use "lmFit" function? >> thanks! >> >> >> Chang >> >> [[alternative HTML version deleted]] >> >> ______________________________**_________________ >> Bioconductor mailing list >> Bioconductor@r-project.org >> https://stat.ethz.ch/mailman/**listinfo/bioconductor<https: stat.e="" thz.ch="" mailman="" listinfo="" bioconductor=""> >> Search the archives: http://news.gmane.org/gmane.** >> science.biology.informatics.**conductor<http: news.gmane.org="" gmane="" .science.biology.informatics.conductor=""> >> > > ______________________________**_________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/**listinfo/bioconductor<https: stat.et="" hz.ch="" mailman="" listinfo="" bioconductor=""> > Search the archives: http://news.gmane.org/gmane.** > science.biology.informatics.**conductor<http: news.gmane.org="" gmane.="" science.biology.informatics.conductor=""> > [[alternative HTML version deleted]]