I do not think that there is any technical limitation that would
you from using it with "non log-transformed data".
However, note that if working on microarray data homoscedasity might
very far from respected and this place the post-fit part of the
pipeline (eBayes or treat) on unknown ground.
On 2011-07-08 11:13, ??? wrote:
> I want to ask some questions about limma package .
> if my data is not log-transform, whether also could use "lmFit"
> [[alternative HTML version deleted]]
> Bioconductor mailing list
> Bioconductor at r-project.org
> Search the archives: