[BioC] Limma Design Matrix: How many levels can each factor have?

[James W. MacDonald]

Xia Han wrote:
> Hi:
> 
> I have a drug treament experiments with 3 factors (cell lines, 
> compounds, and time course). There are 3 cell lines, 3 compounds (DMSO 
> control, compounds 1 and 2), and 2 time points (4hr and 24hr). When I 
> try to build a design matrix by something like this:
> design <- model.matrix(~cell.lines*compound*time)
> I only get 2 cell lines, 2 compounds (DMSO and compound1) and 2 time 
> points.
> If I use something like this, which is really what I want to compare:
> design <- model.matrix(~cell.lines + compound + compound*time)
> I still only get 2 cell lines and 2 compounds in the design matrix.
> 
> Anything I did wrong? Please help.

You didn't do anything wrong, but it appears you don't understand your 
design matrix. This is a big problem, since it is fundamental to 
understanding what you are doing. Have you looked at the limma user's 
guide? There are many examples including time course experiments that 
are explained in some detail.

This is a fairly complex experiment, and you may be better served if you 
can find a local statistician who can help you analyze these data.

Best,

Jim


> 
> Thanks,
> 
> Xia
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor


-- 
James W. MacDonald, M.S.
Biostatistician
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623