[BioC] Yet another nested design in limma

[Paolo Innocenti]

Hi all,

I have an experimental design for a Affy experiment that looks like this:

	Phen	Line	Sex	Biol.Rep.
File1	H	1	M	1
File2	H	1	M	2
File3	H	1	F	1
File4	H	1	F	2
File5	H	2	M	1
File6	H	2	M	2
File7	H	2	F	1
File8	H	2	F	2
File9	L	3	M	1
File10	L	3	M	2
File11	L	3	F	1
File12	L	3	F	2
File13	L	4	M	1
File14	L	4	M	2
File15	L	4	F	1
File16	L	4	F	2


This appears to be a slightly more complicated situation than the one 
proposed in the section 8.7 of the limma users guide (p.45) or by Jenny 
on this post:

https://stat.ethz.ch/pipermail/bioconductor/2006-February/011965.html

In particular, I am intersted in
- Effect of "sex" (M vs F)
- Interaction between "sex" and "phenotype ("line" nested)
- Effect of "phenotype" in males
- Effect of "phenotype" in females

Line should be nested in phenotype, because they are random "strains" 
that happened to end up in phenotype H or L.

Can I design this in limma? Is there a source of information about how 
to handle with this? In particular, can I design a single model matrix 
and then choose the contrasts I am interested in?

Any help is much appreciated,
paolo


-- 
Paolo Innocenti
Department of Animal Ecology, EBC
Uppsala University
Norbyvägen 18D
75236 Uppsala, Sweden