[BioC] Microarray analysis question

Naomi Altman naomi at stat.psu.edu
Wed Jun 18 03:17:55 CEST 2008


Dear Cecilia,
Please remember to cc the list even if you contact someone individually.

Limma and MAANOVA can both handle loop designs and multi-way ANOVA.  The 
way I do this is to use the single-channel normalization as in the Limma 
manual.  For limma, I then proceed with the single-channel analysis as in 
the manual.  "array" is the block.  I am not that fond of setting up the 
contrasts for high order interactions, but it can be done.

I have not used MAANOVA, so I am less than an expert.  However, I would use 
the single-channel normalization and then use RG.MA to convert the 
normalized data back to normalized R and normalized G.  Then I would 
proceed as if each channel were a normalized Affy array, except that there 
would be a random effect for array.

I hope this helps.

Naomi

At 07:16 PM 6/17/2008, you wrote:
>Dr Altman
>
>I was wondering whether you know of any microarray analysis program that 
>can handle loop designs, and three-term interactions? Maanova cannot do 
>what I need. Basically I have Cultivar, Time and Treatment.  I've posted 
>many questions on the message boards but got no useful info. I've used 
>Limma before for other experiments with no problems, but I know that it is 
>not appropriate here.
>
>Any advise wold be much appreciated.
>
>Kind regards
>Cecilia
>
>Cecilia McGregor, Ph.D.
>Department of Entomology
>404 Life Sciences Building
>Louisiana State University
>Baton Rouge
>LA, 70803
>USA
>
>Office Phone: (225) 578-0518
>Lab Phone: (225) 578-0921
>Mobile Phone: (225) 763-9138
>Fax: (225) 578-1643

Naomi S. Altman                                814-865-3791 (voice)
Associate Professor
Dept. of Statistics                              814-863-7114 (fax)
Penn State University                         814-865-1348 (Statistics)
University Park, PA 16802-2111



More information about the Bioconductor mailing list