[BioC] Is a set of four arrays suitable for RMA analysis?

[Rafael A. Irizarry]

i use RMA if i have 2 or more arrays. the roc curves in this paper

http://nar.oupjournals.org/cgi/content/full/31/4/e15?ijkey=EAz2cYYbEWQrE&keytype=ref&siteid=nar

are based on 2 array comparisons.



On Tue, 17 Jun 2003, Wei, Wenbin 
wrote:

> Hello, 
> 
>  
> 
> I have a set of four Affymetrix HG-Focus arrays (including two controls and
> two treatments) to be analyzed. I would like to know whether a set of four
> arrays is suitable for RMA analysis.  In Irizarry et al.
> (http://www.biostat.jhsph.edu/~ririzarr/papers/affy1.pdf
> <http://www.biostat.jhsph.edu/~ririzarr/papers/affy1.pdf> ), the minimum
> number of arrays in a set subjected to RMA analysis was five. 
> 
>  
> 
> In dChip manual, it is mentioned that "For better model fitting and outlier
> detection, the number of arrays in a dChip group is desired to be more than
> 5. " (http://biosun1.harvard.edu/complab/dchip/prepare%20data.htm
> <http://biosun1.harvard.edu/complab/dchip/prepare%20data.htm> ). 
> 
>  
> 
> Has anybody compared various methods available in the Affy package for the
> analysis of a set of four arrays? Which is the best?
> 
>  
> 
> Wenbin Wei
> 
> Division of Cancer Studies
> 
> Birmingham University
> 
> Birmingham, B15 2TT
> 
> UK
> 
> 
> 	[[alternate HTML version deleted]]
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
>