[BioC] limma Common Reference Analysis.

Gordon Smyth smyth at wehi.EDU.AU
Sat May 19 03:50:34 CEST 2007

Well, probably yes, if you followed the code from Section 8.4 "Two 
Groups: Common Reference" without modification.

Section 7.3 "Common Reference Designs" gives an example where the 
reference was labelled with Cy3 or Cy5 in different cases. If you 
used the modelMatrix() function the orientation of the samples will 
be taken into account automatically.

Best wishes

>Date: Thu, 17 May 2007 10:43:00 -0400
>From: "Khan, Sohail" <khan at cshl.edu>
>Subject: [BioC] limma Common Reference Analysis.
>To: <Bioconductor at stat.math.ethz.ch>
>I have analyzed a common reference with 4 groups experimental design 
>by applying the same method described in the limma user's guide for 
>common reference 2 groups.
>However,  all of the reference samples were labeled with 
>Cy5.  Therefore I multiplied all of the resulted ratios with -1.
>I this a valid approach?
>Sohail Khan
>Scientific Programmer
>Genome Research Center
>500 Sunnyside Boulevard
>Woodbury, NY 11797

More information about the Bioconductor mailing list