[BioC] Help: Problems with LmFit

[julin at aecom.yu.edu]

Dear List,
I am working with six 27K human cDNA chips.Three of them are for Average
Primary Tumor Cells and another three are for Invasive cells.In all chips,
the reference in on the red channel, and the tumor samples are on the
green channel. Here is my target file:

ID	FileName	Cy3	Cy5
131	A19.gpr	APTC	Ref
1015	I24.gpr	INV	Ref
416	A25.gpr	APTC	Ref
611	A26.gpr	APTC	Ref
1113	I27.gpr	INV	Ref
10002	I28.gpr	INV	Ref

We would like to find ratios of gene expression of INVasive vs. APTC.I
strictly follow the limma user guide commands in my analysis but I got a
problem with lmFit.Here are my R codes:

> Group<-factor(targets$Cy3,levels=c("APTC","INV"))
> design<- model.matrix(~0+Group)
> colnames(design)<-c("APTC","INV")
> design
  APTC INV
1    1   0
2    0   1
3    1   0
4    1   0
5    0   1
6    0   1
attr(,"assign")
[1] 1 1
attr(,"contrasts")
attr(,"contrasts")$Group
[1] "contr.treatment"

> SNR<- function(x,threshold=1){
+    okred<-x[,"SNR 635"]>=threshold
+    okgreen<-x[,"SNR 532"]>=threshold
+    okflag<- x[,"Flags"]==0
+    as.numeric(okgreen & okred & okflag)
+    }
> RG<-read.maimages(targets$FileName,source="genepix",wt.fun=SNR)
Read A19.gpr
Read I24.gpr
Read A25.gpr
Read A26.gpr
Read I27.gpr
Read I28.gpr
> RG<-backgroundCorrect(RG,method="normexp",offset=50)
Green channel
Corrected array 1
Corrected array 2
Corrected array 3
Corrected array 4
Corrected array 5
Corrected array 6
Red channel
Corrected array 1
Corrected array 2
Corrected array 3
Corrected array 4
Corrected array 5
Corrected array 6
> MA<-normalizeWithinArrays(RG,weights=RG$weights)
> fit<-lmFit(MA,design)
Warning message:
In lmFit(MA2, design) :
  Some coefficients not estimable: coefficient interpretation may vary.

I don't really understand this warning message. Is there anything wrong
with my design matrix?Thank you in advance for any suggestions or
comments!

Best Regards,

Cathy
Albert Einstein Cancer Center