[BioC] Analysis of differentially regulated genes

[r.kandimalla]

Dear all,

I would like to here your comments and suggestions regarding my analysis 
described below.
Im working with genomewide screening of CpG methylation with agilent two 
color CpG arrays with a common reference design.
I have got the data of the future extraction and did loess 
normalisation, applied limma to check for the differentially regulated 
genes. In one of the comparision i have superficial and invasive tumors,
surprisingly i saw zero differentially regulated genes which is quiet 
unrealistic.
Here is the R session:
 > 
design1<-cbind(SM=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0), 
I=c(0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1))
 > colnames(design1)<- c("SM", "I")
 > design1
 > dim(dataset1)
[1] 237220     34
 > fit<- lmFit(dataset1, design1)
 > ngenes <- nrow (dataset1)
 > cont.matrix<- makeContrasts(SMvsI=SM-I, levels=design1)
 > cont.matrix
       Contrasts
Levels SMvsI
     SM     1
     I     -1
 > fit2<-contrasts.fit(fit, cont.matrix)
 > fit2 <-eBayes(fit2)
 > topTable(fit2, adjust="fdr")
 > SMvsI <- topTable(fit2, number=ngenes, adjust="fdr")
 > results <- decideTests(fit2,adjust.method="fdr",p.value=0.05)
 > summary(results)
     SMvsI
-1      0
0  237202
1       0

Best regards,

-- 
Raju Kandimalla, PhD student
Erasmus MC
Department of Pathology
JNI,Room H Be-302
Dr. Molewaterplein 50
3015 GE Rotterdam-NL
phone: +3110-7043093
fax: +3110-7044762
r.kandimalla at erasmusmc.nl
http://www.erasmusmc.nl/pathologie