[BioC] Construct design matrix for qPCR data using Limma package

James W. MacDonald jmacdon at uw.edu
Wed Nov 27 16:40:01 CET 2013


Hi Adam,

That's one way of doing it. Well, except for the last line where you 
substitute in an apparently nonexistent design2 object. Another way 
would be

fact <- factor(rep(c("Control","Case"), c(38,54)), levels = 
c("Control","Case"))
design <- model.matrix(~ 0 + fact)
colnames(design) <- gsub("fact", "", colnames(design))

Best,

Jim



On Wednesday, November 27, 2013 10:29:13 AM, Hongda Chen wrote:
> Dear R-users,
>
> I am facing a problem in constructing design matrix. My data is qPCR data,
> analysing 92  protein markers in 92 samples (two groups). I would like to
> use Limma package to identify differently expressed genes between two
> groups. The first 38 samples were controls, and the other 54 samples were
> cases. So how should I construct the design matrix.
>
> I have tried the following code. Is this right?
> design<-matrix(rep(1,184),nrow=92,byrow=TRUE)
> design[39:92,1]=0
> design[1:38,2]=0
> colnames(design2)<-c("control","case")
>
> Could anyone help me figure out it? Thank you.
>
> Adam
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

--
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099



More information about the Bioconductor mailing list