[BioC] limma question

hdvi hdvi at well.ox.ac.uk
Fri Feb 10 10:12:08 CET 2006


On 10 Feb 2006, at 00:50, Michael Stapelberg wrote:

> Hi.
>
> I am having a few problems with LIMMA that never happened before. The
> first problem is trying to access and download my .gpr files into R.
> Previously I would change the working directory to a nominated file
> that contained all my .gpr files and other related bits. Here is the
> whole script that I have used in the past.
>
>
> Could you please advise me why perhaps I cannot download my gpr files ?
>
> controls<-read.table("blank and ratio.txt")
> morecontrols<-read.table("utility and calib.txt")
> morecontrols2<-read.table("neg.txt")
> y<-as.vector(controls[,1])
> z<-as.vector(morecontrols[,1])
> z1<-as.vector(morecontrols2[,1])
> xx<-c(z,y,z1)
> files<-dir(pattern="*\\.gpr") ??? does not work for some reason
>
Your problem here is that the meaning of star is that the preceding 
item will be matched zero or more times. You don't have any preceding 
items...

\Heidi


> RG<-read.maimages(files,source="genepix")
> RG$genes<-readGAL("mouse 22k gal annotated.gal")
> RG$printer<-getLayout(RG$genes)
> RGb<-backgroundCorrect(RG,method="none")
> RG1<-filter.controls.to.NA(RGb,xx)
> MA<-normalizeWithinArrays(RG1,RG$printer)
> UVscale<-normalizeBetweenArrays(MA)
> UVscaleBALB<-cbind(UVscale[,1],UVscale[,2]UVscale[,3]UVscale[,4]
> UVscale[,5]UVscale[,6])
> UVscaleC57BL<-cbind(UVscale[,7],UVscale[,8],UVscale[,9],UVscale[,
> 10],UVscale[,11],UVscale[,12)
> BALBcA.UVscale <- UVscaleBALB$A
> BALBcM.UVscale <- UVscaleBALB$M
> C57BLA.UVscale <- UVscaleC57BL$A
> C57BLM.UVscale <- UVscaleC57BL$M
> BALBca<-is.na(BALBcM.UVscale)
> BALBcb<-apply(BALBa,1,sum)
> BALBczero<-which(BALBcb==0)
> BALBcx.lmFit <- lmFit(BALBcM.UVscale[BALBczero,],c(1,-1,1,-1,1,1))
> BALBcx.lmFit.Bayes <- eBayes(BALBcx.lmFit,proportion=0.10)
> plot(BALBcx.lmFit.Bayes$coef,BALBcx.lmFit.Bayes$lods)
> topTable(BALBcx.lmFit.Bayes,n=100)
> RG$genes$ID[as.numeric(rownames(topTable(BALBcx.lmFit.Bayes,n=100)))]
> M.UVscale[as.numeric(rownames(topTable(BALBcx.lmFit.Bayes,n=100)))]
>
> MA$M[as.numeric(rownames(topTable(BALBcx.lmFit.Bayes,n=100)))] ???
> could this give me replicate M values in the toptable ranking order ?
>
> C57BLa<-is.na(C57BLM.UVscale)
> C57BLb<-apply(C57BL,1,sum)
> C57BLzero<-which(C57BLb==0)
> C57BLx.lmFit <- lmFit(C57BLM.UVscale[C57BLzero,],c(1,-1,1,-1,1,-1))
> C57BLx.lmFit.Bayes <- eBayes(C57BLx.lmFit,proportion=0.10)
> plot(C57BLx.lmFit.Bayes$coef,C57BLx.lmFit.Bayes$lods)
> topTable(C57BLx.lmFit.Bayes,n=100)
> RG$genes$ID[as.numeric(rownames(topTable(C57BLx.lmFit.Bayes,n=100)))]
> M.UVscale[as.numeric(rownames(topTable(C57BLx.lmFit.Bayes,n=100)))]
> MA$M[as.numeric(rownames(topTable(C57BLx.lmFit.Bayes,n=100)))]
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>



More information about the Bioconductor mailing list