[BioC] Limma with Imagene - reading control status - NO html

Gordon Smyth smyth at wehi.edu.au
Sat Dec 6 01:41:20 MET 2003


The problem is that your gene list doesn't have columns called "ID" and 
"Name". Type

names(RG$genes)

With Imagene data you will probably find a column called "Gene ID". The 
regular expression columns in your SpotTypes file must match columns in 
your gene list.

I have made some updates to limma which should solve your problems. You 
will need to update to limma 1.3.6 available now from 
http://bioinf.wehi.edu.au/limma or very soon from the Bioconductor 
development packages area.

Assuming that your gene list does have a column "Gene ID" (you need to 
check this), you will want a SpotTypes file something like:

SpotType        Gene ID col     cex
Gene    *       black   0.2
Spot    Spot    brown   1
COT     COT     brown   1
Dros    Dros*   brown   1
Cy3     Cy3     green   1
Cy5     Cy5     red     1

You should be able to type:

types <- readSpotTypes()
RG$genes$Status <- controlStatus(types,RG)
plotMA(RG)

to get a color-coded MA-plot. I have used a similar example successfully 
myself with Imagene data.

Gordon

At 03:05 AM 5/12/2003, Plantinga, AED wrote:
>Dear all,
>I am experiencing difficulties with setting the spot types after reading in
>Imagene files:
> > targets <- readTargets() # read in Targets.txt from working dir
> > targets
>SlideNumber FileNameCy3 FileNameCy5 Cy3 Cy5
>1 1 0_Cy3_Slide1.txt 1_Cy5_Slide1.txt Test1 Ref1
>2 2 0_Cy3_Slide2.txt 1_Cy5_Slide2.txt Test2 Ref2
> > files <- cbind(targets$FileNameCy3, targets$FileNameCy5) # This step is
>necessary for imagene files!
> > files
>[,1] [,2]
>[1,] "0_Cy3_Slide1.txt" "1_Cy5_Slide1.txt"
>[2,] "0_Cy3_Slide2.txt" "1_Cy5_Slide2.txt"
> > RG <- read.maimages(files, source="imagene") # read in the files
>Read header information
>Read 0_Cy3_Slide1.txt
>Read 1_Cy5_Slide1.txt
>Read 0_Cy3_Slide2.txt
>Read 1_Cy5_Slide2.txt
> > types <- readSpotTypes() # read from SpotTypes.txt
> > types
>SpotType ID Name Color
>1 Gene * * black
>2 Spot Control Spot brown
>3 COT Control COT brown
>4 Dros Control Dros* brown
>5 Cy3 Control Cy3 green
>6 Cy5 Control Cy5 red
> > status <- controlStatus(types,RG)
>Matching patterns for:
>Error in matrix(unlist(value, recursive = FALSE, use.names = FALSE), nrow =
>length(rn), :
>attempt to set an attribute on NULL
> >
>As far as I know this is what the manual suggests. What am I doing wrong? Is
>there a worked example that uses Imagene files?
>
>Kind regards,
>Edo Plantinga



More information about the Bioconductor mailing list