[BioC] How to use affy's image function on RMA values?
bolstad at stat.berkeley.edu
Thu Jun 9 06:27:22 CEST 2005
I'm a little perplexed by why you want to replace the raw probe
intensities by processed RMA expression values. It makes no sense to do
this and what is happening is that your RMA expression values are being
duplicated many times to fill up the expr slot of the AffyBatch (this
probably explains the diagonal structures you are observing).
Instead this is what I think you really want to do:
Data <- ReadAffy( filenames=c("A1.CEL", etc, "A40.CEL"))
image(Data) # to create/view images of the raw intensities
my.Pset <- rmaPLM(Data)
image(my.Pset,type="resids") #image the residuals from RMA
image(my.Pset,type="pos.resids") #image the positive residuals from RMA
image(my.Pset,type="neg.resids") #image the negative residuals from RMA
image(my.Pset,type="sign.resids") #image the sign of residuals from RMA
Note that image() on PLMsets also will take a "which" argument to see
individual arrays rather than the whole dataset in quick sucession
would show an image of the RMA residuals for the second chip in the
dataset. If you are interested there are a large number of images at
of this type which you could use as a comparison aid on your own images
Hope this helps,
On Wed, 2005-06-08 at 18:47 -0700, Malard, Joel M wrote:
> I would like to compare (visually) the raw intensities of 40 affymetrix
> MOE430A arrays with the corresponding expression values computed by RMA,
> e.g. using the function image.
> It seemed a good idea to replace the exprs component of the AffyBatch
> object with the exprs component of the object returned by justRMA but
> that does not work.
> data <- ReadAffy(filenames=c("A1.CEL", etc, "A40.CEL"));
> exprs(data) <- exprs(justRMA(filenames=c("A1.CEL", etc,
> The error message is:
> > image(data[,1]);
> Warning message:
> data length  is not a sub-multiple or multiple of the number of
> rows  in matrix
> The resulting image shows diagonal lines that shouldn't be there.
> Could anyone put me on the right track?
> Thanks for your help.
> Joel Malard
> [[alternative HTML version deleted]]
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
Ben Bolstad <bolstad at stat.berkeley.edu>
More information about the Bioconductor