[BioC] limma layout query (genepix GAL file)

[Gordon Smyth]

At 08:00 PM 1/10/2005, bioconductor-request at stat.math.ethz.ch wrote:
>Date: Fri, 30 Sep 2005 11:50:58 +0200
>From: Dave Berger <dberger at postino.up.ac.za>
>Subject: [BioC] limma layout query (genepix GAL file)
>To: bioconductor at stat.math.ethz.ch
>Message-ID: <433D0A82.B293F5F at postino.up.ac.za>
>Content-Type: text/plain; charset=us-ascii
>
>Hi
>using R version 2.1.1
>limma version 1.9.6
>
>We know our Slide layout is
>$ngrid.r
>[1] 12
>$ngrid.c
>[1] 2
>$nspot.r
>[1] 7
>$nspot.c
>[1] 32
>$ndups
>[1] 2
>$spacing
>[1] 224
>$npins
>[1] 12
>$start
>[1] "topleft"
>however when we read in layout from GAL file as below
>RG$genes<-readGAL("M64_AtSSH_correct.gal")
>RG$genes$Status<-controlStatus(spottypes, RG)
>RG$printer<-getLayout(RG$genes)
>
>we get
>
>$ngrid.r
>[1] 6
>$ngrid.c
>[1] 4
>$nspot.r
>[1] 7
>$nspot.c
>[1] 32
>
>can you explain why this does not correspond to our expected layout?
>below  is an extract of first 2 blocks from our GAL file
>"M64_AtSSH_correct.gal" from Genepix version 5.1
>The remaining 22 block follow the same row and column patterns.
>
>thanks
>Dave Berger

1. There actually is no foolproof way to determine the printer layout 
information from the columns of a gal file, because GenePix stores only the 
block numbers. Hence getLayout() is forced to assume ngrid.c=4, which is by 
far the most common value. This is actually the reason why the printer 
layout information is not set automatically by read.maimages() for GenePix 
data. There is no way to be sure it is right without manual intervention.

2. You could try using getLayou2() instead, which is an experimental 
function designed to get around the above problem. If that doesn't work, 
set the printer layout information manually.

3. There is actually no need to use readGAL() at all with GenePix data. You 
will see that read.maimages() already populates RG$genes with the relevant 
information.

Gordon