[BioC] Using read.maimages() with Imagene output files.

[Gordon Smyth]

OK, try out limma version 1.1.11 which is available from 
http://bioinf.wehi.edu.au/limma/

I have added "imagene" as a possible value for the argument 'source' for 
'read.maimages'. For Imagene data, the argument 'files' should contain a 
matrix of file names, the first column for the cy3 channel files and the 
second column for the cy5 channel files. I have made "Signal Mean" the 
default for foreground and "Background Median" the default for background. 
Here is a little example:

 > files
      [,1]                     [,2]
[1,] "cy3-pt1-cf-t0-7946.txt" "cy5-pt1-cf-t0-7946.txt"
[2,] "cy3-pt1-wt-t3-8179.txt" "cy5-pt1-wt-t3-8179.txt"
 > RG <- read.maimages(files,source="imagene",names=c("slide7946","slide8179"))
Read header information
Read cy3-pt1-cf-t0-7946.txt
Read cy5-pt1-cf-t0-7946.txt
Read cy3-pt1-wt-t3-8179.txt
Read cy5-pt1-wt-t3-8179.txt
 > MA <- normalizeWithinArrays(RG)
 > RG
An object of class "RGList"
$R
      slide7946 slide8179
[1,]  471.7000 1310.4737
[2,]  478.9474 1192.4118
[3,]  603.3200 1678.8700
[4,]  698.8986 1658.7586
[5,]  428.0727  589.9016
19195 more rows ...

$G
      slide7946 slide8179
[1,]  543.8400 1021.7600
[2,]  492.4182 1033.5517
[3,]  722.1818 1505.1881
[4,]  696.3235 1454.2424
[5,]  460.9800  588.2407
19195 more rows ...

$Rb
      slide7946 slide8179
[1,]     582.0      1563
[2,]     624.0      1609
[3,]     530.0      1615
[4,]     606.5      1429
[5,]     706.0      1256
19195 more rows ...

$Gb
      slide7946 slide8179
[1,]       598    1337.0
[2,]       643    1271.5
[3,]       730    1663.0
[4,]       741    1348.0
[5,]       714    1067.5
19195 more rows ...

$printer
$ngrid.r
[1] 8

$ngrid.c
[1] 4

$nspot.r
[1] 25

$nspot.c
[1] 24


$genes
   Field Meta Row Meta Column Row Column Gene ID
1  test        1           1   1      1 R11726:
2  test        1           1   1      2 R11726:
3  test        1           1   1      3 R37412:
4  test        1           1   1      4 R37412:
5  test        1           1   1      5 R11793:
19195 more rows ...


You can change the foreground/background if you like, for example:

RG <- read.maimages(files,source="imagene",columns=list(f="Signal 
Mean",b="Background Mean"))

Gordon

At 07:29 PM 14/08/2003, Luke Whitaker wrote:
>Dear Gordon,
>
>Thank you for your offer. I am enclosing an example of a (truncated)
>Imagene file. I have deleted 7869 lines from the "Raw Data" part
>of the file partly because the data is confidential but mainly to
>make the attachment a more reasonable size - the original file is
>4Mbytes. I hope this isn't a problem. If it is, let me know and I
>will work something out.
>
>Note that the Imagene format relies on tabs to separate fields,
>and that many of the items can contain spaces. In particular, the
>"Raw Data" column headers often contain spaces (eg "Meta Row",
>"Meta Column", "Signal Mean", "Background Mean" etc etc) and also,
>the "Gene ID" names can contain spaces (eg "(+)Pro25G onG3PDH570 1(60)"
>in the example I have given you).
>
>I believe that as long as the distinction between tabs and spaces
>is observed then there is no ambiguity.
>
>As I said before, there is one file for each dye, and there does
>not appear to be any information in the file itself as to which
>dye the file is for. We encode the dye in the tiff filename, but
>that is not enforced.
>
>Also, I understand that the columns that are output, and possibly
>the actual column names, are to some extent configurable in Imagene,
>so it is probably not safe to assume that the file will always
>contain exactly those columns with those particular column headers.
>
>If there is any further information I can give you then please let
>me know. I would be very grateful if you could let me know how you
>get on.
>
>Thanks,
>
>Luke
>Inpharmatica
>
>
>On Thu, 14 Aug 2003, Gordon Smyth wrote:
>
> > Dear Luke,
> >
> > At 01:31 AM 14/08/2003, Luke Whitaker wrote:
> > >I am attempting to analyse some microarray data that has been
> > >scanned using a program called Imagene and I would like to try and
> > >use limma for the analysis. I was wondering if it might be possible
> > >to use read.maimages() to read these data into R.
> > >
> > >It appears to me that read.maimages() expects both the red and green
> > >signals to be present in a single file. Imagene produces separate 
> files for
> > >the red and green signals. Does anyone out there have any suggestions how
> > >to proceed ?
> >
> > What you say is correct - read.maimages does not support Imagene files, 
> nor
> > does any other function in Bioconductor.
> >
> > If you would be prepared to make available to me some examples of imagene
> > data files, then I will add imagene functionality to read.maimages.
> >
> > Gordon
> >
> > >Thanks in advance,
> > >
> > >Luke Whitaker
> > >Inpharmatica