[BioC] limma and ArrayVision

Gordon Smyth smyth at wehi.edu.au
Sat Sep 27 16:16:32 MEST 2003


Thanks for the example ArrayVision file that you sent me. I now have two 
examples of ArrayVision files from different people - I give the first 3 
lines of the two files below. It is clear that the ArrayVision files are 
designed for reading in to Excel but not for anything else. Note the 
column-names over two lines and the use of non-ascii characters. (Question 
marks "?" indicate that the character was not recognized by my text 
editor.) In the first example there is no explicit information about which 
columns were Cy3 and which were Cy5.

This is what I propose to do for ArrayVision files.
- I will assume that the leading columns (not counting the first column) 
are Cy5 measurements and the trailing columns are Cy3.
- I will assume that columns labeled something like "Median Dens - RFU / 
µm2" are the foreground intensities and that those labeled "Bkgd" are the 
background.

Any other ideas?
Gordon

EXAMPLE 1
LINE 1: ,Ctrl,Ctrl,Ctrl,Ctrl,Data,Data,Data,Data,
LINE 2: Spot labels,Median Dens - RFU / µm2,MAD - RFU / 
µm2,Bkgd,sMedianDens,Median Dens - RFU / µm2,MAD - RFU / µm2,Bkgd,sMedianDens,
LINE 3: R1 - C1 : 1,29.072,11.878,33.825,0.000,1.247,0.299,0.642,0.605,

EXAMPLE 2
LINE 1: ,"Cy5 right","Cy5 right","Cy5 right","Cy5 right","Cy5 right","Cy3 
right","Cy3 right","Cy3 right","Cy3 right","Cy3 right"
LINE 2: "Spot labels","Median Dens - RFU / ?m2","SD - RFU / ?m2","Area - 
?m2","Bkgd","nMedianDens","Median Dens - RFU / ?m2","SD - RFU / ?m2","Area 
- ?m2","Bkgd","nMedianDens"
LINE 3: "B1 : R1 - C1",0.298,0.161,38400,0.303,0,0.499,0.171,38400,0.492,0.002

At 09:15 PM 26/09/2003, alessandro ambrosi wrote:
>Hi all,
>
>first of all thanks again for all your help.
>Reading ArrayVision files sometimes I get the following
>
>> > read.maimages("2706L.csv", source="arrayvision", sep=",")
>>Error in "[.data.frame"(obj, , columns$Gb) :
>>         not all specified columns exist
>
>I noted that ArrayVision files have not a standard number of
>columns and it depends on which informations user decided to
>save. So the interesting columns (relative to intensity and
>background) can change position. One must pay attention that
>
>Ctrl Median Dens -> Rf = 2,
>Ctrl Bkg         -> Rb = 5,
>Data Median Dens -> Gf = 7,
>Data Bkg         -> Gb = 10
>
>correctly.
>Hope it helps.
>
>Best,
>Alessandro
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor



More information about the Bioconductor mailing list