[BioC] beadarray - readBeadLevelData problems

[Mark Dunning]

Hi Felix,

I think this might be a problem with the format of text files created  
by the scanner. When developing beadarray, the files we were using  
always had the same number of rows. However, the file format used by  
Illumina now can have a different number of rows and the package does  
not like this very much :(

We recently replaced readBeadLevelData with a function readIllumina  
which is better for reading bead level data. The function can be  
obtained from the development version of beadarray (version 1.3.1). I  
would recommend downloading this developmental version of the package  
for bead level analysis.

Regards,

Mark

PS please could you make any future replies to the Bioconductor list.  
That way, there is a permanent record of problems with the package  
and the solution that people can search in the future.


On 12 Feb 2007, at 17:10, Engel Felix wrote:

> Hi Mark,
>
> thanks for the help - it works fine for a single array. If I have  
> two arrays specified for reading, I get this error when R starts  
> processing the second array.:
>
> Error in BLData at ProbeID <mailto:BLData at ProbeID> [, i] <- dat1 
> $ProbeID[ord] :
>         number of items to replace is not a multiple of replacement  
> length
>
> Also, our CSV files use the comma instead of fullstop as the  
> decimal seperator, but that is easy to fix.
>
> Thanks again,
>
> Felix
>
> ________________________________
>
> From: M. Dunning on behalf of Mark Dunning
> Sent: Mon 12-Feb-07 17:40
> To: Engel Felix
> Cc: bioconductor at stat.math.ethz.ch
> Subject: Re: beadarray - readBeadLevelData problems
>
>
>
> Hi Felix,
>
> I looked at the files you sent me, and it seems you are trying to  
> read a
> pgm image file into the package. This was the file format that was  
> first
> used in readBeadLevelData, but we recently changed it to accept tif  
> images
> instead (ie the images that come straight off the scanner). Try  
> reading the
> tifs and see if that works.
>
> Hope this helps,
>
> Mark
>
> PS we are currently making some changes to the package which will  
> make the
> reading of bead level data easier and more efficient. The changes will
> shortly be submitted to the development version of the package.
>
>
>
>
> On Feb 12 2007, Engel Felix wrote:
>
>> Dear Mark,
>>
>> we would like to use beadarray for analysis for bead level data  
>> analysis.
>> I use version 1.2.2 of your package and R 2.4.1 on Linux. The  
>> problem I
>> encounter while reading raw image data is that the function
>> readBeadLevelData does return an array with the correct probe  
>> position
>> information, but only zeros for the greenIntensities (as returned  
>> from
>> your external C function). Still, the function returns without error.
>>
>> The location for the test files I use:
>>
>> http://www.dkfz.de/genetics/supplement/engel/beadarray/ 
>> beadarray.zip <http://www.dkfz.de/genetics/supplement/engel/ 
>> beadarray/beadarray.zip>
>>
>> Maybe you can offer some help.
>>
>> Thank you so much.
>>
>> Best regards,
>>
>> Felix
>>
>
> --
> Mark Dunning
> Computational Biology Group
> Department of Oncology, University of Cambridge / Cancer Research UK
> Cancer Research Institute, Li Ka Shing Centre
> Robinson Way
> Cambridge CB2 0RE
> UNITED KINGDOM
> Tel: +44 1223 404295
>
>
>