[BioC] info su illumina

[Giovanni Coppola]

Hello Mark,
thank you very much for your email.
The output from BeadStudio is without normalization.
I am analyzing these arrays building the expression matrix, and  
normalizing between arrays (quantile) with limma.
After linear modeling, the results are biologically sound, but I  
usually do an extensive low-level analysis (eg consistency of  
replicates across the array,  distribution of controls etc), which  
apparently I can't have now.

My only quality measures right now are:
boxplot of detection scores, BeadStDev, and AvgNBeads
scatterplot of (log) signal vs bead SD, showing a direct correlation.
Do you have other suggestions? I will try plotMAXY.
Thanks
Giovanni






On Mar 4, 2006, at 10:58 AM, Mark Dunning wrote:

> Hi Giovanni,
>
> You will be able to analyse this data using 'beadarray'. The  
> package is able
> to read bead-summary data which has been processed by BeadStudio (a  
> single
> averaged value for each bead type on an array) or the bead-level  
> data. If
> you make the modifications that Lynn suggests you will able to extract
> intensities of each bead using the scanning software. It does not  
> sound like
> you have these files though. However, to make full use of the  
> beadarray
> package you would also require files which give the coordinates of  
> each bead
> on each array. Unfortunately, Illumina seem quite reluctant to give  
> out the
> bead centre coordinates at the present time. We are however  
> pressing them to
> make this information available.
>
> I think the best way for you to proceed would be to read the file
> geneprofile.csv into beadarray. The quickest way to do this is to  
> put the
> file in a directory on it's own, change the R working directory to  
> point to
> this and use the "readBeadSummaryData" command from beadarray. You  
> will
> probably need to use the columns parameter to specify the names of the
> columns that are being used in the file. I believe the following  
> command
> will work for the standard output from BeadStudio
>
> BSData = readBeadSummaryData(columns=list(ProbeID="TargetID",
> AvgSig="AVG_Signal", Nobeads="Avg_NBEADS", BeadStDev="BEAD_STDEV",  
> Detection
> ="Detection"))
>
> This will then create a list object similar to the RGList object  
> used in
> 'limma', ie the expression values are in the $R matrix with rows
> representing genes and columns representing arrays. This expression  
> matrix
> can be used for clustering, pca, linear modelling (via limma) or
> normalisation. So far we provide median, quantile or qspline  
> normalisation -
> although methods within the 'affy' package could also be used.  
> Additionally,
> the 'plotMAXY' function will give scatter and MA plots for multiple  
> arrays
> and is a useful way of judging array quality.
>
> btw do you know how the data in geneprofile.csv was normalised by
> BeadStudio? Illumina recommend a rank invariant normalisation plus
> background correction, but that seems like a very bad idea to me  
> looking at
> some of the data it produces!
>
> Please let me know if you have any further questions or comments. The
> package is still in the development stage, so feedback is very much
> appreciated :)
>
> Regards,
>
> Mark
> -------------
> Mark Dunning
> PhD Student
> Computational Biology Group
> Hutchison / MRC Research Centre
> Department of Oncology
> University of Cambridge
> Hills Rd, Cambridge CB2 2XZ
>
> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch
> [mailto:bioconductor-bounces at stat.math.ethz.ch]On Behalf Of Giovanni
> Coppola
> Sent: 04 March 2006 16:54
> To: Lynn Amon
> Cc: Bioconductor
> Subject: Re: [BioC] info su illumina
>
>
> Hello Lynn,
> I am trying to analyze Illumina (Sentrix BeadChip Mouse Ref8) data
> with the beadarray package.
> All I have is:
>
> 1) a geneprofile.csv (matrix with expression values)
> 2) for each array strip, three files (.idat, .jpg, .xml)
> 3) one .sdf file
> 4) metrix.txt
> 5) the illumina CD with beadmap files
> 6) one copy of the BeadStudio software
>
> I assume I can't use the beadarray package, is that correct?
> Can I use the BeadStudio software to obtain bead level text files?
> Thanks
> Giovanni
>
> On Feb 8, 2006, at 10:18 AM, Lynn Amon wrote:
>
>> One possibly helpful note about Illumina arrays is that you need to
>> make
>> a change to the Settings.xml file in the ProgramFiles/Illumina/
>> BeadScan
>> directory in order to get summary or bead-level text files which can
>> then be read in using read.table or with the beadarray package.
>>
>> Change the line:  <SaveTextFiles>false</SaveTextFiles> to
>> <SaveTextFiles>true</SaveTextFiles>
>> and add the following line if you want bead-level data:
>> <SavePerBeadFiles>true</SavePerBeadFiles>
>>
>> Keep in mind that the bead-level files are large especially for the
>> whole genome chips.
>>
>> Lynn Amon
>> Research Scientist
>> Department of Pathology
>> University of Washington
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>