[BioC] Normalization of Illumina Array with separate bgx-file

[Stefanie Busch]

   Hello,

   I'm Stefanie from Germany. So excuse me when my english isn't very well.

   I want to normalize the raw data of the GEO experiment "GSE39313". In this
   experiment the Illumina MouseRef-8 v2.0 expression beadchip was used. I get
   the data with getGEOSuppFiles but I don't know what I should do next. I know
   that I can analyse Illumina chips with the packages beadarray, lumi or
   limma. I have uploaded the raw.txt-file with the different packages.

   For example:
   > Daten.lumi<-lumiR(file="GSE39313_non-normalized.txt")
   Please provide Illumina ID Mapping library!
   > Daten.lumi
   ExpressionSet (storageMode: lockedEnvironment)
   assayData: 25697 features, 48 samples
     element names: detection, exprs
   protocolData: none
   phenoData
     sampleNames: 7539447003_A 7539447003_B ... 7917911031_H (48 total)
     varLabels: sampleID
     varMetadata: labelDescription
   featureData
     featureNames: ILMN_2896528 ILMN_2721178 ... ILMN_2424408 (25697
       total)
     fvarLabels: ID_REF
     fvarMetadata: labelDescription
   experimentData: use 'experimentData(object)'
   Annotation:

   As you can see I have no Annotation information. GEO provide the bgx.file
   for the Chip (Illumina MouseRef-8 v2.0 expression beadchip) but I don't know
   how to connect this. I tried this:

   >
   Daten.lumi2<-addNuID2lumi(Daten.lumi,annotationFile="MouseRef-8_V2_0_R3_1127
   8551_A.bgx")
   > Daten.lumi2
   ExpressionSet (storageMode: lockedEnvironment)
   assayData: 25697 features, 48 samples
     element names: detection, exprs
   protocolData: none
   phenoData
     sampleNames: 7539447003_A 7539447003_B ... 7917911031_H (48 total)
     varLabels: sampleID
     varMetadata: labelDescription
   featureData
     featureNames: WpaZ9x9f_hAnoR.VBE ZhdXp75JftSF3iWLF4 ...
       Zl4J4B0PsjTJ1N19IM (25697 total)
     fvarLabels: ID_REF
     fvarMetadata: labelDescription
   experimentData: use 'experimentData(object)'
   Annotation:

   But there are again no annotation Information. And when I used the exprs()
   function I get this what looks very strange:

   > head(exprs(Daten.lumi2))
                      7539447003_A 7539447003_B 7539447003_C 7539447003_D
   WpaZ9x9f_hAnoR.VBE    4476.8840    5870.2020    5149.9380    5193.5800
   ZhdXp75JftSF3iWLF4     970.4051    1032.8760    1132.4740    1280.3510
   xnx7ijE2CcOSlY55eA    1577.7980    1511.4770    1711.2830    1850.6020
   xPkUXaq37oR_vSi6HY    9862.2040   11187.4000   12783.4400   14389.9900
   3rtRBd9AcK1Qq07fGI     222.3842     236.7587     210.8107     229.3236
   ZqRKnn0K14kTIep1.c    3877.7640    5116.4140    6129.5220    2486.4500


   So I tried to use limma:
   >
   Daten.limma<-read.ilmn(files="GSE39313_non-normalized.txt",other.columns="De
   tection Pval")

   This works well, but I don't know where to get the bgx.file in. And the
   further problem is, when I want to normalize according to the instructions
   of BeadArrayUseCases page 25 I get this error message:

   > maqc.norm<-neqc(Daten.limma)
   Fehler in normexp.fit.detection.p(x, detection.p) :
     Detection p values not found in the data.

   Please can you tell me where is my mistake?


   Stefanie