[BioC] LumiBatch Object - Help

[Michal Blazejczyk]

Hi Sharon,

This is how we do it.  It is a bit complicated as you can see.
The assumption is that variable signalRaw contains a matrix
with raw intensities (one column per sample).  Variable beadSD
contains a similar matrix but for bead sd's.  I think that
bead sd's are required for LumiBatch to work properly.

    library( lumi )

    #############################################
    # Fake feature information
    featureIndices <- 1 : nrow( signalRaw )
    featureNames <- paste( "f", featureIndices, sep = "" )

    rownames( signalRaw ) <- rownames( beadSD ) <- featureNames

    reporterInfo <- data.frame( featureNames )
    names( reporterInfo ) <- "DummyID"
    rownames( reporterInfo ) <- featureNames

    varMetadata <- data.frame( labelDescription = "A dummy Illumina feature identifier" )
    rownames( varMetadata ) <- "DummyID"

    featureData <- new( "AnnotatedDataFrame", data = reporterInfo, varMetadata = varMetadata )

    #############################################
    # Fake sample information
    sampleIndices <- 1 : ncol( signalRaw )
    sampleNames <- paste( "sample", sampleIndices, sep = "" )

    colnames( signalRaw ) <- colnames( beadSD ) <- sampleNames

    pData <- data.frame( sampleNames )
    names( pData ) <- "sampleID"
    rownames( pData ) <- sampleNames

    varMetadata <- data.frame( labelDescription = "A dummy sample ID" )
    rownames( varMetadata ) <- "sampleID"

    phenoData <- new( "AnnotatedDataFrame", data = pData, varMetadata = varMetadata )

    #############################################
    # Create the LumiBatch object
    lumiBatch <- new( "LumiBatch", exprs = signalRaw, se.exprs = beadSD,
        featureData = featureData, phenoData = phenoData )

Best,
Michal Blazejczyk


> Hi,

> I have an Illumina array data in the matrix form. I would like
> to create a LumiBatch object leaving all the slots empty except
> the exprs slot. Is it possible to do that? If so, how?
> I would greatly appreciate any help from you.

> Thanks in advance.

> Kind regards,
> Sharon