[BioC] marrayNorm to exprSet object

[Sean Davis]

Daniel Brewer wrote:
> Hello all,
> 
> I am new to bioconductor and I am running into a couple of problems
> converting a marrayNorm object to an exprSet object.  I convert the
> marrayNorm object to an exprSet using the following:
> 
> library("convert")
> xEset <- as(marrayNorm object, "exprSet")
> 
> The problem is that there does not seem to be any probe level
> information in the exprSet i.e.
> 
>>geneNames(monkey)
> 
> NULL
> 
>>featureNames(monkey)
> 
> Error in featureNames(monkey) : no direct or inherited method for
> function 'featureNames' for this call
> 
> This information is available in the marrayNorm object:
> 
>>norm at maGnames@maInfo[1:10,]
> 
>             ID     Name
> X22170   22170 DXF68S1E
> X21642   21642    ABCA2
> X22209   22209 LOC56990
> X22121   22121    PDZD2
> X24284   24284     PACE
> X66999   66999     CD74
> X110223 110223      DLD
> X110202 110202   EEF1A1
> X21744   21744     GGA2
> X22739   22739 FLJ14566
> 
> I can fix the gene names by doing the following
> 
> geneNames(monkey) <- norm at maGnames@maInfo$Name
> 
> but the same does not work with the feature names.  Is the correct way
> to do this?  Anyway round the feature name problem.

Hi, Daniel.  Welcome to bioconductor.

The geneNames slot of an exprSet is meant to hold the unique name for 
each feature on the array, like the affy probeset id, for example.  It 
can hold the actual names of genes, but then will probably not be 
unique.  In the larger picture, the exprSet is not meant to hold any 
annotation at all.  It was designed to be used in conjunction with a 
bioconductor annotation package.  I personally find exprSets a bit 
difficult to use when I am using my own annotation, just for this 
reason.  What array platform are you using and what are you trying to 
accomplish with converting to an exprSet?

Sean