[BioC] Some problems in combining ExpressionSet objects with the combine() function

[James W. MacDonald]

Hi Manuela,

You don't give the output from sessionInfo(), but I suspect you have an 
outdated version of BioC.

 > combine(japan,denmark,australia)
ExpressionSet (storageMode: lockedEnvironment)
assayData: 500 features, 154 samples
   element names: exprs
protocolData: none
phenoData
   sampleNames: GSM101849 GSM101851 ... GSM327326 (154 total)
   varLabels: position satelite
   varMetadata: labelDescription
featureData: none
experimentData: use 'experimentData(object)'
Annotation: hgu133plus2
Warning messages:
1: In alleq(levels(x[[nm]]), levels(y[[nm]])) : 2 string mismatches
2: data frame column 'satelite' levels not all.equal
 > sessionInfo()
R version 2.13.0 alpha (2011-03-28 r55140)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
  [1] LC_CTYPE=en_US.iso885915       LC_NUMERIC=C
  [3] LC_TIME=en_US.iso885915        LC_COLLATE=en_US.iso885915
  [5] LC_MONETARY=C                  LC_MESSAGES=en_US.iso885915
  [7] LC_PAPER=en_US.iso885915       LC_NAME=C
  [9] LC_ADDRESS=C                   LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] affy_1.29.3     Biobase_2.11.10

loaded via a namespace (and not attached):
[1] affyio_1.19.4         preprocessCore_1.13.6

Best,

Jim



On 7/27/2011 10:17 AM, Manuela Di Russo wrote:
> Hi all,
>
> I'm struggling with the meta-analysis of microarray data. In the user guide of the bioconductor package "MAMA" there is an exemple of meta-analysis in cancer microarray data concerning the comparison of expression profiles in MSI (microsatelite instable) and MSS (microsatelite stable) colon cancer. Data are gathered from three microarray experiments from public databases
>
> The original data sets have been preprocessed and subsampled (same set of 500 genes). All data sets have been normalized and are in log2-scale. The corresponding sample sizes for the three datasets, named denmark, australia  and japan, are 77 (39 MSI and 38 MSS), 36 (5 MSI and 31 MSS) and 41 (16 MSI and 25 MSS), respectively. Each of the datasets is stored as an ExpressionSet object.
>
> I would like to perform a meta-analysis combining the three ExpressionSets into one and using the limma package with a convenient design matrix to model the study effect.
>
> I have two questions:
>
> 1) I tried using the function "combine" in order to combine the three datasets (two by two, too) but I obtain the following error message:
>
> library(MAMA)
>
> library(affy)
>
> load(url("http://math.muni.cz/~xihnatov/DataColonHalf.RData"))
>
> ls()
>
> data<-combine(australia,denmark,japan)
>
> Error in data.frame(<S4 object of class "ExpressionSet">,<S4 object of class "ExpressionSet">, : arguments imply differing number of rows: 36, 77, 41.
>
> I checked the function instructions:
>
> combine(eSet,eSet): Combine two eSet objects. To be combined, eSets must have identical numbers of featureNames, distinct sampleNames, and identical annotation
>
> and the three datasets meet these conditions.
>
> So, can someone help me? Is there a problem in the datasets I downloaded?
>
> 2) How can I access the code of the function "model.matrix" (package : stats)? Typing the name of the function as I usuallly do seems not to work in this case. Is it a "protected" function?
>
> Thank you for your help!
>
> Manuela
>
> ----------------------------------------------------------------------------------------
> Manuela Di Russo, Ph.D. Student
> Department of Experimental Pathology, MBIE
> University of Pisa
> Pisa, Italy
> e-mail: manuela.dirusso at for.unipi.it
> tel: +39050993538
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826
**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues