Hi Iain, I am experiencing the same problem as you on a ExpressionSet generated from justRMA() Did you found a solution? Here is the session: > da.rma <- justRMA(celfile.path="CELfiles/" , destructive=TRUE, phenoData=pd) > da.rma ExpressionSet (storageMode: lockedEnvironment) assayData: 54675 features, 214 samples element names: exprs, se.exprs phenoData rowNames: [...] > class(pd) [1] "data.frame" > pData(da.rma) <- pd > results <- pairwise.comparison(da.rma, "group", c("A","B")) Error in validObject(.Object) : invalid class "PairComp" object: invalid object for slot "pData" in class "PairComp": got class "factor", should be or extend class "data.frame" > sessionInfo() R version 2.7.1 (2008-06-23) i386-apple-darwin9.3.0 locale: en_US.UTF-8/en_US.UTF-8/C/C/en_US.UTF-8/en_US.UTF-8 attached base packages: [1] splines tools stats graphics grDevices datasets utils methods base other attached packages: [1] simpleaffy_2.16.0 gcrma_2.12.1 matchprobes_1.12.0 genefilter_1.20.0 survival_2.34-1 [6] hgu133plus2cdf_2.2.0 affy_1.18.2 preprocessCore_1.2.0 affyio_1.8.0 Biobase_2.0.1 loaded via a namespace (and not attached): [1] AnnotationDbi_1.2.2 DBI_0.2-4 RSQLite_0.6-9 annotate_1.18.0 tcltk_2.7.1 On Feb 4, 2008, at 12:50 PM, IAIN GALLAGHER wrote: > Hi List. > > I'm trying to use the simpleaffy pairwise.comparison procedure on > two groups of arrays. The groups are defined in the phenoData slot > under the heading 'group' and are indicated by numbers i.e. > 1,1,2,2,2,3,3,3 etc. > > If I try: > > results<-pairwise.comparison(test, 'group',c(1,4)) > > where 'test' is an appropriate subset of the data I get the > following error: > > Error in validObject(.Object) : > invalid class "PairComp" object: 1: invalid object for slot > "members" in class "PairComp": got class "numeric", should be or > extend class "character" > invalid class "PairComp" object: 2: invalid object for slot "pData" > in class "PairComp": got class "integer", should be or extend class > "data.frame" > > Is it possible to change the class of the 'group' indicators in my > phenoData slot or is there another way to get around this problem? > > Thanks > > Iain >