I was trying out the DESeq package, whose documentation says: "The CountDataSet class is derived from Biobase?s eSet class and so shares all features of this standard Bioconductor class." So I tried exprs() on it, which didn't work. Of course, CountDataSets have their own counts() function. I was just wondering why they have different names. Why not make count data accessible through exprs()? Is the reasoning behind this due to R internals, or is it biologically motivated?

Dear Ed the documentation of eSet (try typing "class ? eSet" into R after loading Biobase) does not mention an exprs method. Perhaps you are confusing it with ExpressionSet. A schematic interpretation is: - ExpressionSet is for processed datasets and presents a uniform interface for downstream analyses (e.g. machine learning) - Other eSet derived classes (e.g. countDataSet, NChannelSet, the cellHTS class in cellHTS2) are for 'matrix-like' raw data, and reflect their ideosyncrasis, which is useful and needed for technology- specific preprocessing algorithms. E.g. the data in a CountDataSet are non-negative integers. Hope this helps Wolfgang Apr/3/12 5:28 PM, Ed Siefker scripsit:: > I was trying out the DESeq package, whose documentation says: > > "The CountDataSet class is derived from Biobase?s eSet class and so > shares all features of > this standard Bioconductor class." > > So I tried exprs() on it, which didn't work. Of course, CountDataSets > have their own > counts() function. I was just wondering why they have different > names. Why not make > count data accessible through exprs()? Is the reasoning behind this > due to R internals, > or is it biologically motivated? > > _______________________________________________ > 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 -- Best wishes Wolfgang Wolfgang Huber EMBL http://www.embl.de/research/units/genome_biology/huber