Could you provide more explanation or help on how to convert expression data from simple .tsv table into an ExpressionSet object? I have expression values for a list of genes (n rows, each corresponding to one genes) in mixed samples (m columns, each corresponding to a mixed sample). In a separate file, I have the expression values for a subset of these genes (j rows) in pure tissue or cell types (k columns, each corresponding to a pure cell type). I would like to run the CellMix algorithms on this dataset to calculate the fractions of the pure cells in the mixed samples. Some extra explanation would be much appreciated. Thanks, Maria [[alternative HTML version deleted]]

Hi Maria, note that you do not need to use an ExpressionSet object. Plain matrix objects should work as well, as long as the rownames match: x <- as.matrix(read.table('mixed_ data.tsv', sep = "\t")) y <- as.matrix(read.table('celltype_data.tsv', sep = "\t")) res <- gedProportions(x, y, rescale = FALSE, log = FALSE, normalize = FALSE) # unscaled proportions coef(res) # rescaled proportions (sum up to one) scoef(res) If you later want to create an ExpressionSet then it just about wrapping the matrix object: eset <- ExpressionSet(x) but this does not make much sense if you do not have extra phenotypic data or gene annotation package name to associate your gene expression data with. Hope this helps. Please let me know if this works fine for you. Bests, Renaud