I’m working with gene expression microarray data. I want to use the non-parametric method RankProd for identifying differentially expressed (up- or down- regulated) genes (bioconductor package). I have used this method previously and always with satisfactory results. Now I have a large matrix (95 samples) and I want to do a non-paired analysis (disease vs control), but the Rankprod not work and I get strange results. I would like to ask you if there is a limit to the input of the method rankprod.
Thanks in advance, Teresa