Dear all, Im sorry if this is a question not directly related to using bioconductor packages, but at least I used them for the analysis that I need a reference to ;o) I have a biopsy related project with 63 unpaired biopsies in total and two groups where I cannot find any differences using limma, probably due to tissue heterogeneity. I wonder if there is anyone who has published a paper perhaps with different mixes of cell lines and compared that to clean comparisons between the cell lines, in order to draw conclusions about what level of heterogeneity that is possible to detect using e.g. Affymetrix/Agilent or some other platform? Im going to try to publish a negative result and it would be great if anyone has tried to do some quantification around this problem. I have looked around and I found the paper by Holloway et al, "Statistical analysis of an RNA titration series evaluates microarray precision and sensitivity on a whole-array basis", which is a really good paper but here they only have two cell lines and it would be nice to have a look of what it looks like in mixtures. best regards, // Johan [[alternative HTML version deleted]]