Hello Arne, Several publications recently have shown that the distribution of gene abundances seems to follow a power law (one such is referenced at www.the-scientist.com. This should turn into an exponential density on the log scale we often use. The 'hump' we see in signal intensities from microarray experiments probably reflects variable levels of non- specific hybridization to each probe. If we could estimate this, we would be further along in estimating fold changes. regards Mark