Dear all, Maybe this is a stupid question but... Is there a general assumption about the way the highest expression values of a gene (/all genes?) in a series of chips would be, such as log normal, normal...? I know that there are some outliers in these part of the spectrum but what else if they are excluded from analysis ? Is there a paper / study which deals with the distribution of gene expression for extreme values that somebody can suggest me ? thanks again Philippe [[alternative HTML version deleted]]