I have a question about clustering. I' m managing gene expression microarray data and I would like to cluster them in classes. I searched around to find the best clustering algorithm for my data, but since as in all gene expression microarray experiments the number of genes exceed greatly the number of samples being analyzed, the majority of this algorithms assume independency of the variables (so the genes). But in reality we all know that the activity of one gene is not independent from the activity of another gene. My question is: is there an algorithm or a technique able to take into account with simplifications such a dependencies between genes?
Thanks a lot!