I am not an expert, but I am particularly interested in cancer biology. If I understand correctly, two models of melanoma progression are being proposed. According to Haqq et al (2005) the molecular signatures change during tumor progression (like in other tumors). In contrast, you (Harbst et al 2012, this pre-print) propose that CITED1 signature is there from the very beggining (which is very interesting), although you could not analyze CITED1 RNA due to technical reasons. WB from melanoma cell lines fit your proposal. I would like to know by immunocitofluorecence the percentage of cells that express CITED1 in your (223) primary melanomas...