Random forest clustering is attractive for tissue microarray and other immunohistochemistry data since it handles highly skewed tumor marker expressions well and weighs the contribution of each ...