Dear list, I have a set of 5 HuGeneFL arrays which were hybridized with samples from 5 different leukemia patients. So each sample contained lots of leukemia cells (>80%) however were still a mixture of different cell types. For the same patients there are also arrays (hgu133a) which were hybridized with two purified cell compartments: the leukemic cells and another immunophenotypically defined non-leukemic cell type. Besides looking at the detection calls, is there a (more quantitative) way to decide for a set of interesting genes whether the signal measured in the HuGeneFL chips comes from the leukemic cells or the non-leukemic cells or both? Thanks in advance Lewin Eisele ---------------------------------------------------------------------- - Dr. med. Lewin Eisele Innere Klinik - Tumorforschung Universit?tsklinikum Hufelandstr. 55 D-45122 Essen Germany