18 May 2005
The first global analysis of human microarray data has been released. DNA microarrays or gene chips are used to study patterns of cellular gene expression, and have been widely used in the investigation of tumour-specific genetic profiles. A new study has used a global approach to look at a total of 1975 datasets from microarray analyses of cancer cells, comprising 22 different types of tumour, from the Stanford Microarray Database and Whitehead Institute Center for Genomic Research database. The 14,145 genes analysed were assigned to modules, groups of genes that act together to perform specific functions. A total of 456 modules found to represent statistically significant gene expression patterns associated with various functions were analysed with biological and clinical data, to identify conditions in which particular modules were induced or repressed.
The end-product of this process was a map showing the status of modules in different forms of cancer. Some modules were shown to be present in a range of tumour types; for example, an osteoblastic module comprising genes associated with the proliferation and differentiation of cells that produce bone was implicated in breast cancer, lung cancer, hepatocellular carcinoma and acute lymphoblastic leukemia. These tumours may therefore share pathological mechanisms of progression and metastasis. Other modules showed tumour-specific associations, such as a growth inhibitory module found to be repressed in acute leukemia. Only one of the eleven growth suppressor genes present within the module was previously implicated in acute leukaemia; the other ten represent potential novel therapeutic targets. The researchers, whose work is published in the October edition of Nature Genetics, say that the study