Determination of cellular drug targets : searching for functional information in the jungle of microarrays data
We present in this paper an original analysis method for microarray data based on an attribute estimation technique. We show, in the context of radioactive environmental stress, that it is possible to detect with this method a small number of genes which, with high probability, are involved in a response to this stress. Then we study the different processes these informative genes are participating in. Moreover, we draw a ranking of the genes that can turn out to be useful to discriminate between the two conditions (irradiated versus non irradiated) of the experiments. Using this method, we highlighted the transcriptional response of yeast cells to low doses of radiation.