Bioinformatics methods analysing gene expression data from complex chemical exposure.

The complexity of gene expression data from DNA-microarraysassays, make classic statistics difficulty applicable, because of the high dimension of the data set, of the low cardinality (asmall number of vectors) and of noise. Computational automatic learning methods concur to inferring knowledge from the genomic and proteomic data, using a limited number of samples and in absence of meaningful a prioriknowledge of the problem. This is particularly helpful in case of complex toxicological experiments involving multi chemicals exposure. In this poster we briefly show the applications of Bioinformatics methods togene expression analysis in the adrenals of mice chronically exposed to Atrazineand Arsenate in drinking water