The application of modern high throughput genomic and metabonomic technologies to the field of toxicology will provide significant breakthroughs and advances. By simultaneously looking at data on gene expression and metabolite concentrations, a more accurate and complete picture of cellular behavior can be determined. Using statistical and mathematical algorithms applied to high throughput data, toxicant exposure characterization will be computed including identification of type of exposure and estimates of dose amount. In the current proposal, we intend to apply the latest multivariate linear and non-linear statistical and mathematical techniques to find subtle and complex patterns in the data that are consistent with validated samples of known toxic exposure. By using supervised machine learning techniques, training data with cross-validated measurements will be used to quantitatively measure the accuracy of the proposed statistical techniques. By applying these techniques to a wide variety of both public and in house data samples consisting of gene expression data and metabolomic NMR concentration data, small correlations and patterns can be measured and used to characterize the type and amount of environmental and toxicant exposure to host organisms. The basic research carried out in this proposal can result in a useful analysis tool that has a broad applications in drug discovery as well as diagnostic applications in monitoring of host organism exposure to harmful substances.
