The prediction of metabolic alterations on genome scale is still providng a challenge with respect to its application to medical research. While metabolomics approaches are on the rise, methods to either integrate these data with other omics-approaches or methods that are able to predict metabolic alterations based on expression changes are still missing or very limited. As disease dependent metabolic changes can be measured in the blood, the prediction of metabolic biomarkers would considerably improve minimal invasive disease diagnostics.