Providing tools for information retrieval in -omics databases, for organizing and navigating phenotypic data and for mining literature on proteins, lipids and metabolites.

Designing predictive models of diabetes complications based on phenotypic measurements and candidate biomarkers emerging from all SUMMIT work packages. Different modelling strategies such as SVMs and Bayesian Networks are being exploited to rank the genes and biomarkers from the datasets as biologically plausible candidate biomarkers.

Exploiting the in silico probabilistic models of diabetes complications to evaluate the impact of the candidate markers as surrogate endpoints in clinical trials and their effectiveness in drug registration trials.