The structure/function relationship is fundamental to our understanding of biological systems at all levels, and drives most, if not all, techniques for detecting, diagnosing, and treating disease. The predominant means of collecting structure/function data in biomedicine is reductionist and has thus led to a proliferation of complex data (e.g., gene expression arrays, digital images) that capture only a fraction of this relationship because they are limited to one organizational scale each. For example, gene sequence and expression data capture the structure and activities of individual genes but fail to explain how these genes collaborate to control cellular and tissue-scale functions. As biomedical research moves toward more systems-based, multiscale approaches for uncovering structure/function relationships, we are hindered by the lack of integration between different forms of reductionist data. In this talk we will present novel techniques to vertically integrate the information from multiple scales to build data driven models. Our techniques are based on building graphs and extracting graph theoretical features for clustering, classification, and prediction. Hosts: Edward Knightly and A.C. Antoulas