ABSTRACT Pancreatic cancer is an aggressive, painful malignancy with a 5 year survival rate of less than 5% and median survival of about 4 to 24 months. The overall lack of success of chemotherapy treatments underscores, at least in part, the incomplete understanding of how pancreatic adenocarcinoma (PDAC) grows and invades local and distant tissues. There continues to be novel discoveries with respect to genetic mutations that predispose individuals to developing PDAC, histopathological features with prognostic implications, and clinical imaging (functional metabolic and anatomic, e.g. PET/CT) that define clinical stage and guide treatment options. However, overall, each diagnostic modality in large part, remain independent of one another. Consequently there is a need to assess and integrate measurements from the cell, tissue, and organ levels with multiple imaging modalities and to evaluate the concordance across different types of measurements and spatial scales in order to characterize the “molecular omic-type to clinical phenotype” relationship. Linking measurements from these different disciplines, radiology, pathology, and high-throughput molecular measurements, will help to establish more definitive phenotypes of PDAC that are consistent across multiple spatial scales. Recently many of the changes in PDAC have evolved around changes in the metabolic aspects of the tumor microenvironment and the interactions between tumor cells and the surrounding stromal cells, thus we focus on identifying metabolically driven/related phenotypes. We hypothesize that characterizing the PDAC microenvironment tumor-stroma interface at the cell and tissue levels will enable assessment of glycolytic and inflammatory characteristics of tumors in relation to specific histological tumor characteristics (e.g. perineural invasion and lymphovascular invasion) and this will lead to 1) a non-invasive imaging test to identify PDAC phenotypes and 2) new potential cellular targets for treatment. A data-driven systems biology approach using transcription-signaling-metabolic network reconstructions and constraint-based modeling will be used as the tie that binds a wide array of disparate data, including cellular microscopy, 3D preserved architecture tissue imaging, and PET/CT imaging in order to define a non-invasive functional, metabolic phenotyping measurement of PDAC that corresponds to cellular phenotypes. The constraint-based genome scale integrative modeling framework allows the simultaneous analysis of multiple types of data in interacting populations of cells. Additionally through the acquisition of multi-parametric cross- sectional imaging, it will be possible to evaluate whether or not metabolic changes on the cellular level have any manifestations on the tissue level. The environment, team, and resources assembled at UCLA are poised in a unique position to carry out these studies and achieve success in developing a non-invasive imaging approach for phenotyping P...