Pancreatic ductal adenocarcinoma (PDAC) is estimated to become the second cause of cancer-related deaths in the EU by 2040. Unlike other tumors, PDAC has shown no measurable improvement in 10-year survival in the last five decades, and the median survival after diagnosis is 8-10 months. While 90% of PDACs harbor mutations in KRAS and three tumor suppressor genes (CDKN2A, TP53, SMAD4), they display extensive intra- and inter-patient molecular and functional heterogeneity that is associated with clinical outcomes. Yet, PDAC functional heterogeneity is largely ignored in current diagnostic and treatment approaches. COMBAT-PDAC aims to leverage PDAC transcriptional and functional heterogeneity to develop and validate combinatorial strategies to target phenotypically diverse PDAC cell populations within each tumor. Our consortium will bring together a highly complementary set of competences and resources, including multiple omics and mechanism-focused functional genomics approaches, computational data mining and machine learning, production of innovative monoclonal antibodies (mAbs) and CAR-T cells, functional analysis in relevant preclinical models, molecular pathology and clinical expertise. Building upon our preliminary findings on subpopulation-specific cell-surface molecules and targeting agents, our specific aims are to: (i) identify and validate combinations of biomarkers and targets comprehensively covering heterogeneous PDAC cell populations (WP1); (ii) generate and characterise monoclonal antibody-based tools, including adaptable CAR-T cells, for combinatorial targeting of distinct PDAC subpopulations (WP2); and (iii) mechanistically define the role of targets in PDAC progression, possibly leading to the identification of novel mechanisms for diagnosis and therapy (WP3). Through this multidisciplinary approach, COMBAT-PDAC will identify actionable combinatorial biomarkers and develop theranostics agents targeting non-genetic heterogeneity in PDAC.