Abstract Pancreatic ductal adenocarcinoma (PDAC) is the 3rd leading cause of cancer-related deaths in the United States. This is largely due to the fact of high propensity of patients presenting at metastatic stages resulting from symptom ambiguity and a complete lack of general population screening programs. As the overwhelming majority of patients (85%) present with surgically unresectable disease, the 5-year survival rate is dire at only ~8%. Outside of what is known of the commonly mutated genes in PDAC, which include KRAS, TP53, SMAD4, and CDKN2A, little molecular information exists that has yielded clinical knowledge that may have an impact on survival outcomes. Therefore, a concerted clinical and research paradigm shift towards higher-resolution investigations of the genomics of PDAC is an area of highest public health significance. In our laboratory, we have leveraged single-cell sequencing approaches in both cell lines and patient-derived tissues to develop models for molecular underpinnings of PDAC and as potential tools for therapeutic stratification. First, to demonstrate the ability to model cellular heterogeneity of tumor populations, single-cell RNA and DNA sequencing was applied to widely used patient-derived PDAC cell lines. Using a bioinformatics pipeline, were able to differentiate between supposedly identical cell lines that had evolved rich subclonal architecture that may provide insight into complex in silico landscapes that naturally evolve as these cells are continuously used by laboratories. We then characterized single-cell gene expression profiles of patient-derived organoids (PDOs) to understand their heterogeneous composition and applied an in silico therapeutic drug prediction model to identify and validate combinatorial strategies targeting subpopulations. These present multiple phenotypic traits similar to that of the parental tumor including histoarchitecture, oxygen biology, epigenetic marking, and differentiation status. Orthogonal validation of predicted drugs was confirmed through a high-throughput drug screen of 764 agency-approved candidates on two PDOs. This strategy demonstrated the utility of single-cell transcriptomic profiling from PDOs in effectively identifying heterogeneous enriched pathways and molecular subtypes within cultures to predict viable treatment options using a pharmacogenomic approach. Citation Format: Maria E. Monberg, Vincent B. Pagan, Alexander Semaan, Jaewon J. Lee, Paola A. Guerrero, Senthil K. Muthuswamy, Anirban Maitra. Single-cell profiling reveals subclonal vulnerabilities to therapy in patient-derived 2D and organoid models abstract. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2019 Sept 6-9; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2019;79(24 Suppl):Abstract nr A33.