AbstractBackgroundGastric cancer is among the most lethal human malignancies. Previous studies have identified molecular aberrations that constitute dynamic biological networks and genomic complexities of gastric tumors. However, the clinical translation of molecular-guided targeted therapy is hampered by challenges. Notably, solid tumors often harbor multiple genetic alterations, complicating the development of effective treatments.MethodsTo address such challenges, we established a comprehensive dataset of molecularly annotated patient derivatives coupled with pharmacological profiles for 60 targeted agents to explore dynamic pharmacogenomic interactions in gastric cancers.ResultsWe identified lineage-specific drug sensitivities based on histopathological and molecular subclassification, including substantial sensitivities toward VEGFR and EGFR inhibition therapies in diffuse- and signet ring-type gastric tumors, respectively. We identified potential therapeutic opportunities for WNT pathway inhibitors inALK-mutant tumors, a significant association betweenPIK3CA-E542K mutation and AZD5363 response, and transcriptome expression ofRNF11as a potential predictor of response to gefitinib.ConclusionsCollectively, our results demonstrate the feasibility of drug screening combined with tumor molecular characterization to facilitate personalized therapeutic regimens for gastric tumors.