The article presents a computational approach for de novo drug design using deep interactome learning, which combines graph neural networks and chemical language models (CLMs). This method, named DRAGONFLY, aims to generate molecules with specific chemical and pharmacological properties without the need for application-specific reinforcement, transfer, or few-shot learning. DRAGONFLY leverages an interactome-based deep learning framework to create compound libraries tailored for specific bioactivity, synthesizability, and structural novelty. The study evaluates DRAGONFLY's performance in structure-based de novo design, specifically for generating ligands targeting the human peroxisome proliferator-activated receptor (PPAR) subtype gamma. The top-ranking designs are chemically synthesized and characterized, demonstrating potent PPAR partial agonist activity with favorable selectivity profiles and no significant off-target interactions. Crystal structure determination confirms the anticipated binding mode. The successful outcome advocates for the use of interactome-based de novo design in bioorganic and medicinal chemistry, enabling the creation of innovative bioactive molecules. The article also discusses the advantages and limitations of DRAGONFLY, including its ability to incorporate desired physical and chemical properties, and its potential for exploring new chemical spaces.The article presents a computational approach for de novo drug design using deep interactome learning, which combines graph neural networks and chemical language models (CLMs). This method, named DRAGONFLY, aims to generate molecules with specific chemical and pharmacological properties without the need for application-specific reinforcement, transfer, or few-shot learning. DRAGONFLY leverages an interactome-based deep learning framework to create compound libraries tailored for specific bioactivity, synthesizability, and structural novelty. The study evaluates DRAGONFLY's performance in structure-based de novo design, specifically for generating ligands targeting the human peroxisome proliferator-activated receptor (PPAR) subtype gamma. The top-ranking designs are chemically synthesized and characterized, demonstrating potent PPAR partial agonist activity with favorable selectivity profiles and no significant off-target interactions. Crystal structure determination confirms the anticipated binding mode. The successful outcome advocates for the use of interactome-based de novo design in bioorganic and medicinal chemistry, enabling the creation of innovative bioactive molecules. The article also discusses the advantages and limitations of DRAGONFLY, including its ability to incorporate desired physical and chemical properties, and its potential for exploring new chemical spaces.