Living cells can be engineered to perform neuromorphic computation using genetic circuits that emulate neural dynamics through regulation and molecular interactions. Our work combines simulation, experimental validation, and formal optimization to create scalable biomolecular neural networks capable of analog computation, feedback control, and self-adaptive learning.
We engineer morphogenesis as an executable genetic program, designing circuits that direct cell fate, differentiation, and self-organization into defined 3D structures. Through synthetic gene networks linking lineage specification to spatial patterning, our work has established design rules, computational frameworks, and genetic control systems that guide multicellular organization—from stem cell–derived tissues to programmable organoids for disease modeling and regenerative medicine.
We pioneered transcription-independent genetic control in mammalian cells, developing RNA-only logic circuits using RNA-binding proteins, miRNA-tuned elements, and self-amplifying RNA replicons that function in vivo without genomic integration. These advances—from model-driven replicon design and CRISPR- or small-molecule-regulated RNA circuits to clinically validated saRNA therapies—culminated in programmable RNA medicines such as STX-001, which has shown favorable safety and antitumor activity in Phase I trials.