Neural-network–like computation can be implemented directly within living cells using genetic, enzymatic, and CRISPR-based circuits. Recent work demonstrates that biomolecular networks can operate as linear and nonlinear classifiers, enabling pattern recognition. These results establish a foundation for neuromorphic computing using biological substrates.

Biomolecular circuits operate in environments where transcriptional and translational resources are inherently shared. This line of research examines how resource competition alters the behavior of protein and genetic neural networks. Sequestration-based designs reveal strategies that allow reliable computation even under strong cellular constraints.

We apply principles of feedback and adaptive control to regulate gene expression and cellular behavior. Our work focuses on biomolecular controllers that restore and stabilize dynamic behaviors, including sustained oscillations and cell-fate regulation, in the presence of uncertainty and disturbances.

Predictable biological behavior requires circuits with well-defined and tunable dynamics. Work in this area focuses on sequestration-based architectures that generate pulsing, oscillatory, and computational responses. These design principles support the construction of modular and scalable circuits for complex biological functions.