Designing robust synthetic circuits for cellular contexts is crucial in synthetic biology, but faces challenges due to low predictability and lengthy troubleshooting. This paper introduces Cyberloop, a testing framework that accelerates the design and implementation of biomolecular controllers. It uses real-time cellular fluorescence measurements fed to a computer simulating candidate stochastic controllers, which then compute control inputs and feed them back to cells via light stimulation. Experiments with yeast cells and optogenetic tools characterized various biomolecular controllers, tested the effects of non-ideal behaviors like dilution, and demonstrated functional improvements with network modifications. The analysis provides conditions for optimal biomolecular controller performance, minimizing biological implementation pitfalls.

Sant Kumar, Marc Rullan, Mustafa Khammash