Abstract
Controlling cells endowed with the genetic toggle switch has been suggested as a benchmark problem in synthetic biology. It has been shown that a carefully selected periodic forcing can balance a population of such cells in an undifferentiated state. The effectiveness of these control strategies, however, can be mined by the presence of stochastic perturbations and uncertainties typically observed in biological systems and is therefore not robust. Here, we propose the use of feedback control strategies to enhance robustness and performance of the balancing action by selecting in real-time both the amplitude and the duty-cycle of the inducer molecular signals affecting the toggle switch behavior. We show, via in-silico experiments and realistic agent-based simulations, the effectiveness of the proposed strategies even in presence of uncertainties and stochastic effects. In so doing, we confirm previous observations made in the literature about coherence of the population when pulsatile forcing inputs are used but, contrary to what proposed in the past, we leverage feedback control techniques to endow the balancing strategy with unprecedented robustness and stability properties. We compare via in-silico experiments different control solutions and show their advantages and limitations from an in-vivo implementation viewpoint.
