Three Pillars to Stochastic Control: Autoregulation, Noise and Feeback

Synthetic biology seeks to develop modular bio-circuits that combine to produce complex, controllable behaviors. Modern synthetic biology design processes have focused to create robust components to mitigate the noise of gene expression and reduce the heterogeneity of single-cell responses. However, deeper understanding of noise can achieve control goals that would otherwise be impossible. We explore how an “Optogenetic Maxwell Demon” could selectively amplify noise to control multiple cells using single-input-multiple-output (SIMO) feedback. Using data-informed [1] stochastic model simulations and theory, we show how an appropriately selected stochastic SIMO controller can drive multiple different cells to different user-specified configurations irrespective of initial condition [2]. We explore how controllability depends on cells’ regulatory structures, the number of cells controlled, the number of cells observed, and the accuracy of the model used. Our results suggest that gene regulation noise, when combined with optogenetic feedback and non-linear biochemical autoregulation, can synergize to enable precise control of complex stochastic processes.

[1] M Rullan et al., Mol Cell, 70, 2018
[2] P Szymanska et al., Phys Biol, 12, 2015