External control of a genetic toggle switch via Reinforcement Learning

TL;DR

This paper presents a reinforcement learning-based method to externally control a synthetic genetic toggle switch, demonstrating its effectiveness through simulation and potential for real-world biological applications.

Contribution

It introduces a sim-to-real approach for controlling genetic switches using reinforcement learning trained on simplified models and applied to realistic biological systems.

Findings

Reinforcement learning successfully stabilizes the toggle switch in simulations.

The approach shows promise for in-vivo biological control.

Simulation results suggest practical applicability in synthetic biology.

Abstract

We investigate the problem of using a learning-based strategy to stabilize a synthetic toggle switch via an external control approach. To overcome the data efficiency problem that would render the algorithm unfeasible for practical use in synthetic biology, we adopt a sim-to-real paradigm where the policy is learnt via training on a simplified model of the toggle switch and it is then subsequently exploited to control a more realistic model of the switch parameterized from in-vivo experiments. Our in-silico experiments confirm the viability of the approach suggesting its potential use for in-vivo control implementations.