Engineering synthetic cells using next-generation robotics and machine learning

Bottom-up synthetic biology has ushered in a new era of synthetic cell science where biomimetic entities including microrobots are constructed from non-living and living components to create tailorable structural elements and complex life-like behaviours. Synthetic cells can model systems to unravel biological processes including signal transduction and have the potential to act as microrobots for therapeutics, agrochemical delivery, diagnostics and regenerative medicine. Although significant progress has been made in developing synthetic cells with individual behaviours such as motility, biosynthesis and communication, a crucial lack of high-throughput production and screening technologies hinders the design of next generation systems.

Here, we propose to unlock this potential by developing robotic synthetic cell production methods and coupling these with machine learning-powered feedback. Such methods will be used to design and develop a suite of new synthetic cells that are capable of sensing, computation and biosynthesis using a variety of molecular parts from membrane proteins to DNA circuitry. Integration of machine learning methods will facilitate rapid data analysis and inform future experimental design, unlocking new high-throughput production workflows. Such processes will be critical in translating fundamental synthetic cell technologies to tackle societal challenges in medicine and industry, acting as new delivery systems, microreactors and diagnostics.