Ant Systems-Based DNA Circuits

Abstract

A major challenge in the development of evolvable, autonomous, and programmable biomolecular machines is the introduction of the ability to cope with external changes. In the present study, DNA strand displacement was chosen as the main framework for modeling a DNA circuit capable of complex computational mechanisms such as decision making and reinforcement learning. Our goal was to design a DNA-based system that is adaptive and reactive to external stimuli. Our design is based on a computational algorithm inspired by a natural system, namely the ant foraging system, additionally consisting of geometrical components such as nanostructures or DNA origami. The correctness of our algorithm was verified in silico via quantitative measurement of reaction kinetics. Our results indicate that the circuit design responds correspondingly, regardless of the initial conditions under limited threshold (in contrast to the currently available DNA strand displacement systems). Furthermore, we discuss potential applications, which include decision making capable machine, and reusable DNA circuits.