In-vitro DNA-based finite state machine

Abstract

Molecular robotics and other autonomous molecular machines, like their conventional-mechanical counterparts, are expected to perform intelligent tasks under minimum external supervisions. One strategy to accomplish such complex design is to represent internal states of the machines by using finite state automaton (or also called finite state machine/FSM). The transition between states is triggered by external stimulus (can be input signals, sensing data from the environment, or the communication with other machines in the case of multi-agent systems). While there have been many proposals in implementing deterministic transitions by DNA reactions, e.g. by DNA strand displacement cascades, the experimental procedure still remains a challenge. Moreover, in this work, we also propose a new design for stochastic transitions, which also may be applied to arbitrary stochastic DNA computations.