Programmable DNA-based finite automata

Abstract

Computation using DNA has many advantages, including the potential for massive parallelism that allows for large number of operations per second, the direct interface between the computation process and a biological output, and the miniaturization of the computing devices to a molecular scale. In 2001, we reported on the first DNA-based, programmable finite automaton (2-symbol-2-state) capable of computing autonomously with all its hardware, software, input, and output being soluble biomolecules mixed in solution. Later, using similar principles, we developed advanced 3-symbol-3-state automata. We have also shown that real-time detection of the output signal, as well as real-time monitoring of all the computation intermediates, can be achieved by the use of surface plasmon resonance (SPR) technology. More recently, we have shown that it is possible to achieve a biologically relevant output, such as specific gene expression, by using a reporter-gene as an output-readout.We cloned the input into circular plasmids, and thereby achieved control over gene expression by a programmable sequence of computation events. Further efforts are currently directed to immobilization of the input molecules onto a solid chip to enable parallel computation, where the location of the input on the chip represents specific tagging.