Autonomous forward inference via DNA computing

Abstract

Abstract Recent studies direct the research into building DNA computing machines with intelligence, which is measured by three main points: autonomous, programmable and able to learn and adapt. Logical inference plays and important role in programmable information processing or computing. Here we present a new method to perform autonomous molecular forward inference for expert system. A novel repetitive recognition site (RRS) technique is invented to design rule-molecules in knowledge base. The inference engine runs autonomously by digesting the rule-molecule, using a class IIB restriction enzyme Ppil. Concentration model has been built to show the feasibility of the inference process under ideal chemical reaction conditions. Moreover, we extend to implement a triggering communication between molecular automata, as a further application of the RRS technique in our model.