Molecular Computing Machines

Abstract

Biopolymers such as nucleic acids and proteins encode biological data and may be viewed as strings of chemical letters. While electronic computers manipulate strings of 0’s and 1’s encoded in electric signals, biologically encoded data might, in principle, be manipulated by biochemical means. During the last decade, several approaches to compute with biomolecules were developed, and the field has become known as biomolecular or DNA computing. The approaches varied widely with respect to the model of computation they employed, the problems they attempted to solve, and the degree of human intervention. One approach focused on the application of the Turing machine model and, more generally, string-processing automata to biomolecular information processing. Its goal is to construct computers made of biomolecules that are capable of autonomous conversion of an input dataencoding molecule to an output molecule according to a set of rules defined by a molecular program. Here we survey the field of biomolecular computing machines and discuss possible future directions.