An approach to computational complexity: nondiscrete biomolecular computing

Discrete molecular computing devices based on digital device simulation and nondiscrete biomolecular computing devices based on the overall principles of information processing that use molecular-level mechanisms to create simple biomolecular systems are described. It is argued that discrete molecular computing only extends the choice of materials available in modern semiconductor device technology, but does not provide an opportunity to use physical mechanisms of information processing different from those available in conventional materials and devices. Nondiscrete biomolecular information processing devices effectively solve some problems with high computational complexity-for example, image processing operations such as contour enhancement and line extraction. The information processing potential of biomolecular systems is reviewed.<>