NEC scientists awarded patent for design of "DNA Turing Machine"

Princeton 18 September 1998 A computer which simultaneously can solve a series of any given problems through the use of DNA molecules is truly universal. In fact, the invention was revolutionary enough for awarding a US patent to Dr. Warren Smith and Dr. Allan Schweitzer, two scientists at the NEC Research Institute in Princeton. Their concept has been baptized the DNA Turing Machine in analogy with the original computer out of 1937, and according to the name of its designer, Alan Turing, a British computer scientist. Today's digital computers are all inspired on this model.

Advertisement

A computer which simultaneously can solve a series of any given problems through the use of DNA molecules is truly universal. In fact, the invention was revolutionary enough for awarding a US patent to Dr. Warren Smith and Dr. Allan Schweitzer, two scientists at the NEC Research Institute in Princeton. Their concept has been baptized the DNA Turing Machine in analogy with the original computer out of 1937, and according to the name of its designer, Alan Turing, a British computer scientist. Today's digital computers are all inspired on this model.

The first Turing Machine consisted of a virtually infinite length of tape, a read/write head, a method to place the tape whether left or right, and a unit to control both the tape movement and writing. The new DNA Turing Machine is able to explore the various possible computational outcomes to a given problem. As a result, the solving time is being reduced by a factor of N for an N-molecule/processor DNA computer. The machine for instance can start up a procedure in which DNA strands with enzymes are cut. The DNA subsequently is melted and bases are added to the strands.

The scientists Smith and Schweitzer inspired their invention on the work of Leonard Adleman. In 1994, this researcher at the University of Southern California achieved to solve a computational problem, using chemical units of DNA. In this way, Adleman was able to accurately calculate the shortest travel distance between a range of different cities on a map. Afterwards, it was Professor Richard Lipton at Princeton University who proved that the same set of primitive DNA operations allowed to solve a series of so-called satisfiability problems.

The two researchers at NEC now continue this earlier scientific work by trying to solve no matter which computational problem by means of DNA. The difficulty consists in scaling up the design in order to compete with the conventional silicon-based computers. How to avoid the exponential building up of errors, which in fact comes down to the decrease in yields over a sequence of chemical reactions? Even if a great number of DNA Turing Machines were connected in parallel, the different operations would still take too long to beat a traditional computer.

Smith and Schweitzer are therefore investigating the recent developments in biotechnology. The biological phenomenon, referred to as "RNA editing" seems to hold in store some promising potential. Profound research has shown that some microbes are able to "edit" their RNA, thus allowing characters to be cut, pasted, inserted, deleted, and overwritten. It is this type of mechanism that enhances the storage of the microbes' genes in a compressed form. Only if the need occurs to generate protein, the genes regain their full size. It is most likely that the editing operations suffice to implement a general-purpose computer. We just have to wait and see.


Leslie Versweyveld

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]