WEIZMANN INSTITUTE SCIENTIST DESIGNS THE FIRST GENERAL-PURPOSE MECHANICAL COMPUTING DEVICE
15 June 2000 - Weizmann Institute of Science

The first general-purpose mechanical computer designed for biomolecular and
pharmaceutical applications has been developed by Prof. Ehud Shapiro of the
Computer Science and Applied Mathematics Department at the Weizmann Institute
of Science. The mechanical computer will be presented today at the Fifth
International Meeting on DNA-Based Computers at the Massachusetts Institute of
Technology.

In terms of the logic behind it, Shapiro's mechanical computer is very similar
to biomolecular machines of the living cell such as the ribosome. Therefore, a
future biomolecular version of the device may ultimately lead to the
construction of general-purpose programmable computers of subcellular size. If
scientists succeed to build such a computer, it may be able to operate in the
human body and interact with the body's biochemical environment, thus having
far-reaching biological and pharmaceutical applications.

"For example, such a computer could sense anomalous biochemical changes in the
tissue and decide, based on its program, what drug to synthesize and release in
order to correct the anomaly," Prof. Shapiro says.

The Turing machine
Unlike existing electronic computers, which are based on the computer architecture developed by John von Neumann in the U.S. in the 1940s, the new mechanical computer is based on the Turing machine, conceived as a paper-and-pencil computing device in 1936 by the British mathematician Alan Turing. The theoretical Turing machine consists of a potentially infinite tape divided into cells, each of which can hold one symbol, a read/write head, and a control unit which can be in one of a finite number of states. The operation of the machine is governed by a finite set of rules that constitute its "software program." In each cycle the machine reads the symbol in the cell located under the read/write head, writes a new symbol in the cell, moves
the read/write head one cell to the left or to the right, and changes the control state, all according to its program rules.

Although the Turing machine is a general-purpose, universal, programmable
computer and is key to the theoretical foundations of computer science, it has
not been used in real applications. Shapiro's mechanical device embodies the theoretical Turing machine, and as such is a general-purpose programmable computer.

The mechanical computer
The device employs a chain of three-dimensional building blocks to represent
the Turing machine's tape, and uses another set of building blocks to encode
the machine's program rules. In each cycle the device processes one "rule
molecule." The device is designed so that the processing of the molecule
modifies the polymer representing the Turing machine's tape in accordance with
the intended meaning of the rule.

At the conference, Shapiro will present a 30-cm high plastic model of his
mechanical computer. He hopes that in the future, with the advent of improved
techniques for the analysis and synthesis of biomolecular machines, the actual
computer could possibly be built from biological molecules, so that it would
measure about 25 millionths of a millimeter in length, roughly the size of a
ribosome.

The computer and the ribosome
In fact, Prof Shapiro designed the mechanical computer with the ultimate goal
of implementing it with biological molecules. The computer is not more
complicated than existing biomolecular machines of the living cell such as the
ribosome, and all its operations are part of the standard repertoire of these
machines. These operations include the mechanical equivalents of polymer
elongation, cleavage and ligation, as well as moving along a polymer and being
controlled by coordinated structural changes.

The ribosome is the molecular machine of the living cell that performs the
final step of interpretation of the genetic code by translating messenger RNA,
which is transcribed from DNA, into protein. A key similarity between Shapiro's
mechanical computer and the ribosome is that a "program rule" molecule
specifies a computational step of the computer similarly to the way a transfer
RNA molecule specifies a translation step of the ribosome.

The computer is similar to the ribosome in that both operate on two polymers
simultaneously, and their basic cycle consists of processing an incoming
molecule that matches the currently held molecules on the first polymer,
elongating the second polymer, and moving sideways. However, unlike the
ribosome, which only "reads" the messenger RNA in one direction, the computer
edits the tape polymer and may move in either direction.
A future interactive biological computer

The computer design may allow it to respond to the availability and to the
relative concentrations of specific molecules in its environment, and to
construct program-defined polymers, releasing them into the environment. If
implemented using biomolecules, such a device may operate in the human body,
interacting with its biochemical environment in a program-controlled manner. In
particular, given a biomolecular implementation of the computer that uses RNA
as the tape polymer, the computer may release cleaved tape polymer segments
that function as messenger RNA, performing program-directed synthesis of
proteins in response to specific biochemical conditions within the cell.
Such an implementation could give rise to a family of computing devices with
broad biological and pharmaceutical applications.

http://www.weizmann.ac.il

More News:
  • For June 2000
  • From Weizmann Institute of Science
  • For University