DNA computing started in 1994 when
Leonard M. Adleman, a University of Southern California researcher,
solved a version of a mathematical problem called the “traveling
salesman problem” using DNA in a test tube. The problem is to find
a route that passes through a certain set of cities exactly once,
given the constraint that only certain routes between cities are
possible. This first DNA computer found the solution for a
seven-city version of this problem by simple trial and error, without
using the logic circuits found in traditional microprocessors.
The problem itself is not difficult to
solve -- conventional computers can solve the traveling salesman
problem for thousands of cities. Adleman's achievement was
significant because it demonstrated that computing can be done using
DNA molecules, rather than by electrical circuits.
Three years later, University of
Rochester researchers created simple Boolean logic gates made from
DNA. Logic gates are the basis of modern computers' processing
abilities, and fashioning them from DNA was a major breakthrough.
The development of logic gates opened the doors to sophisticated
DNA-based processors.
In 2002, scientists at the Weizmann
Institute of Science in Israel unveiled a working microprocessor made
not of silicon, but of protein structures and DNA molecules. A year
later, they had engineered a computer made of a single strand of DNA.
This amazing device was vanishingly small, since DNA is nothing more
than a large molecule, yet could perform 330 trillion operations per
second (0.33 PFLOPs). Compare this to the (current) world's fastest
supercomputer, the IBM Roadrunner, which can perform about 1.03
PFLOPs.
DNA computing is a form of parallel
computing, in which processes are undertaken simultaneously.
Conventional processors complete solve problems by completing their
constituent tasks one at a time. Parallel computers can dramatically
outperform conventional computers in solving some types of problems.
Silicon-based parallel computers are the cutting edge of
high-performance computing technology, but they are large and draw
enormous amounts of electrical power. Compared with traditional
parallel computing hardware, DNA computers have the advantages of
being much smaller and of not requiring electricity.
Just as traditional computers make use
of microchip hardware, DNA computers have their own version of
microchips, which are already in wide use. These devices are called
DNA microarrays or “gene chips,” and are manufactured by several
companies, including Affymetrix and Illumina. Gene chips evolved
from a DNA-analysis technique called Southern blotting. In a gene
chip, short segments of DNA called probes are attached to a glass or
silicon surface; these then “hybridize” (pair up) with fragments
of the DNA being analyzed. Gene chips are used in bioinformatics,
with applications ranging from forensic analysis to cladistics. In
computing, they may become especially useful for reading the output
of a DNA processor.
Because they utilize parallel
processing, DNA computers will likely be useful in fields that use
“fuzzy logic,” such as encryption. Specialized computers called
computational genes will have wide-ranging medical applications in
detecting and treating diseases. DNA may replace silicon as the next
king of miniaturization and speed in computing.