Ben Hirschler and Patricia Reaney
LONDON: News last month that scientists had built the first programmable
computer made from the molecule that carries our genes has brought the vision of
computing with DNA one step nearer.
Israeli researchers have developed a DNA computer so tiny that a trillion of
them could sit in a drop of water and perform a billion operations per second.
The idea of following Mother Nature's lead and using DNA to store and process
information took off in 1994, when Leonard Adleman of the University of Southern
California first used DNA in a test tube to solve a simple mathematical problem.
Since then a dozen research groups around the world have jumped into the
field -- which fuses biology and information technology -- in a bid to harness
the inherent ability of strands of DNA to perform trillions of calculations at
the same time.
That massive simultaneous problem solving at a nanoscale is a potential way
of getting round the limits of the silicon chip, which scientists believe cannot
be scaled down much further. The famous double-helix molecule found in the
nucleus of all cells can hold more information in a cubic centimetre than a
trillion music CDs, with data stored as a code of four chemical bases --
adenine, thymine, cytosine and guanine, or A, T, C and G.
These chemical "letters" like to link up with particular other
ones, which means strands with complementary letters stick together. These
linkages can then be "read" using naturally occurring enzymes, giving
scientists a way of finding hidden patterns in complex datasets.
The achievement of researchers at Israel's Weizmann Institute in getting DNA
to perform calculations automatically follows a breakthrough last year by a team
at the University of Wisconsin who successfully anchored strands of DNA to a
glass slide, opening the door to eventual DNA computer chips.
But harnessing DNA's potential as a microprocessor remains a challenge and
many scientists believe it will only ever complement rather than replace
silicon-based computers.
Hybrid machines
"I think in the future we might have hybrid machines that use a lot of
traditional silicon for normal processing tasks but have DNA co-processors to
take over specific tasks for which it is best suited," said Martyn Amos, a
lecturer at the University of Liverpool who wrote the first PhD in DNA
computing.
While a conventional desktop PC is designed to perform one calculation very
fast, DNA strands produce billions of potential answers simultaneously, which
may make them suitable for solving "fuzzy logic" problems that have
many possible solutions rather than the either/or logic of binary computers.
Although DNA offers an intriguing new medium for computing, scientists have
yet to crack the problem of "scalability," or the capacity to expand
to solve huge problems that existing computers now do.
Adleman's initial test-tube calculation solved the so-called "travelling
salesman problem" by working out the shortest route between seven cities
linked by 14 one-way roads. All possible permutations were created in a test
tube and the correct ones filtered out using biochemical reactions. The snag is,
the bigger the problem the more DNA you need -- and, with current techniques,
the numbers can get out of hand.
"It has been estimated that if you scaled up Adleman's problem to 200
cities from seven, then the weight of DNA required to represent all the possible
solutions would exceed the weight of the Earth," said Amos.
That is one the reasons why computer giant International Business Machines
Corp is focusing on other ideas such as carbon nanotubes and quantum computing,
based on atoms rather than biological material.
Clever cells
If it seems unlikely we will be popping down to the local PC store for a DNA
computer any time soon, the clever molecule may yet have other applications that
could bring the technology closer to the pharmaceutical industry.
Professor Ehud Shapiro of the Weizmann Institute, the head of the Israeli
research team, believes DNA nanomachines could in the future operate within
human cells, monitoring potentially disease-causing changes and synthesising
drugs to fix them.
"The living cell contains incredible molecular machines that manipulate
information-encoding molecules such as DNA and RNA (its chemical cousin) in ways
that are fundamentally very similar to computation," he said.
"Since we don't know how to effectively modify these machines or create
new ones just yet, the trick is to find naturally existing machines that, when
combined, can be steered to actually compute."
Another use could be in building diagnostic tests inside a "smart"
bacterium by re-engineering its genome to include a small logic circuit that
could, for example, be activated by the presence of a certain chemical. The
whole field of DNA computing remains at the very early
"proof-of-principle" stage but could start to become a reality in the
next five to 10 years, Amos believes.
(C) Reuters Limited.