for the fact that everything, including DNA and proteins, is made
from quarks, particle physics and biology don't seem to have a lot
in common. One science uses mammoth particle accelerators to explore
the subatomic world; the other uses petri dishes, centrifuges and
other laboratory paraphernalia to study the chemistry of life. But
there is one tool both have come to find indispensable:
supercomputers powerful enough to sift through piles of data that
would crush the unaided mind.
Last month both physicists and biologists made announcements that
challenged the tenets of their fields. Though different in every
other way, both discoveries relied on the kind of intense computer
power that would have been impossible to marshal just a few years
ago. In fact, as research on so many fronts is becoming increasingly
dependent on computation, all science, it seems, is becoming
"Physics is almost entirely computational now," said Thomas B.
Kepler, vice president for academic affairs at the Santa Fe
Institute, a multidisciplinary research center in New Mexico.
"Nobody would dream of doing these big accelerator experiments
without a tremendous amount of computer power to analyze the data."
But the biggest change, he said, was in biology. "Ten years ago
biologists were very dismissive of the need for computation," Dr.
Kepler said. "Now they are aware that you can't really do biology
Researchers have long distinguished between experiments done in
vivo (with a living creature) and in vitro (inside a glass test tube
or dish.) Now they commonly speak of doing them in silica — as
simulations run on the silicon chips of a computer.
There are computational chemistry, computational neuroscience,
computational genetics, computational immunology and computational
molecular biology. Even fields like sociology and anthropology are
slowly succumbing to the change. At the Santa Fe Institute, computer
models are used to study the factors that might have led to the rise
and fall of complex cultures like the Anasazi of Chaco Canyon and
Mesa Verde — a kind of artificial archaeology.
Scientists still devise hypotheses to be tested in the laboratory
or in the field. But a new step has been added to the scientific
process: More and more often, the experimental data that emerge are
used to generate computer simulations. A network of nerve cells or a
complex molecule comes to life as an animation on a phosphorescent
screen — to be electronically prodded and poked, manipulated with a
fluidity not possible in the real world.
In the course of this augmentation of the scientific mind, the
volume of data that needs to be analyzed has increased from a
trickle to a torrent, with physicists and biologists making the
Early last month, Brookhaven National Laboratory in Upton, N.Y.,
unveiled precise new measurements of something called the anomalous
magnetic moment of the muon. For months scientists gathered
information about how streams of these particles, ejected from an
accelerator, wobbled as they coursed around inside the world's
largest superconducting magnet — a donut-shaped ring more than 40
feet in diameter. Details aside, the take-home message of the
experiment was that the revered Standard Model, a longstanding
theory describing the particles and forces of the universe, may be
But reaching that conclusion required a monthlong computational
marathon in which more than a trillion bytes of data were processed
by a dozen computers. Then, just to be safe, the information was
processed again by another bank of computers using different
A trillion bytes is the equivalent of a thousand one-gigabyte
hard drives — hundreds of thousands of Napster downloads. But that
was just a fraction of the information needed to produce the
competing computer models of the human genome revealed the following
week by Celera Genomics and the
publicly financed International Human Genome Sequencing Consortium.
Generating Celera's computerized genomic map required
scrutinizing some 80 trillion bytes of data using what the company
describes as "some of the most complex computations in the history
of supercomputing." For this and other biological projects, Celera
has assembled what is believed to be the largest civilian
supercomputing operation in the world. The rival genome consortium,
which relied on less computationally intensive techniques, had to
yoke together 100 Pentium-powered PC's at the last moment to
assemble 400,000 snippets of DNA into its own picture of the genome.
When the number-crunching was done, both teams were surprised to
find that there may be far fewer genes than had long been believed —
30,000 instead of 100,000. The realization may lead to a rethinking
of how the complexity of life unfolds from the genetic code.
Physicists, more than biologists, have been accustomed to working
this way. Extreme computing has been an important part of their
field since the days of the Manhattan Project. Supercomputers at
government research centers, processing data at unprecedented
speeds, simulate some of the complexities of a nuclear explosion or
the impact of a meteor striking earth.
In more abstract realms, a whole field called lattice quantum
chromodynamics has sprung up, studying the strong nuclear force,
which holds together the nuclei of atoms, by modeling how quarks and
gluons cavort on a four-dimensional grid of artificial space and
time. In the grandest simulations of them all, cosmologists play
with computer models of the universe, tweaking the parameters of
creation and running the big bang again and again.
With the genome project, biologists are now upstaging everyone,
including physicists, in their sheer demand for computing power. And
reconstructing the genome is just the beginning. Figuring out how
the 30,000 genes, played like piano keys, give rise to the rhythms
and melodies of life is going to take even more calculating power.
Earlier this year Celera joined with Sandia National Laboratory in
Albuquerque, N.M., and Compaq computers to begin developing the
hardware and software needed to move into biology's next phase.
For years physicists have worried that many of the bright young
students who would once have joined the quest to discover the laws
of nature were being diverted instead into computer science. Last
month a leader in the software industry, Larry Ellison, the chief
executive of Oracle, predicted that the focus of the intellectual
excitement will shift again.
"If I were 21 years old," he said at a company conference in New
Orleans, "I probably wouldn't go into computing. The computing
industry is about to become boring. I'd go into genetic
Maybe it wouldn't matter. Whatever field he chose, he would
eventually end up doing computer science.