No, this gibberish is not a misprint. This string
of letters A, C, G and T – which represent the four chemical bases or building
blocks of life – is a small portion of a human DNA sequence. Within this
entire sequence of 1,821 characters are genes, some of which are chemically
activated, and thus "turned on" or expressed. Expressed genes make RNA, which
in turn makes proteins that are the building blocks of cells and control their
function.
photo by Stanley Leary  |
|
Bioinformatics is the future of life sciences, says Georgia
Tech Prof. Mark Borodovsky. He created GeneMark, one of the most
used and accurate software programs for finding genes in genomic
DNA sequences. | |
Finding the genes and proteins they code for within this relatively small
sequence is a large task in itself. But magnify the problem to three billion
sequenced letters – the result of the Human Genome Project, a draft of which
was completed last year. How can scientists possibly manage this amount of
data – and more that is on the way – and then mine it for valuable genetic
discoveries to aid in diagnosing, treating and even preventing disease,
perhaps based on an individual's genetic makeup?
In a word, the answer is bioinformatics. Its definition varies, depending
on the scope of one's interpretation. To bioinformatics pioneer Mark
Borodovsky, a professor of biology and mathematics at the Georgia Institute of
Technology, the term refers to the interdisciplinary science that combines
mathematical, statistical and computer science methods to interpret biological
data and answer biological questions.
"Bioinformatics is the future of life sciences," says Borodovsky, who
created GeneMark, one of the
most used and most accurate software programs for finding genes in genomic DNA
sequences. "A wealth of new genomic data conceals endless gems of biological
knowledge and, with bioinformatics, researchers can generate new ideas and
test new hypotheses much faster."
The definition of bioinformatics has broadened in the past several years
from analysis of gene sequence data to the process of integrating many kinds
of biological data to find new meaning, says Jim Prestegard, a professor of
biochemistry and molecular biology at the University of Georgia.
"Now we're realizing that to have a full understanding of how biological
systems work and how they relate to disease, you need more than analysis of
genomic data," Prestegard says. "It is still necessary to deal with enormous
amounts of data, so bioinformatics is the same today in that sense.... But now
we need to integrate sequence information with information on gene expression
and the character of protein products. Ultimately, we need to integrate our
observations with physiological function of complete organisms and someday
with clinical records."
Whatever the definition of bioinformatics, most scientists agree
the field holds great promise not only for new discoveries, but also for
increasing the speed at which those findings are delivered. "The interesting
thing now is this paradigm shift in biology," says Scott Hemby, an assistant
professor of pharmacology and psychiatry at Emory University. "We've gone from
analyzing one gene at a time to tens of thousands in one fell swoop. This a
leap of science."
photo by Gary Meek 
|
|
Scott Hemby, director of the DNA Microarry Facility at
Emory University, is combining the power of bioinformatics with
microarray technology to find genes that are critical in disease
formation. Microarray technology allows researchers to analyze
10,000 to 20,000 genes simultaneously. (300-dpi
JPEG version - 417k) | |
This paradigm shift will make it possible to see major results from today's
basic scientific research in one-half to one-third of the time, Hemby adds.
This shift has not gone unnoticed by economic developers and entrepreneurs,
including those in Georgia. With the development of high-tech industry as its
ultimate goal, the Georgia Research Alliance
(GRA) plans to devote a significant portion of its proposed budget for fiscal
year 2002 to a bioinformatics initiative. Over the next two years, the money
would fund 10 to 12 new "GRA Eminent Scholars," equipment and facilities at
six Georgia research universities – Georgia Tech, the University of Georgia,
Emory University, the Medical College of Georgia, Georgia State University and
Clark Atlanta University.
Already, the GRA has funded 12 Eminent Scholars and their labs to do
research associated with bioinformatics and life sciences. Plans call for the
creation of a Center for Bioinformatics Research, which will support life
science studies in all GRA institutions. Meanwhile, Georgia Tech is preparing
students for careers in bioinformatics with a master's degree program that
began in 1999. Funded by the Alfred Sloan Foundation, it was the first such
program in the nation.
"We're populating our universities with the best and brightest scientists
who are doing cutting-edge work that will lead to the best scientific
discoveries at our universities," says Mike Cassidy, president of the GRA. "In
turn, that will become a magnet that will attract more companies in genomics
and bioinformatics to the state."
To plan its bioinformatics initiative, the GRA staff talked at length with
its current Eminent Scholars. "We want to know what drives them, what their
passion is about, what their science is and what they think that science can
be applied to," Cassidy explains. "That's how we're going to build a whole new
industry in the state of Georgia. We don't have a lot of traditional
biotechnology industry in the state, but we're convinced that this is going to
be very big. Unless we do something very aggressive and very bold, we'll be
left behind. So we're moving rapidly to make sure we're making the right
investments in the state."
Their efforts seem to be paying off already with the development of
biotechnology business incubators, the rise in start-up companies – some of
which academic researchers have founded – and the recent moves of two
established companies to Georgia. EmTech Bioscience is a commercial research
and development center focussed on the life sciences. EmTech, a partnership
between the GRA and Georgia Tech's Advanced Technology Development Center, is
operated by both Emory and Georgia Tech. NuTec Science, an established
bioinformatics and supercomputing company, moved to Georgia last summer and
became the anchor corporation at EmTech.
Meanwhile, the Center for Applied Genetic Technologies (AGTEC) at the
University of Georgia in Athens is an incubator for agriculture-related
biotech companies, including start-ups ProLinia and AviGenics. Also, Merial
Ltd., a major manufacturer of animal vaccines, recently relocated its
headquarters to Atlanta. Its plans call for the construction of a pilot
vaccine development plant in Athens. Both animal and human vaccines will be
manufactured there, bringing new high-tech jobs to Georgia.
"Georgia's initiative in bioinformatics has a huge amount of potential,"
says Gary Schuster, dean of the College of Sciences at Georgia Tech. "It's
interesting that five years ago, almost nobody knew what bioinformatics was.
Now we are charting its course, and that can be difficult. But almost everyone
agrees that bioinformatics is at the core of all the promises the 'Biological
Revolution' has to offer. It is being able to manage huge data sets, being
able to mine huge data sets and determine what's important and what's not,
whether it's for drug discovery or fundamental knowledge."
Many researchers – from molecular biologists to bioengineers to
animal scientists – are fascinated by the prospects of bioinformatics and
dependent upon its analysis of biological data. Few are developing the tools
of bioinformatics. Among those is Borodovsky with the creation of GeneMark
software and the latest version GeneMark.hmm (for the Hidden Markov Model upon
which its algorithms are based). The former has been used to completely
decipher the genomes of dozens of bacterial species, including the infamous E.
coli.
photo by Gary Meek 
|
|
Robert Nerem, director of Georgia Tech's Petit Institute
for Bioengineering and Bioscience, believes the next generation of
bioinformatics tools, available in five to 10 years, will be based
more so than current ones on predictive mathematical models. (300-dpi
JPEG version - 435k) | |
Both versions can analyze DNA sequences from higher organisms, including
humans, but the latter more accurately predicts the boundaries between introns
(non-coding regions) and exons (regions coding for proteins). In higher
organisms, a gene starts in a sequence and then can be interrupted by an
intron. Then the gene proceeds again with several more interruptions.
A Belgian researcher's study published in the November 1999 issue of the
journal Bioinformatics found that GeneMark.hmm was the most accurate among
several other state-of-the-art programs in predicting genes in plants.
GeneMark is licensed by about 150 companies and institutions worldwide,
including Harvard, Stanford and Columbia universities, the National Center for
Biotechnology Information, the Max Planck Institute and the European
Bioinformatics Institute. Others use GeneMark via the software's Web site
servers at Georgia Tech and in the United Kingdom. The U.S. server analyzes
about 3,000 sequences a month.
Borodovsky's research is funded by the National Institutes of Health
through 2003. In the next several years, he plans to improve and diversify
GeneMark, transforming it into a whole family of academically based software
programs that are regularly updated and widely available. Borodovsky founded
Gene Probe, a small spin-off company that provides technical support and
distribution for GeneMark programs.
The next generation of bioinformatics tools, available in five to 10 years,
will be based more so than current ones on predictive mathematical models,
says Robert Nerem, director of both Georgia Tech's Petit Institute for Bioengineering and
Bioscience (IBB) and the Georgia Tech/Emory Center for the Engineering of
Living Tissues. Better tissue substitutes will depend upon these models, which
should predict function of these materials both in the laboratory and the
body.
IBB researchers are developing such models, as are a host of academic and
corporate researchers across the nation. At Georgia Tech, involved faculty
members are: Gang Bao, Robert Guldberg and Joe LeDoux in the Georgia
Tech/Emory University Department of
Biomedical Engineering; Ray Vito and Cheng Zhu in the School of Mechanical
Engineering; Athanassios Sambanis in the School of Chemical Engineering; and,
of course, Borodovsky.
Meanwhile, researchers at GRA institutions are using the current
power of bioinformatics to make valuable leaps forward in basic science,
biotechnology and medicine that could revolutionize diagnosis and treatment of
disease. Much of this work is collaborative, particularly in biomedical
engineering, because of degree program and research partnerships between
Georgia Tech and Emory.
photo by Gary Meek 
|
|
The forces that control the structure of DNA – and
ultimately chemical activation of some genes – are a hotly debated
scientific question. Georgia Tech Professor Loren Williams is
addressing these questions in his structural biology laboratory.
(300-dpi
JPEG version - 495k) | |
"Tech researchers need Emory scientists, who understand what the
applications should be," says Don Giddens, chairman of the joint Department of
Biomedical Engineering and a GRA Eminent Scholar. "Conversely, Emory
researchers need the people who understand the technology and who can do the
modeling."
Moving the department's research beyond bioinformatics is Giddens' goal.
"Engineering is a field that implements things," Giddens says. "So that's our
connection. We want to plug into bioinformatics and be a part of it, but we
are stressing how to move that information into the medical and clinical
arenas."
Plans call for a GRA Eminent Scholar in the joint department; this
researcher would focus on Giddens' vision of "bioinformatics and beyond."
Giddens also cites two important ongoing projects. They are: (1) a basic
scientific study by Gang Bao of the physical forces that change or disrupt
DNA; (2) a technology under development by Joe LeDoux for delivering genes for
gene therapy.
Researchers in the Georgia Tech School of Electrical and Computer
Engineering also depend upon bioinformatics. Professor William Hunt is
developing microelectronic biosensor devices for molecular detection of target
molecules. Knowing the target molecules around which to design the biosensor
is the product of biomedical research and bioinformatics.
One application of Hunt's research could be for rapid protein screening.
For cancer screening, a biosensor would know the proteins to search for in a
urine sample. The result might be an early detection of a specific cancer.
Hunt's current devices can detect as little as a 3-hertz shift in a
250-megahertz resonant frequency of a surface acoustic wave biosensor. The
shift occurs when molecules bind to the surface of the device. That
sensitivity level is equivalent to about 10 parts per billion, but Hunt does
not know if it will be sufficient for detecting proteins for early cancer
screening. To that end, he is evaluating the acoustic design of his devices,
as well as the design of the chips on which the microelectronic system
resides.
While he awaits more information on the appropriate target molecules, Hunt
is focusing on the development of increasingly more sensitive devices. "What I
am working on has to be part of a collaborative effort," Hunt says. "I can
build sensitive devices, but unless I know what target molecule to look for, I
am just an electrical engineer building neat stuff."
Just down the hall from Hunt, Assistant Professor Bruno Frazier has begun a
five-year collaborative research project funded by the National Institutes of
Health. He and his collaborators from Princeton University, the University of
Virginia and Agilent Technologies are miniaturizing sample preparation for
biological and biochemical analysis. Specifically, Frazier is developing ways
to prepare blood for genetic analysis. The technology will allow scientists to
take a blood sample and extract the important components, such as DNA, from
it.
"This project is one component in a total analysis system that includes
sample prep, an analysis component and a detection/bioinformatics component,"
Frazier explains. ".... Ultimately, this system could be used for minimally
invasive diagnosis or treatment of certain types of cancers or other
diseases."
Frazier's work occurs at the interface of electrical systems, such as an
integrated circuit, with biological components, such as cells. Specifically,
he approaches the sample prep system's task of cell sorting and selection
through electrophysiology, the study of electrical properties of cells. In the
case of blood, Frazier uses electrical components, such as a conductor, to
measure its passive electrical properties (e.g., resistance and capacitance).
He has found that certain types of cells have certain electrophysiological
signatures. The micro system device can use those signatures to sort red and
white blood cells and to sort and detect certain types of cancer cells.
Researchers will use the tools of bioinformatics to analyze data gathered
by Frazier's micro system. Then they can visualize the data with a contour
plot, for example. This process is what gives the data meaning, Frazier says.
In the project's final phase, researchers plan to build a portable biochemical
analysis device that will incorporate Frazier's system.
Researchers at Emory University are combining the power of
bioinformatics with DNA microarray technology to find genes that are critical
in disease formation. DNA microarray technology allows researchers to analyze
10,000 to 20,000 genes simultaneously. At the GRA-funded DNA Microarray
Facility at Emory, researchers put about 10,000 spots of DNA on a microscope
slide. Then through a process called reverse transcription, researchers label
RNA and hybridize it with DNA on the slide.
"It's like putting a puzzle together," explains Hemby, director of the
facility. "All the pieces are there. You just have to put them back together
in the right places."
Scientists then bombard the slide with light under a microscope that is
like a laser scanner. What is emitted is either green or red. Using the tools
of bioinformatics, researchers then analyze the intensities of emitted light
to determine the abundance of genes and possibly whether the genes are
mutated, either of which could be indicators for disease. Many diseases have
genetic components; these include certain types of cancers and psychiatric
disorders, such as schizophrenia, autism and drug abuse. Hemby is studying
psychiatric disorders.
The evaluation and results of microarray experiments are what Hemby and his
colleagues call a gene expression profile. They have patented a technique for
creating the profiles for schizophrenia and drug abuse.
"When we learn more about diseases, we will be able to diagnose based on
what genes are expressed," says Mike Kuhar, a GRA Eminent Scholar and chief of
the Neuroscience Division at Emory's Yerkes Primate Research Center.
"So the genetic profile provides us with a bull's eye to shoot at with
medications. What medication may be right for me might not be right for you."
For example, Hemby is collaborating with Emory Professor and GRA Eminent
Scholar Rafi Ahmed to identify a subset of genes responsible for HIV patients'
tolerance to medications called highly active antiretroviral therapies
(HAART). After the researchers identify the subset, Hemby's lab will use a
smaller candidate gene chip to expedite this approach of Ahmed's research.
Eventually, it may be possible for patients to use a self-test device to see
if they are developing drug tolerance.
This is an example of the individualization of medicine, Hemby says. Other
Emory researchers are working to tailor medicines to best treat certain types
of cancer. Hemby plans to use microarray and bioinformatics technologies to
create a database of genetic profiles for a variety of psychiatric diseases,
all of which scientists believe will yield a specific expression profile.
In other research at Emory, Kuhar and his colleagues used bioinformatics
technology to study a new neurotransmitter in the brain; it is called the CART
peptide for Cocaine and Amphetamine Regulated Transcript. CART helps control
food intake and is partly responsible for the feeling of satiety. The finding
may eventually lead to new medications for obesity.
"Imagine the power of bioinformatics," Kuhar says. ".... Without a single
experiment in the laboratory, just by sitting at a computer, the uncovered
information at the terminal drives the direction of subsequent research.
Bioinformatics told us that CART was a neuropeptide, so our research went in a
different direction than it would have if we found that CART was another kind
of protein."
Animal scientists have also discovered the power of
bioinformatics, as well as microarray analysis. At the University of Georgia,
Associate Professor Steve Stice is cloning livestock to make more copies of
the best animals. Eventually, he hopes to make genetic changes in the clones
to produce animals that are more disease resistant, use fewer resources and
are more environmentally friendly (e.g., manipulating the genes that control
animal odors).
photo by Rick O'Quinn  |
|
At the University of Georgia, Associate Professor Steve
Stice is cloning livestock to make more copies of the best
animals. He says bioinformatics helps his research team determine
what genetic information is useful.
| |
"Scientists are finding more genes all the time, so there are many
possibilities for what we can do," says Stice, also a GRA Eminent Scholar. "I
like to say the pigs are piggybacking on the mouse and human genome projects.
We don't know as much about the genes of pigs and cattle, but we are learning
a lot more."
In fact, the Human Genome Project has revealed that there is an 80 to 90
percent similarity in genetic makeup between humans and farm animals. The
difference occurs in how those genes are expressed, and in what and how many
proteins are produced, Stice explains.
"That's where bioinformatics comes in," he says. "It's not just the ability
to know the genes, but to know how they interact in the body. That's the
really important part."
Stice is collaborating with Hemby at the DNA Microarray Facility to
determine why some cloned cattle embryos develop better than others. Now, only
one in 50 cloned embryos eventually produce offspring. Microarray technology
will allow Stice to compare gene expression in embryos that succeed in
development and those that don't. Then he can devise a genetic technique for
dealing with that difference, he says.
"This work is still in a very early stage," Stice says. "We are still
learning how to use this technology with cattle. But bioinformatics is so
important. It will help us determine what genetic information is useful."
In the vitally important arena of basic scientific research,
bioinformatics is also playing a role. At Georgia Tech, Loren Williams in the
School of Chemistry and Biochemistry is focusing on structural biology. Genes
are expressed in both the sequence of DNA bases and the three-dimensional
shape of DNA. The forces that control the structure of DNA are a hotly debated
scientific question – one that Williams is addressing. Scientists do not yet
understand how a linear meter of DNA gets folded to fit into a tiny cell
nucleus, or how a cell recognizes which part of the DNA to unfold and read.
Williams uses X-ray diffraction to help answer these questions. In X-ray
diffraction, the researcher takes a sample protein or piece of DNA or RNA,
purifies it and then crystallizes it – essentially dehydrating the sample in a
chemical "soup." At that point, he exposes the sample to an X-ray beam and
collects the diffraction pattern. Some of this work is done on the synchrotron
beam lines at Argonne and Brookhaven national laboratories, which provide
especially brilliant X-ray sources. Cameras equipped with light-sensitive
integrated circuits called charge-coupled devices collect the X-ray
diffraction patterns from the synchrotron, and high-speed computers process
the data. It takes about three months for scientists to determine the
three-dimensional structure of a macromolecule by X-ray diffraction. Within a
couple years, the process will take only three days and eventually only three
hours or so, Williams predicts.
photo by Gary Meek 
|
|
Georgia Tech Professor William Hunt is developing
microelectronic biosensor devices for molecular detection of
target molecules. One application of Hunt's research could be in
rapid protein screening for early cancer detection. (300-dpi
JPEG version - 990k) | |
The slow part of Williams' research is interpreting the three-dimensional
structures, trying to understand molecular interactions. For example, he is
interested in why a DNA macromolecule takes on a particular shape. "The
structural information is starting to come in to us too fast," Williams says.
"Now we have a bioinformatics problem. Bioinformatics gives us a way of asking
the whole database these questions."
Learning why macromolecules take particular shapes is potentially useful in
application-driven research such as drug design. But the information is also
critical for understanding fundamental phenomena in DNA bending and protein
folding. The applications of such basic research are never obvious, but
neither were the applications of the Internet when it was first built,
Williams says. Quoting from the movie "Field of Dreams," Williams hastens to
add, "If you build it, they will come."
"Basic research leads to unexpected things," Williams says. "Bioinformatics
has given scientists huge databases to which they can ask questions. We are
only limited by our imaginations."
At the University of Georgia, structural biologists use both X-ray
diffraction, or crystallography, and nuclear magnetic resonance (NMR)
technology to study proteins. They are collaborating with researchers from
Georgia State University, the University of Alabama at Birmingham and
Huntsville, and a Huntsville company called Research Genetics to form one of
seven recently funded NIH pilot centers for structural genomics. Eminent
Scholar and Professor B.C. Wang is the lead investigator, and GRA Eminent
Scholar Prestegard is one of seven other co-investigators. The goal of the
center is to provide structural information on proteins in a volume useful to
bioinformaticists, Prestegard says.
"We're working with a very new paradigm in science," he explains. "....
Historically, structural biologists have worked from a recognized biological
function back to fundamental information, such as the three-dimensional
structure of proteins and gene sequences. Now, we are almost turning that
approach upside down. The philosophy of structural genomics is to start from
the genes.... Even if we don't know the function of the protein, we can
crystallize it and use X-ray crystallography or NMR to get structural
information on it."
X-ray crystallography delivers about 80 percent of structural protein
information; NMR provides the remainder. Prestegard uses NMR, the same
technology used in biomedical imaging, but on an atomic scale. Typically,
determination of a complete protein structure (i.e., knowing where every atom
is placed) using NMR would take three to four weeks of data acquisition and
then a lot more time for data analysis. Instead of pursuing this course,
Prestegard's research team developed an efficient technique for studying only
the "backbone fold" of a protein. They believe the backbone fold's
restrictions on placement of the various chemical side chains will help
identify a protein's function. The research is helped, in part, by the highly
sophisticated laboratory equipment purchased with funds from the GRA.
"We believe this technique will be a very useful stepping stone that, along
with the tools of bioinformatics, will allow us to make functional predictions
from a more fundamental level," Prestegard says. "This technique is increasing
the rate of data acquisition from three to four weeks to just three to four
days.
".... There's a lot of work to be done," Prestegard says. "It's not just
getting the basic structure. We need to bring together information from a lot
of sources like gene transcription profiles and mass spectrometry data on
expressed proteins. In my view, that's where modern bioinformatics is going.
We're not only focusing on analysis. The real payoff is in integrating the
information. That's the frontier."
At the GRA, Cassidy wants Georgia on the bioinformatics frontier
and fast.
"There will be a rapid explosion in bioinformatics in the next five years,"
Cassidy says, "particularly in the tools needed to decipher and interpret
genomic data. Estimates indicate that bioinformatics will become a $49 billion
industry in the next four to six years."
photo by Rick O'Quinn  |
|
University of Georgia Professor Jim Prestegard and his
colleagues have developed an efficient technique for studying only
the "backbone fold" of a protein. They are using the technique and
the tools of bioinformatics to help identify protein function.
| |
Long-term economic development prospects related to bioinformatics are in
therapeutics and pharmaceuticals. Researchers in this arena are expected to
increase their drug discovery targets from about 400 to 4,000 over the next 20
years as bioinformatics matures and individualized medicine becomes
commonplace, Cassidy says. The pharmaceutical industry is expected to increase
from $300 billion a year to $300 trillion during that timeframe.
Georgia's academic researchers are excited by the prospects, too, but they
caution that many factors will affect the state's ultimate success in
bioinformatics-related economic development.
"There is a long way between knowledge accumulation and a medical treatment
revolution," Borodovsky says of the promises touted by the Human Genome
Project. "There has been a fundamental, revolutionary change in biological
science. The Human Genome Project has been a groundbreaking thing. Scientists
are changing how they do experiments, how they collect information before
doing experiments, how they interpret results. Everything is different from
the way it was 10 years ago.
"But there is still a process of hard scientific work to go inch by inch
with generating and testing ideas, and making sense out of data," Borodovsky
says. "It's a process of trial and error. It's not just a jump into a gold
mine."
Nerem believes Georgia could become one of several hubs for
bioinformatics-related industry. He is particularly encouraged by the
relocation of NuTec Science to Atlanta. More companies and more Eminent
Scholars, particularly in molecular and cell biology, are needed to build a
solid infrastructure, he says.
Others agree on the point of bringing more GRA Eminent Scholars to Georgia.
"We need more investment," Prestegard says. "But all of this takes quite a
while. Georgia is doing the right things. It has a tremendous advantage in the
GRA. Not many states have such a mechanism."
Williams hopes to see a "perfect synergy" between the long-term basic
scientific research of academia and the short-term applied research of
industry. "We can push the envelope of science if companies and academia will
share their resources," he says.
Giddens cites the complementary capabilities of the six GRA universities.
There are those developing the tools of bioinformatics, those gathering the
data, those analyzing and interpreting the data, and those who will integrate
the data with new tools of bioinformatics.
"We don't yet know the potential of bioinformatics," Williams says. "The
databases are getting bigger. We have new kinds of information. We are getting
better algorithms. It's inconceivable what we can learn with these new tools."
For more information, contact Robert Nerem,
Petit Institute of Bioengineering and Bioscience, Georgia Institute of
Technology, Atlanta, GA 30332-0405. (Telephone: 404-894-2768) (E-mail: robert.nerem@ibb.gatech.edu);
or Gary Schuster, College of Sciences, Georgia Institute of Technology,
Atlanta, GA 30332-0365. (Telephone: 404-894-3300) (E-mail: gary.schuster@cos.gatech.edu);
or Mike Cassidy, Georgia Research Alliance, 50 Hurt Plaza, Suite 980,
Atlanta, GA 30303 (Telephone: 404-332-9770) (E-mail: mcassidy@gra.org)
Contents
Research
Horizons GT
Research News GTRI Georgia Tech
Send questions and comments regarding
these pages to Webmaster@gtri.gatech.edu
Last updated: Feb. 16, 2001
