|
|
Deciphering the Genetic Code Researcher's computer program analyzes DNA from around
|
When Mark Borodovsky arrived in the United States from Russia in 1990, he
did not envision the American dream upon which he was about to embark.
See sidebar story:
A Genetics Primer
image courtesy of
National Institutes of Health
Georgia Tech professor of biology Dr. Mark Borodovsky deciphered
the complete bacterial genome sequence of Haemophilus influenzae -- the
structure of which is depicted here -- using GeneMark, a powerful software
program he developed. Now his research team has used the program to
annotate, in biological terms, more than 10 bacterial genomes, helping
to unravel the genetic code of these organisms.
A scant
eight years later, the Georgia Institute of Technology professor has developed the
world's most-used computer program for deciphering bacterial DNA; is developing
one for human DNA, and has seen his family flourish in his adopted country.
Indeed, most people born in the United States live their entire lives without being
profiled in Newsweek, which touted Borodovsky's field of bioinformatics as a "hot
specialty" and credited him with creating it.
See related story:
From Moscow to Atlanta
What Is Bioinformatics?
A few short years ago, scientists developed the tools to decipher long strands of
DNA into strings of their four underlying bases, the adenine (A), thymine (T),
cytosine (C) and guanine (G), that make up the genetic code. But it was arduous
work to take strings of bases and separate them into functional units, or genes,
which govern traits. While still living in Russia, Borodovsky decided the computer
would be a natural tool to manage vast amounts of genetic
information.
"There are certain mathematical models which help biologists do their work,"
Borodovsky says. "In 1985, while I was still in Moscow, I came up with the idea of
using Markov models to decipher genetic information."
The Russian mathematician, Andrey Markov, introduced his models early in
the 20th century. Borodovsky believed Markov models could portray genes by the
frequency of certain combinations of bases in known genes, contrary to non-genes.
Therefore these probabilistic models could be applied to DNA sequences to predict
where genes would lie on bacterial DNA.
Borodovsky did his initial work in Moscow from 1985-1989, laying the theoretical
groundwork for his model. Then his research stalled as he looked for biologists to
test his approach.
"The general economic situation and isolation of Russian scientists from the rest
of the scientific world was not conducive to testing my ideas," Borodovsky recalls. "I
needed research biologists who were sequencing DNA to compare my computer
predictions with experimentally verified genes."
Mark Borodovsky's research strives to annotate DNA after it has been sequenced.
For non-molecular biologists, that statement may be as clear as, well, the process of
using a computer to translate DNA into genes and proteins.
To fully understand the import of Borodovsky's research, one must have basic
knowledge of genetics. Following are a few concepts regarding some of the
overriding principles of this science.
Human cells contain 23 pairs of chromosomes. Each chromosome contains a
continuous double-helix strand of deoxyribonucleic acid (DNA). Four substances,
called bases, compose DNA: adenine, guanine, cytosine and thymine. In genetics,
their shorthand is A, G, C and T, respectively. The bases are bound in pairs, one of
each pair on each strand of DNA, in a precise manner (A binds with T, and C binds
with G).
Total human DNA is about 4.5 billion base pairs. These base pairs are divided
into 50,000 to 100,000 genes that control all aspects of the human condition, from
development to eye color to the origin of diseases.
Researchers now know that perhaps less than 5 percent of all of our DNA results
in genes. The rest of our DNA consists of base pairs that do not contain genetic
information and create gaps between and inside genes. These places of meaningless
DNA are called introns and intergenic regions. The actual gene portions of DNA are
called exons. Genes include more than genetic information; they also include codes
to signal their beginning and end, much like a capital letter clues the reader into the
beginning of a sentence and a period, the end.
Information stored in DNA is transferred to cell mechanisms producing protein
molecules via the processes called transcription and translation. In transcription, the
RNA copy of the gene containing DNA fragments is made, and the introns are
removed, leaving only the bases constituting genes. In an oversimplification, what's
left at this point is a string of letters that needs to be divided into meaningful words
to yield information. Through research, scientists figured out that three adjacent
bases are read together as a unit called a codon. Each codon corresponds to one of 20
specific amino acids, which are the building blocks of proteins. Finding genes,
protein coding regions, was the major problem in annotating DNA sequences up to
the size of a whole genome.
Borodovsky's computer programs take the DNA strand after molecular
geneticists have deciphered the order of the bases and interpret it into exons and
introns and make predictions about what proteins will result from this sequence.
His earlier work created a program to interpret bacterial DNA. Unlike human DNA,
bacterial DNA does not have introns inside of genes, making it simpler to predict
where one gene begins and ends. The ability to use computers to interpret DNA has
increased the speed and accuracy with which geneticists around the world are able
crack the genomes of various organisms.
The following year, Borodovsky made
a decision that would forever alter the
course of his life. In 1990, he and his family traveled to Atlanta, and he visited the
Georgia Institute of Technology.
"A scientific meeting at Georgia Tech made clear that my research work might
have a level of support incomparable to what I had back in Russia," Borodovsky
recalls. "Professor Roger Wartell, then newly appointed chair of the School of
Applied Biology, encouraged me to think about the perspectives that were just
unthinkable in Moscow. On the other hand, I saw that another dream might come
true. The tradition of religious freedom, along with the opportunities for education
for children, was something that the Soviet Union at that time was unfamiliar
with."
So instead of returning to Moscow at the end of the Atlanta visit, the
Borodovsky family decided to remain in the United States — even though they had
with them only the belongings they brought in two suitcases from Moscow.
The decision to live in the United States proved fruitful. Within one year,
Borodovsky met Dr. Fred Blattner from the University of Wisconsin. Blattner had
sequenced a significant portion of the DNA of Escherichia coli. This DNA piece
should have contained new genes, but their locations were not known. Borodovsky
analyzed the sequence using the early version of his computer program, called
GeneMark. The GeneMark predictions were later shown to be correct. In 1992,
convinced of the accuracy of GeneMark, Blattner employed Borodovsky's method to
analyze all of the raw DNA sequence data produced by his laboratory.
Another scientist, at Emory University in Atlanta, also was an early advocate of
GeneMark.
"In the early 1990s, I was sequencing a gene in the worm Caenorhabditis elegans,
and Mark Borodovsky contacted me to see if I was interested in testing his program.
This was a great opportunity to work with someone in town, especially since the
popular software program at the time was very difficult to use," recalls Dr. Guy
Benian, an assistant professor of pathology and cell biology at Emory. "GeneMark
gave very accurate predictions and was instrumental in annotating the gene."
The access to researchers, such as Blattner and Benian, who could test GeneMark
"was exactly what was missing in Russia," Borodovsky says.
By 1992, Borodovsky, in collaboration with James McIninch, an undergraduate at
Georgia Tech, had created a full version of GeneMark. When asked about the name
GeneMark, Borodovsky says it works on a number of levels.
"GeneMark marks genes, which appeals to biologists; it is based on the Markov
model, which mathematicians appreciate; and since my name is Mark, it is
meaningful to me personally," says Borodovsky, smiling.
Subsequent use of GeneMark showed it was a powerful tool for finding bacterial
genes. Researchers from around the world have sent their DNA fragments via e-
mail to the GeneMark e-mail server, which predicts locations of genes. After
mapping gene locations, the computer program compares the newly predicted
protein sequence to known ones in a database. This determines protein function.
The protein analysis is done in collaboration with the
National Center for
Biotechnology Information at the National Institutes of Health (NIH).
Publications about GeneMark in scientific journals caught the attention of
researchers at the Institute for Genomic Research (TIGR). The TIGR scientists were
pioneers in sequencing the complete genomes of numerous common bacteria.
Understanding the genomes of key microorganisms may increase understanding of
human genetics because lower organisms have some genes that correspond to
human genes. Also scientists can design new drugs based on knowledge of
disease-causing bacteria.
Borodovsky was asked to help decipher the first complete bacterial genome
sequences. GeneMark was used on Haemophilus influenzae, Mycoplasma
genitalium, Methanoccocus jannaschii
and Heliobacter pylori, helping to unravel
the genetic code of these organisms. Now more than 10 bacterial genomes have been
decoded, or annotated in biological terms, with the use of GeneMark. Also,
GeneMark has been used to annotate parts of genomes of other organisms,
including fungi, plants, insects, rodents and primates.
"GeneMark is faster and more efficient than other algorithms and is more
accurate than others in making predictions about where genes are," says Dr. Bruce
Roe, a professor of chemistry and biochemistry at the University of Oklahoma. Roe,
who holds the George Lynn Cross Distinguished Research Professorship, runs one
of the eight human genome centers in the United States. His lab is sequencing a
number of bacteria, in addition to human chromosomes. Roe uses GeneMark to
annotate the bacteria work in his laboratory because it is more than 98 percent
accurate, he says.
The Next Level: The GeneMark Family of Programs
The latest step Borodovsky has undertaken uses GeneMark Genesis as its base to
make even more sophisticated predictions — this time for the genomes of
eukaryotic, or higher organisms. (The cells of eukaryotes, including humans, have
nuclear membranes, paired chromosomes and complex cell division patterns.
Prokaryotes, such as bacteria, are single-cell organisms with no nuclear membranes.
They lack many of the more complex structures of eukaryotic cells, and divide
simply through such mechanisms as budding.)
"Deciphering bacterial DNA is simpler than deciphering human DNA since its
genes run continuously, without gaps. The genes of human DNA may be divided
into pieces, called exons, with non-coding genetic material between the exons. These
spacers in the genes, called introns, were hard to detect by a computer algorithm.
Also, eukaryotic DNA is much longer, with an average gene size of 10,000
nucleotides," Borodovsky explains.
Therefore, the predictions of where eukaryotic genes lie on a strand of DNA
must include predictions of the boundaries between the exons, which contain the
genetic information, and introns, which are the non-coding regions. To create a
computer program to achieve this, Borodovsky has employed another model, called
Hidden Markov Models or HMM. His most recent NIH grant will fund
incorporation of HMM into GeneMark, making the program responsive to the
boundaries between genes and introns. GeneMark.HMM was developed in
collaboration with Georgia Tech researcher Dr. Alexander Lukashin. The test of the
program demonstrated its "state-of-the-art accuracy," says Borodovsky, meaning,
when tested against current means of finding eukaryotic genes, GeneMark.HMM
performed at least as well as the best current methods.
GeneMark.HMM will fill a need, as evidenced by early demand from scientists.
Even before information about GeneMark.HMM has been published in a scientific
journal — the traditional method of disseminating information for the community
— almost 30 researchers have expressed interest to one of Borodovsky's graduate
students, John Besemer, who gave a poster presentation on GeneMark.HMM at a
recent conference on the eukaryotic organism Chlamydomonas reinhardtii.
A Start-Up Company
and a New Bioinformatics Degree Program
"The GeneMark programs are the type of research programs that should be
incorporated into the user-friendly environment that is easy to understand and
used by every biologist," Borodovsky says. "So we formed a start-up company to
make them more user-friendly."
The company, called GenePro Inc., supported by the NIH grant, will
commercialize GeneMark programs by making them more readily available and
portable to different computer platforms. The company will also provide technical
assistance and other services that cannot be done under the auspices of university
research.
While his company strives to maximize uses of the GeneMark programs,
Borodovsky is also heading Georgia Tech's new interdisciplinary master of science
degree program in bioinformatics. But his decision to add teaching to his busy
schedule was a challenge, he says.
"For six years, until 1996, I have had a strange feeling that I was not yet a full-
fledged part of the Georgia Tech community since it emphasizes teaching,"
Borodovsky says. "When I became a professor, the addition of teaching greatly added
to my busy schedule. But thanks to the strong moral support of Dr. Gary Schuster, I
was able to make this move. In the past two years, I have taught four new courses. It
was difficult, but a very satisfying experience. I am eager to see that students like my
lectures. It is not easy, and I should say that I do not have success all the time."
Now, Borodovsky is devoting time to the new degree program, which is being
supported by the Sloan Foundation. "I think that the major factor that will make
this program a success is the enthusiastic support of Georgia Tech administration
and faculty, including professors Anderson Smith, Roger Wartell, Robert Nerem,
Leonid Bunimovich and Sham Navathe."
This new degree program is expected to start in the fall of 1999. Then,
Borodovsky plans to move his lab to the new Parker C. Petit Institute of
Bioengineering and Biosciences Building, where space is specifically designed for
multidisciplinary research programs.
Bioinformatics employs mathematics, computer science and biology. It has
developed from the need to more quickly and efficiently manage and make sense of
the staggering amounts of genetic information contained in DNA strands.
Even while GeneMark was being used successfully to annotate the genes of
bacteria, Borodovsky was refining his program. GeneMark has been successful in
making predictions because it could "learn" based on previous knowledge,
Borodovsky says.
His next version, called GeneMark Genesis, became necessary
when TIGR scientists wanted to sequence the genome of the bacterium
Methanoccocus jannaschii, for which there were no experimentally studied
segments available to train the Markov models. The new program developed by
Borodovsky and graduate student William Hayes "learned Markov models from
anonymous sequences based on the grammar of the genetic code," Borodovsky
explains.
photo by Stanley Leary
Dr. Mark Borodovsky heads Georgia Tech's new interdisciplinary
master of science degree program in bioinformatics. The curriculum
development is supported by the Sloan Foundation.
Now with the whole family of GeneMark programs developed, Borodovsky,
along with his former undergraduate and graduate student Dr. James McIninch,
want to find a way to make these popular programs more accessible to the biological
science community.
For more information,
For more information, you may contact Dr. Mark Borodovsky,
School of Biology, Georgia Tech,
Atlanta, GA, 30332-0230. (Telephone: 404/894-8432)
(E-mail:
mark.borodovsky@biology.gatech.edu).
Contents |
Research Horizons |
GT Research News |
GTRI |
Georgia Tech
Send questions and comments regarding these pages to
Webmaster@gtri.gatech.edu