Church Hopes to Make DNA Decoding Accessible

Katherine M. Gray

George Church shows off the machines that analyze DNA sequences. He hopes to significantly reduce the cost of sequencing human DNA.

George M. Church wasted no time in revolutionizing the study of genetics.

In his sophomore year at Duke University in the 1970s, Church—who is currently a professor of genomics at Harvard Medical School (HMS)—designed a program that folded up one-dimensional DNA structures into cloverleaf-like structures.

The method he devised is “still the most common way people look at three dimensional structures,” says Church, who today operates a lab that analyzes and synthesizes genomes and specializes in technology development.

His discovery was significant not only for its scientific merits, but also because it relied heavily on the use of computers. And whereas his research was computer-aided, to that point “almost all the rest of biology was not,” he says.

Today, as director of Harvard’s Personal Genome Project (PGP)—which promises to make DNA sequencing a faster and more cost-efficient process—his quest to improve our knowledge of genetics with the aid of computer science continues.


Because only 0.1 percent of DNA accounts for the differences in human traits—including weight, hair color and disposition to diseases—scientists must decode millions of genomes before they can find the particular genes that relate to variations across individuals.

PGP, which was launched in January, will decode the genomes of 10 research subjects, including Church, this year.

Church hopes the project will allow his team to get “large information chunks” on the subjects by the end of the year. Their DNA has already been partially sequenced, and the first stage of the program is more than halfway complete.

The project is unique because the subjects will not remain anonymous—however, the results of their sequencing will be shown to both subjects and experts together before the subjects choose whether or not to release their results to the public. All of the people whose genomes are being decoded are required to have at least a master’s degree or above in genetics to ensure they understand the information they will be given.

“If there are risks, I’m in a good position to see it first,” Church says of his participation in the project, adding that there is usually a tendency in science away from self-experimentation, to prevent what he calls “a Dr. Jekyll and Mr. Hyde sort of thing.”

“But we’re not treating anybody here,” he says. “I think it has been very helpful to visualize it more thoroughly.”


As a Harvard Ph.D. student, Church was one of 12 scientists who proposed decoding the human genome back in 1984.

Despite the price tag—decoding a genome cost $3 billion by 2003—he said the effort was worthwhile because, like the computer industry, it was important to first make the scientific discovery and decrease the process’s cost afterwards.

“I think that was a reasonable balancing act,” he says.

Today, decoding a human genome costs $20 million, but Church says that upcoming technology may cause the price to fall as low as $200,000.

The ultimate goal for genomics researchers and businesses is to push costs below $10,000.

Church says that at that cost, individuals could afford to make a once-in-a-lifetime investment to have their genome decoded, get information on their genetic makeup, and perhaps ultimately prevent life-threatening diseases.

The human genome was first sequenced using an updated version of a method developed in the 1970s by Frederick Sanger. That method, while reliable and accurate, is also expensive and time-consuming. To mitigate the cost, some companies are currently using sequencing-by-synthesis methods, including the base extension method and Church’s preferred ligation method.

The idea behind driving down costs is similar to that in the computer industry. Church says that to decrease costs, genomists must decrease manual labor as much as possible and create denser, smaller samples of DNA and enzymes.

Both ligation and base extension focus “on one position at a time in a microbead full of identical DNA strands and millions of beads per microscope slide,” Church says.


Church expects PGP to expand and for millions of people to volunteer to have their genome decoded over the next decade.

He hopes that by the time this happens, the potential for insurers to discriminate against individuals with a disposition towards disease will be mitigated by the sheer number of people who have had their genome decoded.

Church adds that while the genome have been partially sequenced, several months of work still remain.

“We’re not at the stage where a highly-sophisticated analysis can be done. We’re getting there very quickly,” he says of PGP.

PGP, he says, is a type of hypothesis-driven research; the subjects may tell experimenters about a tendency towards motion sickness, for example, and scientists can then look at a possible correlation in that individual’s genome.

“An expert examining himself is going to be much more effective than examining someone else,” he says. “There’s this sort of introspection that can occur.”

He adds that while most medical research focuses on individual organs, genomics takes a more holistic approach.

“You should have personalized genomics, personalized physiology, personalized medicine,” he says. “Where each person’s different, and each body is an integrated whole.”

—Staff writer Katherine M. Gray can be reached at