George Chrch, visible man
George Church is professor of genetics at Harvard medical school and also heads the Lipper Center for Computational Genetics, MIT-Harvard/ Department of Energy Genomes to Life Center, and the National Institutes of Health (NIH) Center for Excellence in Genomic Science. Church’s Harvard lab is a member of the genome-sequencing technology development project of the NIH-National Human Genome Research Institute.
He’s also the inspiration behind a number of companies including Genome Therapeutics (since 1989), the sequencing part of which merged with Agencourt Biosciences in 2003, arguably the largest current-generation gene-sequencing company (recently acquired by Beckman Coulter for more than $100 million (depending on earn-outs) and a 2005 spin-off called Agencourt Personal Genomics. He has also contributed ideas to Codon Devices, a “synthetic biology” company that constructs large-scale integrated genetic circuits for anything from pharmaceutical manufacturing to biosensors and smart materials; Codon has funding from Kleiner Perkins (including Vinod Khosla personally) as well as Flagship and Alloy.
Most recently, Church led Harvard’s research project to design a faster and therefore cheaper way to sequence genomes, reported August 4 in the journal Science.. This followed by four days an announcement by company 454 Life Sciences, based in Branford, CT, of a similar achievement.
**Three magnitudes down, two more to go **
Both groups have automated and miniaturized the process, which makes it much cheaper, faster and more accurate. The very first human genome sequence took 13 years and cost $2.7 billion (though the second was much quicker!). The current cost – if someone wanted it – would be about $20 million, or the price of Dennis Tito’s trip into space. Church’s and 454’s separate but similar approaches drop the cost to about $2 million. Their ultimate goal is about $1000, though Church points out that even $20,000 would be compatible with our current medical system: not for everyone, but akin to a complex operation.
But 454 sells its equipment for $500,000, whereas Church’s group is aiming for a more “community-oriented” effort. Church expects his approach to be used by 454’s three major competitors (including Agencourt) and, indeed, labs all around the world. In fact, he says “You can use equipment that’s currently available in most labs, for about $150,000, starting with a digital camera and a microscope; everyone has those. “The science paper includes step-by-step instructions, although they may not be for “anybody”: The final words of the paper are: "We collected ~786 gigabits of image data from which we gleaned only ~60 megabits of sequence. This sparsity - one useful bit of information per 10,000 bits collected - is a ripe avenue for improvement. The natural limit of this direction is single-pixel sequencing, in which the commonplace analogy between bytes and bases will be at its most manifest."
“I like commerce,” says Church, who is loosely connected with some 20 companies as an advisor or scientific contributor (as well as more closely with Agencourt and Codon). “But here, commercially, we’re going to race to the bottom. We’ll run workshops and do everything we can to spread the technology. Agencourt may want to become Amazon or eBay and do useful applications, not make money on the "browser", which spreads freely academically. We wrote the paper to make it extremely enabling. It’s a total cookbook: where to order the parts, how to use them…the opposite of how a commercial entity would write a paper.”
He describes the process, which can use tissues as simple as blood or swabs from a mouth: The material is immobilized into beads on a slide, while various solutions flow through it slowly. The system works basically by matching fragments from the sample onto a reference genome. That is, you can’t do the first instance of any species’ genome this way. You have to start with the multi-million-dollar model. But after that, it’s more like checking a new document against a reference copy for subtle changes, or fitting complementary pieces onto a long jigsaw puzzle. “I’m mostly a scientist, but when I touch engineering I get this rush of excitement,“ says Church. “You have to ignore all the things you can’t do, and not beat yourself up over what you could have done long ago. It’s so arbitrary, but so wonderful when you just make something work!”
But Church isn’t content merely to create the technology; he understands that some people don’t want their genomes sequenced, and many more don’t want to share their information. “I see privacy attitudes in three buckets,” he says. “Some people want total privacy; they might not even want to know the information themselves. In the middle, people want just you and your own health-care provider to see it. And on the other end, a large group of researchers could see it. These views all exist; my colleagues think the norm veers towards the private.” But here too, Church is an engineer as well as a scientist, with his Personal Genome Project (see resources page for URL). “We don't seek controversy,” he says, “but we do seek a safe way to explore extremes in order to arrive at a reasonable middle.”
After 12 months Church’s Personal Genome Project (PGP) has been approved by the Harvard Medical School Internal Review Board, which vets the ethics of all human-subjects research proposals at the Med school. Its mission is not to expedite research but to ensure proper treatment of subjects, so this approval is a significant win.
The idea is very simple: to sequence the genomes of individuals – “ however many we can afford, initially” – and to publish them along with the full medical records of those individuals, publicly identified. Then, the idea is to see what kinds of activities and research the presence of such information will foster. And how will the individuals involved feel about it after the fact?’
Church has already gone ahead and put his own medical records online and will do the same with his genome; he’ll be the first research subject. “It’s already very useful,” he says cheerfully. “I was giving a medical seminar one day, and a hematologist in the third row told me I should get my cholesterol checked…He said ‘I looked at your Web page and you’ve been taking Lovastatin. You should have checked after 6 weeks whether it was working.’ And indeed he was right. I had expected the drug to take care of things, but my cholesterol level was up to 288. Following up on his suggestions brought it down to 150.”
The PGP study is carefully designed to meet ethical standards, and it’s a social as well as a technical experiment. The volunteers (we are one, tentatively, though we haven’t seen the fine print) don’t get free medical care, payment or any other benefits that might be considered coercion to say yes. The volunteers are not supposed to be representative, but rather to be articulate, well-informed people who will take the time and trouble to learn about the science and medicine behind the project, and to be spokespeople for genomic research and, by example, openness about medical matters. Yes, we can understand why a certain proportion of people might legitimately want privacy for themselves and family members, just as some people do about other matters, but openness should be perceived as, at best, generosity with data, rather than exhibitionism.
Says Church: “Ultimately, to do epidemiology and association studies we need genome and phenome data, both of which are currently expensive.. but both types of costs can drop dramatically. The phenome data costs could drop via data-mining in medical records and the genomics will drop via technology initiatives from the NIH & DOE. The more patients feel comfortable with the dual use of medical records for health care and research, the more everyone could benefit.”
He’s optimistic that he’ll get the volunteers, and he also expects a high-end, early adopter market to emerge, along with a more coerced group of people desperate to understand their own anomalous conditions. “I’m looking for Oedipus,” he says. “I don’t want them to poke their eyes out, but they must want to know everything. Consider how much people love their objects – homes and cars…. They could have the same fascination with their bodies and genomes.” Certainly, Ray Kurzweil and Larry Ellison come to mind, along with any number of age-defying Hollywood starlets.
“Imagine having 200 physicians,” he says. “Even in medicine, there’s the wisdom of crowds.
It could be a social phenomenon, which could be good: the weather, or football, or your genome. It’s up to the individual to learn to be witty about their genome. The first thing is to be brief. Find out what you share genetically. If you don’t want to bore them about cholesterol, find out what you do have in common, like kidney disease. And of course, sometimes there’s good news. You might find out that you can marry your cousin with no special risk.”