Google, Microsoft May Help Usher in Personalized Medicine Wave, Says George Church
The genomic era hasn’t yet produced a revolution in personalized medicine, but it’s coming, says Harvard University geneticist George Church. Major tech companies like Google and Microsoft are making it their business to help people keep track of their health data—side-by-side with their genome sequence data (if they’ve got it). The adoption of these technologies has been slow to date, but combined with a new policy push for electronic medical records in Washington D.C., it just might move medicine away from one-size-fits-all approach that’s been the standard for so long, Church says.
That was the most interesting idea I picked up from talking with Church after he spoke at the Xconomy Forum on biotech innovation that we held recently at Biogen Idec headquarters in Cambridge, MA. People perked up their ears when Church talked about genomics and personalized medicine, since he’s one of the world’s leading thinkers on those topics, and has also worked hard to apply his ideas at a number of emerging biotech companies. The list includes Cambridge, MA-based Knome, a provider of genomic interpretation services; South San Francisco-based LS9, a renewable fuel company; and Mountain View, CA-based Complete Genomics, a gene sequencing company that has brought down the price of a genome to $5,000.
Here’s an edited account of my conversation with Church after the forum:
Xconomy: You mentioned earlier that you think Google and Microsoft are doing interesting things in terms of fostering greater usage of genomic data. How is that?
George Church: It’s not really about genomics so much as it is about personally controlled health records.
X: Ok, so what kind of impact does this have on your work?
GC: To the extent that these things are a controlled vocabulary, that’s important. To the extent that it makes people feel like they own their medical records, they can then share them more easily than filling out a form or asking the physician to give them something, and then getting a bunch of photocopied sheets that need to be transcribed onto a computer. This makes it much easier for people to share it for research.
I would love to see a wave of enthusiasm where people say ‘I’m going to share my genome and my medical traits, so that we can all benefit.’ Because right now it’s largely uninterpretable. But if everybody shares, it becomes interpretable. It greatly changes the ability to do research if the genome and the traits are both in the hands of the individual, and it really costs them nothing to push a button. But they need to think very deeply about what the consequences might be.
X: How might this work? With something like the $5,000 genome being proposed by Complete Genomics, does an individual go in to the doctor, get one of these sequences done, and dump all the results into Microsoft HealthVault or Google Health?
GC: That’s exactly right. And if you do a partial genome, it already costs even less than that. Very, very quickly—surprisingly quickly—people will be getting this data and they will want to put it together with all their other health records. Whether it’s separate or together, they’ll be able to share both, because it will be there sitting on their disk drive at home. It’s just a matter of condensing them so there’s benefit to them, their family, and to society.
X: How do you do that?
GC: We’re trying many different strategies. Ranging from 23andme, which gets people interested in social networking, family ancestry, and medicine in an educational format. That’s one experiment. Another one is the Personal Genome Project, which really has had a remarkable response. We just opened [a new phase of the project] on DNA Day, April 25, and we’ve already got 11,000 volunteers [for 100 slots]. That could easily grow and flower in unexpected ways that we’ve seen happen with various Internet phenomena like Facebook, Wikipedia, all sorts of crowdsourcing phenomena. We’ll keep trying experiments until we get one that hits. It’s extremely important.
Unlike most science, where you can buy your chemical and go into your dark room and work on it, this is something that has to have a vibrant, exciting interface with the public, or else you have nothing to work on.
X: You said before you were worried that not enough people will participate. How many people do you need to participate to create a really meaningful pool of data for people like you to run queries against it?
GC: We’re approved by our Institutional Review Board to do 100,000. I think lots of interesting things will happen before that. Lots of interesting things will happen at 100 or 1,000 people, or 10,000 people—every factor of 10. But more is better. If we got 1 billion people, that would we be way better than 100,000.
It’s such uncharted territory, we really don’t know what kind of good experiences and bad experiences people might have. But if a lot of people report good experiences, and they can see it percolate in society, then it will really take off. It may not require the usual kinds of markets and establishment support, in the same way that Wikipedia didn’t really require the usual kinds of financial returns. It became cheap enough that an occasional donation of time and money was adequate.
X: What’s your personal experience been like? Your genome has been out there in the public domain for a while. What’s happened?
GC: My medical records have been out since 1999. My genome has been out since earlier this year.
It’s still early days. The 10 members of the Personal Genome Project have their genomes and medical records online, and we find things. For example, one of them has a potential predisposition to hypertrophic cardiomyopathy. It’s not something where you’d ordinarily walk into a doctor’s office and say, ‘hey, test me for that, because gee, I’m feeling a little hypertrophic cardiomyopathy coming on today.’ The way you find out is you drop dead during sports or exercise. That’s kind of too late. So he’s going to be sure to get a checkout now.
Now, the Personal Genome Project is not only about delivering diagnostics, like in that case. We feel responsible, because if we turn something up, we should let them know. That sort of thing will greatly expand with time as we learn more.
X: Has there been any downside? Have any of your concerns about putting this data out in the open been confirmed, or maybe assuaged?
GC: One thing that’s happened since we put it up is the Genetic Information Non-Discrimination Act has passed. The downsides that we were worried about were that maybe people would over-interpret it. They might say ‘Gee, you have a gene here for silliness, or Machiavellianism,’ or something like that. If you were running for office, it might take a while to convince an uninformed public that people who make claims like this don’t know what they’re talking about.
We weren’t really worried about it happening right away. It’s not really an attractive honeypot to attract people who want to abuse it. To a large extent, I wasn’t that worried, and I think that’s true of a lot of the volunteers we have so far. We could be wrong. We certainly go to great extremes on our website, and the consent form, and the tests people take, to walk them through potential scenarios. I think most people who sign up recognize that they’re just scenarios and they’re actually fairly unlikely. There’s lower risk here than say, volunteering to join the armed services, or going into space flight, or driving a car, that a lot of people do anyway.
X: Is there something the President could do on this project to help? Have you talked to him about this?
GC: I have talked to his staff, when he was a Senator. He had a great, surprisingly good staff, a PhD and an MD, who were really on top of healthcare issues and genomics issues. He’s got all kinds of problems [to deal with], but he’s in a great position to make fundamental changes. He’s got a great science staff now, one of the best in the history of the United States. I know there’s some bills coming up to support health research.
I think there’s a movement that he could support for specifically targeting personalized medicine. It will happen anyway, but it would be nice if he got behind it. Basically, what we’re looking for now from an economic standpoint is something comparable to what happened in the Clinton Administration. They started out, the economy wasn’t so great. Because of the computing, and Internet revolution, things turned out great. There was a budget surplus. That could happen again this time around. It’s probably not going to be computers and the Internet again, this time it’s probably going to be biotech and medical genetics.