Personal Genome Project: Nature and Nurture, Warts and All
Baltimore neurologist Clifford Andrew, MD, PhD, has always been a fan of evidence-based medicine.
That's why, in 1977 and just 5 years out of medical school, he volunteered to participate in a major clinical trial, the Physicians Health Study II.
That's also why – fast-forward 35 years – he has joined the Personal Genome Project, which is aimed at melding the genetic and environmental data of 100,000 people in a publicly available database.
"It's an absolutely marvelous concept," Andrew told MedPage Today. If the project works as intended, he added, it will "be a trove, a wealth of information, for doctors around the world."
The idea is relatively simple. Take 100,000 Americans, sequence all 3 billion base pairs of their DNA, add in their medical records and environmental exposures, and compare with their phenotypic traits.
Make the resulting information available to anyone and everyone, with the idea that they'll use it to understand how genes and environment influence each other to produce an individual, warts and all.
It's the sort of holistic approach that should have been the norm since the modern age of genetics started, says George Church, PhD, of Harvard University, the project's principal investigator.
"In principle, we should have been doing it in the Human Genome Project," Church told MedPage Today.
But that enormously complex and costly project – like most of the genomic research since – focused on the DNA and not on the whole person.
With the breakthrough human genome map in hand, though, scientists have been eagerly exploring the world of human genes, but only rarely the much more complicated problem: the world and humans and their genes. Because of that approach – and with some exceptions – genetic findings tend to be limited to correlations between gene patterns and disease.
The Personal Genome Project is looking for harder data. "Our focus is on causality, not correlation," Church says. "But you need a cohort where you can study everything about a person as an individual, not just a de-identified piece of an individual."
It's that idea that drew Clifford Andrew to the project. And it's why he has tried to get his former classmates to join and why he hands out flyers for the project to some of his patients. "I think most physicians would understand the importance of this," he says.
Pulling Together the Cohort
Church came up with the idea in 2004 and it took a year before he could get review board approval for the first 10 volunteers. It took a couple more years to get approval for the full-scale 100,000-genome project. Now the project has about 2,400 entries – each representing a participant on its website.
And, the project is starting to metastasize. Led by University of Toronto geneticist Stephen Scherer, PhD, a Canadian version kicked off last year and there has been interest from other parts of the globe, Church says.
The first 10 American participants included some well-known names, including investor Esther Dyson, and Harvard psychologist Steven Pinker, PhD as well as -- not coincidentally – Church himself. He is PGP1 or – in the words of the project website – "guinea pig #1".
Sure enough, if you browse Church's online entry, you get an idea what he's about. There are reports on his genome sequence, his medical records (high cholesterol and cardiac arrhythmia) family history (maternal grandmother had gallstones), and a host of other individual traits. Multiply by 100,000 and you would have an enormous wealth of data.
But getting to that stage is not a quick or easy process – Church and his colleagues are still only a few steps along the way. Among the issues:
- How to organize the data. For example, medical records don't all follow the same format, so the project managers have had to develop supplemental questionnaires that allow medical histories to be associated with ICD codes.
- Privacy. Participants agree to have their data made public, although their names are usually withheld. But the investigators spent a lot of time making sure participants understand there's no guarantee that medical data won't be re-identified. Indeed, Canada's Scherer says the consent process takes longer than sequencing the exome – all the genes in the DNA – which can be done in about 2 hours.
- Cost.
The latter, though, is the "only roadblock really," Church says. Currently, it costs between $1,200 and $2,000 to do the genetics for an individual – roughly a million-fold decrease since the $3 billion Human Genome Project. The per-genome price continues to fall, but it's still significant, so the project relies on support from individuals and from some nonprofit organizations. As the utility of the data set increases, he says, it's likely that support will grow.
"We've never been in a big rush," Church says.
Interestingly, the Canadian project is on a slightly different footing. Lacking grant support – at least so far – Scherer and colleagues are asking would-be participants to cover the costs. You'd think that would be even more of a roadblock than it is for the U.S. project.
But, in the wake of a national newspaper article about the project late last year, about 500 people volunteered, all willing to pay their own way and many offering to pay for someone else. "The numbers are pretty impressive," Scherer told MedPage Today. "We have more than enough to get going and keep us busy for a while."
Benefits Already
There have already been some therapeutic benefits for participants, although that is not the main goal. Church himself, for instance, is on medication for high cholesterol. An email from a physician who had read his file suggested that – based on his reported diet – he should be doing better. Perhaps his medication was not right for him? After checking, Church switched drugs and has seen his levels drop.
Another participant learned he had a JAK2 mutation that was causing a coagulation problem, manifested as spots on the retina and pain in the thighs. The treatment, luckily, was simple – baby aspirin.
But those sorts of things – while clearly important to the individuals – are to some extent beside the point. It's the science that counts and on the face of it, scientific applications might seem some distance away. If you will eventually have 100,000 or more datasets, what can you do with a couple of thousand?
In fact, researchers are already beginning to use what's available.
For instance, Church says, the FDA and the National Institute of Standards and Technology are partnering to develop what they're calling "genomeinabottle" – a standard DNA sample that can be used to develop new diagnostic or other genetics-based tools. "They looked around the world for suitable DNA samples and came to the conclusion that none of them had been properly consented, other than the Personal Genome Project," Church says.
One in-house project is a contest to see which researchers can develop the best software to predict traits from the genome. The scientists will be able to see the DNA sequence of 50 participants – Clifford Andrew is one of them – but the rest of the data are being held back temporarily. It's "kind of like a bake-off," Church says.
Canada's Scherer notes that even conventional genetics can benefit. "For every single study we do," he says, "we are severely lacking control data." The thousands of personal genomes could help fill that gap.
"Our hope is that this will give us insight into precision medicine," Church says, "not only identifying which combination of genes and environment cause a particular trait, but how we combine lots of information from lots of such studies to treat individuals."
In a way, the Personal Genome Project may help medicine come full circle. Early doctors, who knew very little, focused on the individual patient because they could do nothing else. With the rise of modern medicine and such tools as clinical trials – doctors began to understand the similarities among patients with the same disease. Now, Church says, his project may allow the "physician of the future" to regain that focus on a single patient, using all the possible facts, not just genes or even broad-brush disease states.
"In the end, no two of us are really the same," he says. "No matter how finely you dissect it, you're always unique."
Michael Smith
North American Correspondent
North American Correspondent for MedPage Today, is a three-time winner of the Science and Society Journalism Award of the Canadian Science Writers' Association. After working for newspapers in several parts of Canada, he was the science writer for the Toronto Star before becoming a freelancer in 1994. His byline has appeared in New Scientist, Science, the Globe and Mail, United Press International, Toronto Life, Canadian Business, the Toronto Star, Marketing Computers, and many others. He is based in Toronto, and when not transforming dense science into compelling prose he can usually be found sailing.
.png)
No hay comentarios:
Publicar un comentario