As the director of the Stanford Center for Genomics and Personalized Medicine, Michael Snyder was already a believer in improving people's health by analyzing their genomes. On April 11, 2011, he became a member of the choir.
That's when he got the official word that he had type-2 diabetes. He might never have spotted the condition so early, except that he'd had his own genome sequenced, and he and his lab members had discovered that he had a genetic predisposition.
"We didn't expect it," Snyder said in an interview in his campus office. "I happened to go in and get the glucose test because my genome said I was at high risk."
It turned out that Snyder's blood sugar had been climbing even as he and his team analyzed his genome. Glucose tests yielded increasingly higher results, and his diabetes was diagnosed. In essence, the researchers had been watching the inception of a disease.
Since Snyder had a physical exam only every two or three years and didn't know of any diabetes history in his family, the condition might have gone undetected for a long time. But now he was able to swiftly cut dessert from his diet and dramatically increase his exercise.
"I had one bite of wedding cake in 2011," he confessed. His blood-sugar levels improved.
For Snyder, the moral of the story is clear. While complete genome-sequencing is not yet widely used, its promise for people's health -- and that of their families -- can't be denied, he believes. In his case, his diagnosis inspired his siblings to also get glucose tests, and in some cases, start exercising to improve their own blood-sugar levels.
"I'm a believer in the future," he said.
Snyder, who also chairs Stanford's genetics department, has headed the Center for Genomics and Personalized Medicine since its birth in 2010. It's an interdisciplinary effort with an ambitious mission: to continue analyzing the human genome (a person's complete set of DNA) and its molecular makeup and then translate findings into individualized medicine. Researchers believe DNA holds valuable information on predicting, diagnosing and treating conditions as diverse as cancer, schizophrenia and asthma and anticipating which medications may work better for a particular patient.
Through genomics, researchers are also developing vaccines made from DNA and RNA. In addition, they're seeking to lower health care costs through preventive care: as many describe it, moving medicine from "diagnose and treat" to "predict and prevent."
"I've never seen a more exciting time in medicine than now," said center co-director Stephen Galli, who also chairs Stanford's pathology department.
It's been a banner 23 years for the young field of genomics. The U.S. Human Genome Project, an international enterprise led by the U.S. Department of Energy and the National Institutes of Health, officially kicked off in 1990. The aim was to "discover the complete set of human genes and make them accessible for further biological study, and determine the complete sequence of DNA bases in the human genome," according to the project's website.
The project was expected to last 15 years, but researchers and technology were ahead of their game. By 2000, leaders had announced that a "working draft" DNA sequence of the human genome had been completed. In 2003, the project was declared finished, all 3 billion DNA letters in the human genome successfully sequenced. The science of genomics was off and running.
Researchers at Stanford and numerous other institutions saw the promise early on. With whole genomes being sequenced, they could look for genetic variations: differences between individuals' genomes that could signal genetic diseases, risks for diseases and potential ways that people could react to medications, pathogens and other environmental influences. Many variants have no effect, but the ones that do can have dramatic ramifications.
Since the first human genome has been sequenced, the process has continued to improve, and the cost and time needed for sequencing has dropped significantly. The Human Genome Project cost about $2.7 billion in fiscal-year 1991 dollars. Today, a complete sequencing of a genome can be done for a few thousand dollars, in a day.
People also use the term "the $1,000 genome" to refer to a future when complete sequencing -- and a personalized plan of medicine coming out of it -- will be accessible to the public at large.
On a smaller scale, there are now companies that provide commercial genotyping to the public, providing a partial DNA analysis from a saliva sample.
"It has some value," Snyder said of the practice. One of the firms, the Mountain View-based 23andMe, promises to give clients insight into what diseases they may be at risk for. It also offers a look at a client's ancestry -- in some cases, helping seek out "new-found relatives."
While genomics was developing as a science, Snyder was working at Yale University. In the early days, researchers studied genes one at a time; he took part in an early project that analyzed thousands of genes at once. Full-genome sequencing looked like an exciting future.
"The cost of sequencing was dropping in 2005 and 2006. We saw that you'd be able to sequence for an ordinary person," he said. By the start of 2009, the year he came to Stanford, two genomes were sequenced, with more to come. The center he now heads was established the following year.
Stanford's genomics program, which includes an off-campus sequencing facility in Palo Alto with machines from the Hayward-based Illumina Inc., is one of a few in the country to receive funding from the National Human Genome Research Institute at the National Institutes of Health. Other institutions that have gotten grants from the institute include the California Institute of Technology, Harvard Medical School and Yale University.
Snyder describes much of the genomics work done at Stanford today as clinical research rather than clinical. Lab researchers studying the genome are exploring a variety of questions. Which genes are linked to heart disease? What can genes reveal about predispositions to cancer or whether a tumor will grow? How is genetic material related to autoimmune responses, autism, asthma? Stanford has also launched a repository for genetic samples to create a database that could aid in future study.
In some cases, sequencing is also used in individual patients' care at the university's hospital system. For example, a patient with an unusual syndrome that baffles doctors might have his genome sequenced in hopes of finding out what the condition is and what existing drugs might work for it.
"For a long time, we've known how to use individual genes," Galli said. For instance, if a person has cystic fibrosis, doctors would know to test his or her child for it, or to tell a patient about potential risks to his or her unborn child. Today's use of genes is broader, Galli said: "We're able to tell you that you have a specific mutation and that this particular medicine might work well for you."
But with full-genome sequencing not yet widely available, genomics is still more a science of the lab than of the physician's office. The average doctor has not been trained on how to interpret sequencing results and to communicate with patients about them, nor are all doctors convinced that the technology has widespread use in individualized medicine, Snyder said.
"We're still cave people in terms of our prowess at this."
That's why Stanford's work in genomics is an interdisciplinary effort. It's not just about the researchers delving into questions of science in the lab.
As genomics moves toward being translated into widespread personalized medicine (or precision medicine, as some call it), many other questions arise: how to interpret the data that emerge from labs and studies; how to train doctors and other health care workers to deal with the data; how to help patients and doctors grapple with the ethical questions that arise; how to protect patients' privacy. More broadly still: Who owns genetic information, and how should society use it?
Stanford faculty members in various disciplines are involved with genomics, extending beyond the genetics department. There are people from pediatrics, pathology, developmental biology, bioengineering, computer science, biomedical ethics, psychiatry.
Atul Butte perhaps best personifies the interdisciplinary approach. He's an associate professor of pediatrics and genetics and the chief of the division of systems medicine. Along with his M.D. and doctorate, he has a degree in computer science, where he started out.
"I still love to code when I have time to do it," he said during an interview in his campus office. There, his framed diplomas and stacks of science magazines are joined by a sign from a TED conference, where he gave a talk called, "What if you outsource three double-blind mice?"
While Butte's colleagues are in the lab or the clinic generating data, he's helping them make sense of the findings and turn the data into clinical recommendations. He and his team perform statistical analyses on genomic data, look for patterns as they map, compare and analyze the data that other researchers from all around the world have put out there, often on the Internet.
"We all benefit from sharing what we can contribute," he said. "It's one thing to learn from a big data set. It's another thing to learn from two of them. ... What do 10 researchers see? What do 100 researchers see?"
Once scholars like Butte have analyzed the data, concrete ideas may emerge that can help patients. For instance, a study of cancer patients might yield a clue to which genetic variant is connected to an adverse reaction to a particular drug.
Butte has also focused on finding new ways to use old drugs. Often, a medication already approved by the FDA for one condition may be beneficial for another. In one case, Butte analyzed publicly available data from the Internet on lung-cancer patients and saw that the tricyclic antidepressant desipramine had a surprising positive effect on the cancer. He notes that "drug-re-purposing" has a long history (Viagra, for instance, was originally an angina drug) but not to this extent.
In true Stanford tradition, Butte and others started a company. He's a founder and scientific adviser at NuMedii, a Palo Alto firm that translates the "Big Data technology" developed in Butte's lab into finding new uses for existing drugs.
Butte also makes use of technology developed by other firms. He holds up a chip made by Affymetrix Inc. in Santa Clara. Called the GeneChip, the measurement tool scans DNA samples to seek out genetic variations.
"People use these to figure out patterns," he said. "This chip is amazing."
Snyder and Galli agree that the data analysis done by Butte and his team is a key component to their work at the center. Without the results from studies being properly interpreted, doctors would all be as lost as the pharmacist in a cartoon popular with people working in genomics. There are many versions of the comic, but basically, as Galli puts it: "Someone walks into a pharmacy, hands a pharmacist a piece of paper and says: 'This is my genome. Tell me what I need.'"
With all the interpretation, disciplines and training needed in genomics, there's another popular joke in the field, too: "We're heading toward a $1,000 genome and a million-dollar interpretation."
When the topic of genomics comes up among people outside the field, the discussion can be very different. It often boils down to one central question: Would you have your genome sequenced? Which really means: Would you want to know what you're at risk for?
Other questions are inextricably woven in: What would you do if you found out you were at risk for a serious condition? Would you want your family to be tested? Issues of psychology, ethics, responsibility are forever entwined with genomics.
At Stanford, the Center for Biomedical Ethics plays a major role in the university's work in genomics, with center Director David Magnus actively involved. He and his colleagues are used to being brought in to help researchers, physicians and others address ethical issues in many disciplines. As a young science, genomics raises even more issues, some of them unprecedented.
One of the most common questions that many institutions are grappling with is one of the most fundamental, Magnus said: What sequencing results should be given to patients?
"It's easy to say we can give patients whatever they want, but it might not mean anything" to someone not trained in genetics, he said. This goes back to the need for more training of physicians in helping patients deal with sequencing results.
In addition, he noted, genomics' young age means there are many things that even high-level scientists haven't figured out yet.
"Whenever you do a full sequencing of anyone, you'll find variations that will be of unknown significance," he said.
With a genetic variation, the stakes can end up being very high: matters of health and sickness, life and death. As sequencing technology continues to become more widely available, it means more high-stakes choices that patients have to make about their own health and the health of their relatives. Overall, genomics, Galli said, "is going to increasingly require patients to play a big role in taking control of their own health care."
That starts with the first decision: whether a person wants to have his own genome sequenced. While doctors grapple with the question of how much information to give to patients, they also have to realize that some patients won't want it.
Galli uses the example of the risk for developing early-onset Alzheimer's disease. Some people are planners and will want to know. They'll want to make sure that their families will be provided for, that they'll get to do what they want early in life.
"Even they may be sorry they know it, but they'll ask for it," he said.
Other people will want to know only if something can be done about it -- if their future condition will be treatable. Still others won't want to know at all.
"The patient needs to have a choice," Galli said. "It involves how you think about your own future, your children's future, whether you're going to have children. What's right for one person may not be right for another."
Another question of ethics related to genomics is one of privacy. Can people's health data really be kept anonymous when study results are shared on the Internet? Can people be identified by their genetic information, even if their names are kept out of it? What could be done with those data?
Magnus said that privacy is one of the major ethical issues that keeps coming up around genomics. But he points out that privacy has always been a concern in health care. In one sense, protecting privacy is the same as it's always been: having good computer security, building firewalls, keeping names off study data. Many large databases used for genomics research are restricted to vetted researchers and not open to the public at large, he said.
With genetic information, though, security may need to go farther, he said. In some cases people could theoretically look at public records, such as voter rolls, and combine them with study databases to try to identify individuals.
"That's a new challenge, and people are working to find ways of dealing with the bioinformatics challenge," Magnus said. "People really are trying to think through policies and try to do the best we can. It's always hard to anticipate every curve."
Patients can also play a role in protecting their own privacy. Magnus has heard stories of people going on social media and sharing details they've learned about their own genome, whether it's serious information from a full sequencing or something that seems more innocuous, obtained from a partial genotyping from a commercial company.
"People may think it's innocuous to share things on Facebook, like the inability to taste cilantro, and yet the details might be sufficient to identify people in a larger database and now know whether they're at risk for early Alzheimer's, and an employer could decide not to hire them," Magnus said.
Such horror stories are largely hypothetical at this point, Magnus noted. "We're trying to focus on the concrete issues that we can address now."
Back in his office, Snyder, ever the advocate, painted a brighter picture of the future. He imagines a day when the average person will go to the doctor and not get just a few limited tests. Instead, anyone will be able to get a full genome sequencing, meaning that a patient could be tested for "hundreds of thousands of things" at each doctor's visit. Patients will be able to anticipate and respond to risk factors, adverse medication responses. They'll be able to make lifestyle changes and improve their health, and work with their doctors to design therapies for anticipated problems.
"Genomics is a no-brainer," he said. "It needs to become standard care."