$1 million for your genome map: Welcome to the future.

[Guest guest]

Welcome to the Future by David Ewing Duncan November 2007 Issue With a little help from Google, secretive Silicon Valley startup 23andMe is betting you’ll want to Web-surf your own DNA. But is the science ready? Gadgets The Business of Genomics View Slideshow More than 50 years after the discovery of the double helix, its profit potential remains unrealized. See All Video & Multimedia RELATED CONTENT From Portfolio

The Danger of Personal Gene Tests National News Predicting Your Future View Interactive Feature What you might learn about yourself from new genetic data. See All Video & Multimedia This just in: Illumina C.E.O. Jay Flatley has the “sprinting” gene, a cluster of DNA that is associated with professional athletes and Olympic track stars, those human beings whose muscles twitch fast instead of slow. Flatley, a stocky 55-year-old industrial engineer, chuckles at this discovery.On the other hand, he lacks a fairly common gene variant linked to Type 2 diabetes. That’s good news, he says, zooming in on a virtual human body that’s now floating on his computer screen. He clicks on the figure’s leg, and up pops a page describing the gene for restless legs syndrome, a malady that causes some people to flail around at night, resulting in sleeplessness. The webpage includes graphics, descriptions, and links to

studies and references, including PubMed, an online health science library. “This site will tell me if I have a gene for obesity,” he says, “or whether my earwax is wet or dry. I’m not kidding.”My fast but fantastic voyage through the inner Flatley comes during a visit to his company’s San Diego headquarters, where he has allowed me to peek at an early version of the site being developed by a Silicon Valley startup, 23andMe, which is named after the number of paired chromosomes in every human. In stealth mode until a launch planned for later this year, 23andMe is using Illumina’s powerful gene-reading chips to build a company that is among several hoping to cash in on personalized genomics—the tailoring of an individual’s health care and lifestyle based on DNA and biomarkers.When 23andMe goes live, customers will be asked to spit into short plastic tubes or to swab cells from inside their cheeks and mail in the sample. 23andMe

will then analyze the data in relation to reams of information about ailments, treatments, diet, and ancestry and compare the results with those of thousands of others who have been tested for genes associated with diseases and other traits.The cybersynthesis will be channeled into a customized DNA diary on the website. The company is considering a feature that would allow people to then link their personalized pages to those of others who share their DNA—fellow sprinters, say, or people at risk for Alzheimer’s—just as you can now link to college chums on Facebook. 23andMe has not discussed pricing, but competitors are talking about charging upwards of $2,000 a person.As wonderful as this sounds, 23andMe is plunging into very controversial territory. Scientists and ethicists warn that the understanding of most genes remains a work in progress—and, as a result, users could make life decisions based on incomplete or

erroneous science. “Just because we have identified a gene doesn’t mean its function or its impact has been thoroughly understood or that having a gene has any real predictive value,” says Francis Collins, who led the international consortium that sequenced the human genome and now directs the National Human Genome Research Institute.This admonition, however, has not deterred 23andMe and other online companies from entering what they believe is a potentially gigantic market. “We’ve only scratched the surface of understanding the human genome,” admits a source familiar with 23andMe. “But the company will be very responsible with the information it will make available to consumers. 23andMe is working closely with the research community.” The startup’s buzz comes from its close association with Google. The co-founder of 23andMe is a 34-year-old former biotech investor named Anne Wojcicki, who is married to Google co-founder Sergey Brin. Wojcicki met Brin when her sister rented her Menlo Park, California, garage to Brin and Google co-founder Larry Page as their first office. Wojcicki’s father, Stanley, is a physics professor at Stanford University,

and she has a degree in biology from Yale University. Before 23andMe, she worked as an investor at a San Francisco hedge fund, Passport Capital. In 2006, she joined Linda Avey, a 47-year-old business-development expert with extensive experience with major gene-technology companies, to form 23andMe. In May, Google invested $3.9 million in the tiny startup, part of $10 million raised in a recent round of private funding, according to VentureBeat, an online news site covering Silicon Valley. This influx of capital reportedly allowed the company to pay back a $2.6 million personal loan from Brin. Other investors include Genentech, the Menlo Park venture capital firm Mohr Davidow Ventures, and New Enterprise Associates.Brin has said that he

wants to unleash Google’s search technology to make sense of genomics, which so far has yielded rafts of genes and studies that have yet to be well organized. Google has offered few details, and its representatives declined to comment for this article, but biotechies eager for a fusion of stodgy life science with hip I.T.—dare I call it Web 3.0?—are closely watching 23andMe as a possible vehicle for Brin’s vision. Other contenders include Navigenics, a startup planning to launch early next year; DNA Direct, which already offers individual gene tests online; and DeCode Genetics, which conducts original research and is developing new drugs. In 1998, famed geneticist Craig Venter stood in front of a crowd and held up a credit card with his picture on it. In five years, he said, everyone would have a card like this. A magnetic strip on each card would contain all a person’s DNA information.In those heady days, his prediction seemed believable. The Human Genome Project, a government-funded program to map and analyze all the genes of human beings, was nearing completion—with a privately funded effort led by Venter in hot pursuit—and the promise of personalized medicine was fueling a boom in genomics companies. Unlike 23andMe’s effort to sell tests and information about one’s genes online, the companies of a decade ago were trying to develop personalized drugs. Big Pharma invested heavily, startups raised billions of dollars,

and stock prices soared.In 2003, five years after Venter’s forecast, genomics had scored few successes, personalized medicine still had not yet arrived, and the genomics bubble had burst. What happened? Companies discovered that the science of linking one’s own genetics to new drug compounds was far more complex and expensive than originally anticipated. In 2002, Venter was ousted from his company, Celera (he says he quit), which reorganized itself into a more traditional diagnostics business.While commercial genetics struggled, the science of gene identification exploded. Researchers using

ever more powerful sequencers and computers feverishly combed through the 3 billion nucleotides of the human genome—all those A’s, C’s, G’s, and T’s. Geneticists have now strongly linked nearly 2,000 genes to highly heritable rare diseases by combining genetics and computers in a science called bioinformatics. About 50 have been strongly linked to more common diseases, says Francis Collins, but the science is moving so fast he expects that number to increase to 500 by 2009. Every few days, new genes are discovered and linked to diseases, everything from male pattern baldness to schizophrenia. “A Moore’s law for bioinformatics is under way,” says Randy Scott, C.E.O. of a diagnostics company named Genomic Health. Moore’s law is the axiom that computing power doubles every 18 months. “Like mainframes to personal computers, this is an unstoppable force.” A few meaningful genetic tests are now commercially available. For instance, thousands of

women a year are tested for variants of two genes associated with breast cancer, BRCA1 and BRCA2, at a cost of between $300 and $3,000 per test. Breast-cancer patients are tested for a variation of the HER2 gene to see if they might benefit from Genentech’s anticancer drug Herceptin. The Oncotype DX test from Genomic Health analyzes 21 genes and uses a mathematical equation to assess the likelihood of cancer recurrence and to assess treatment options. Roche Diagnostics manufactures DNA tests that identify patients who might suffer from the side effects of antidepressants; sales of such tests are expected to reach $1 billion in a few years. This summer, the Food and Drug Administration recommended that patients take a genetic test before being prescribed warfarin, a blood thinner that can cause dangerous clots or excessive bleeding. Thousands of pregnant women a year also have their fetuses tested for rare genetic disorders such as Tay-Sachs

disease. These tests are typically conducted one at a time, under the supervision of a physician, and the patients are usually already afflicted or have a family history of or ethnic proclivity to a disease. In contrast, a new wave of companies has begun providing testing on genes associated with diseases directly to consumers. One pioneer is DNA Direct, founded in 2003 by Ryan Phelan, a medical-information entrepreneur who developed an early internet site that became an important part of WebMD. DNA Direct offers 12 online tests aimed at clients who have some indication they might contract a disease. Costs range from $199 for simpler tests to $3,456 for a more complicated panel; the fees cover phone access

to genetic counselors.23andMe is planning to test for perhaps dozens of genealogical factors, genetic markers for disease, and other traits that make you you. The company has not yet revealed if it will offer genetic counseling.Some analysts predict that the genetic-testing market 23andMe is entering could be worth a staggering $12.5 billion by 2009. Naturally, this has attracted the interest of Web entrepreneurs. They see an industry that is largely unregulated (so far) and costs only a few million dollars to enter—the price of a few brilliant programmers, a website, and marketing—and are betting that people will pay to test their own DNA directly. One indication of the potential market is that online medical-information companies are starting to make real profits. WebMD, for instance, attracts 40 million users a month and expects to net more than $30 million this year, mostly from ad sales. “I’m convinced there is an early-adopter market here,” says Sue

Siegel, former president of Affymetrix and now a venture capitalist at Mohr Davidow. “Millions of people are used to getting health-care information online.”The big question is whether healthy customers will pay $2,000 for gene tests. “Someone needs to be the Quicken of health care,” argues one biotech insider. “This is not for the faint of heart. It will be competitive. You also have the regulatory factor, which is an unknown, and much of this still needs to be validated in the clinic.”Indeed, many longtime biotech investors who are much closer to the science remain unconvinced. “I tend to be more on the skeptical side for a genomics consumer play because, by and large, the information is not useful to the individual,” says Doug

Fambrough of Oxford Bioscience Partners, a biotech venture capital firm headquartered in Boston. “First of all, the information is hard to understand for anyone who is not a trained molecular biologist. It has to be boiled down. Also, the science has not yet arrived. We can tell a single trait here and there, but what people really want to know is how this affects their life. We can’t yet make accurate predictions about this.” Until the science improves, Fambrough thinks that the service will merely capitalize on the vanity of the more affluent. The big market, he says, is still with hospitals and physicians. “To make money, you have to tap into the medical market, and you have to get insurers to pay for it. If you have a medical test for

everyone, this is worth billions.”Many physicians and ethicists are also uncomfortable with the idea of websites delivering genetic results directly to consumers, especially for diseases that have no cure. “I don’t think that patients should be told they have a gene for Alzheimer’s online, or even over the phone,” says Josh Adler, an internist and the medical director of ambulatory care at the University of California at San Francisco. Currently, Alzheimer’s is incurable and largely untreatable. “Doctors should be telling patients this sort of information,” Adler says.“This should not be taken lightly,” agrees Jonathan Rothberg, co-founder of the gene-sequencing company Raindance Technologies and a geneticist whose previous company, 454 Life Sciences, sequenced the genome of famed geneticist James Watson. Rothberg says he’s a fan of online gene services but cautions that they need to be used responsibly. “People shouldn’t be surfing through serious medical

data. Some of it is fun—genealogy is fun. But I had a cousin who killed herself when she found out she had Huntington’s disease.” Critics also see little value in testing healthy people for a wide range of possible diseases. “We don’t take an M.R.I. for everything, and I don’t order every test for every person,” says Harvard geneticist and physician David Altshuler, a key figure in the Human Genome Project. “Those who do are scamming people. It’s the idea that just knowing something is useful—well, maybe, maybe not.”Part of medicine’s resistance comes from ignorance, counters Lee Hood, a biotech pioneer, M.D., and the president of the Institute for Systems Biology in Seattle. Most physicians have little training in genetic testing, he says, so they have no idea what to tell their patients who hear about tests or want them done. “Health-care professionals need to be pushed to understand this information,” National Human

Genome Research Institute director Francis Collins agrees. “Most are not ready for this.”Ten miles up Highway 101 from 23andMe’s Mountain View office is another online gene-testing startup called Navigenics, which is also in beta mode and plans to launch in early 2008. Focusing on health genetics, it plans to offer consumers a panel of 15 to 20 gene tests, along with detailed information on medical conditions and genetics. “The tests they will run are all clinically validated,” says Brook Byers of Kleiner Perkins Caufield & Byers, the big Silicon Valley venture capital firm, which is investing in the company.Navigenics C.E.O. Mari Baker, previously the head of Johnson & Johnson’s BabyCenter

website and the product manager for Quicken, explains that initially the company will offer tests only for treatable diseases. “Customers will always have the option to exclude certain conditions from their report,” Baker says. One controversial test is for a gene linked to Alzheimer’s disease. “We believe that there are enough promising therapies in development for Alzheimer’s that it makes sense to include it, since recognizing it early is important.”Navigenics’ market is intended to be people who are healthy and affluent. Customers will be charged between $2,000 and $3,000 to have nearly a million genetic markers tested on a gene chip manufactured by Affymetrix. But Navigenics’ site won’t release all of the data collected by the chip, only the designated panel of gene tests. The company plans to offer information and telephone support from genetic counselors, and a subscription to its service will last a year. “Your DNA will be on file, and we’ll test it against

new findings,” says Amy DuRoss, Navigenics’ head of policy and business affairs. Navigenics was co-founded by David Agus, a famed oncologist and geneticist at Cedars-Sinai Medical Center in Los Angeles, and the company has set up a task force to deal with bioethics, patient rights, and data security.Navigenics is positioned to go after the millions of Americans who are spending money outside of the traditional health-care system to stay healthy, buying vitamins and dietary supplements, gym memberships, and diet and self-help books. “These are people who want to take control of their health,” Baker says. “They will own their own DNA and their results. It’s up to them who they want to share it with; it’s their call. The results will be sent to the patient, not to a doctor.” But Navigenics is also talking to major health-care centers around the country—the Mayo Clinic, the Scripps Research Institute, and the Stanford University Medical Center.Upstart sites

like 23andMe and Navigenics are about to be joined by older genetics companies—including survivors of the late-1990s genomics boom—that run large research labs and develop drugs. One of these is Iceland’s DeCode Genetics. Its DNA studies on the isolated island’s population have led to many of the clinically validated genes—those proved to be strongly associated with a disease—appearing recently in the news. DeCode C.E.O. Kári Stefánsson says that the company plans to launch a direct-to-consumer site next spring. “We have the original data and they don’t,” says Stefánsson about his I.T. competitors.But once people are done surfing through information on their earwax and fast-twitch muscles, will these companies be able to handle the tough questions that consumers really want answered: Will I get diabetes? Will I contract leukemia? Will I have a heart attack? When? The short answer is no. Most people who carry a gene

associated with a disease never contract the malady, while others among the afflicted do not carry the gene at all. Even if a test is valid, customers will need to get used to dealing with probabilities of contracting diseases, rather than a simple “you will” or “you won’t.” For instance, the TCF7L2 gene, associated with Type 2 diabetes, has a variant carried by more than a third of the population that increases their chances of contracting the disease by 45 percent. Harboring the TCF7L2 variant is only one factor at work, though. Activating the gene usually requires eating lots of Big Macs and guzzling sugary sodas, but consuming lettuce, lean meat, and fruit juice can help prevent the disease. Finally, there are possibly hundreds of other genes that have something to do with diabetes, many of them unknown, with different combinations triggering the illness in different people. “We need to remember that genomics is only part of the equation,” Lee Hood

says. “Environment and other factors play huge roles.”One more bit of complication is that, strictly speaking, most genetic tests are not really sequencing genes—those long strings of nucleotides that instruct the body to grow a tooth or create white blood cells to attack an incoming virus. They are after something much smaller: tiny variations in the genome—99 percent of which is identical in all humans—that account for differences like height, eye color, and susceptibility to disease. These variations, called single nucleotide polymorphisms, or SNPs (pronounced snips), are read by gene-chip technology like Illumina’s. The problem is that these chips don’t cover all the SNPs that make us different. In fact, they can detect only 10 percent of all 10 million common SNPs. Researchers direct the search to the “high-value” regions where certain genes cluster, but this system is designed primarily for scientists looking for genes, not for individuals hoping to

pinpoint their own specific disease risk or traits.To chisel out every last detail of your inner you, you would need to sequence your entire genome. Recently, this was done for James Watson—the co-discoverer of the double-helix shape of DNA—and geneticist Craig Venter. “Gene chips don’t compare to whole genome sequences,” Venter says. “What we need is more genome sequences, thousands of them, to make sense of genetics for individuals.”No one disagrees with this. But the cost of sequencing an entire genome is nearly $1 million. That’s a relative bargain compared with the $3 billion price tag for the Human Genome Project, but it’s prohibitive for a large-population study. “When sequencing is as cheap as buying a Chevrolet, then we’ll be able to make this science really work,” Watson says.I ask Hood, who has met with Google and has long been a maverick bridging the worlds of biology and I.T., do Web entrepreneurs truly understand the limitations and

pitfalls of this science? “They absolutely do not,” Hood says. “The heart of predictive medicine is in getting clinical validation and working out the fundamental biological systems—how genes and proteins and other elements interact. I don’t think that most of the Web 2.0 crowd entirely gets this.”So far, the world of consumer genetics has been a Wild West. A handful of genetic tests are regulated by the federal government (such as the one doctors order before prescribing Herceptin). Otherwise, companies have operated without fences. Pressure is building, however, for the feds to establish rules and standards for direct-to-consumer gene tests. Last year, the Federal Trade Commission issued a Facts for Consumers report that bluntly warns consumers to be wary of all companies that offer do-it-yourself genetic tests and promise to tell you that your genes can reveal a

definitive risk factor for developing a particular condition. The report says the F.D.A. and Centers for Disease Control and Prevention “aren’t aware of any valid studies that prove these tests give accurate results.”Last summer, Congress investigated one area of direct-to-consumer DNA testing known as nutrigenetics, which creates personal genetic profiles meant to let people customize their diet and purportedly improve their lifestyle. The resulting report found that four leading nutrigenetics companies were selling results that were not verified by science, plus lifestyle advice that was generic and obvious to anyone trying to eat right and be healthy. F.D.A. officials agree that this industry should be more tightly controlled, and this summer, the agency issued guidelines for companies offering genetic tests that claim to be able to predict medical outcomes. (23andMe and Navigenics are careful to

say they are not offering predictions or medical advice.)As I’m leaving Illumina, I ask Jay Flatley about another major issue: whether consumers will be able to keep their DNA information private from employers, potential spouses, insurers, and anyone else. He says this information can’t stay under wraps for long. “People will find out about it,” he says. At the National Human Genome Research Institute, Francis Collins agrees, adding, “You need to have laws to protect this information.” He has worked for years to promote a bill in Congress to prevent genetic discrimination and protect personal DNA information. Last spring, the House of Representatives passed the bill, and President Bush says he will sign it if the Senate also approves it. It has been held up in the Senate, partly because of concerns from insurance companies, but Collins thinks it will eventually pass.Doug Fambrough of Oxford Bioscience Partners wonders whether trying to sell genetic tests

to consumers could backfire. “People may try buying these products, but if they find it’s not useful or are afraid it won’t be handled properly, this will set back the prospect of bringing this new business to consumers,” he says. On the other hand, the public might embrace this information in just the way that 23andMe seems to be envisioning—as cool, fun, and informative. It’s possible that, like Illumina’s C.E.O., people will want to know more and will link up with others with such shared genetic proclivities as tallness or colon cancer. If this really takes off, maybe we will see a new type of genetic paparazzi who try to snag DNA from stars—taken perhaps from a lip imprint on a wineglass. "Lindsay Lohan’s Genes Revealed!" could be on a future cover of People magazine. And what about politicians? What’s to stop an enterprising political operative from snooping on the genetics of presidential candidates? We might find out that Hillary Clinton and

Rudy Giuliani have the novelty-seeking gene and Barack Obama does not—or that John Edwards carries a gene for dry earwax. What if the candidates have the gene for Alzheimer’s disease or prostate cancer—particularly given that this does not mean they will ever contract these maladies? “We don’t want people misusing this information,” says Gregory Stock of the University of California at Los Angeles Program on Medicine, Technology, and Society. “We don’t want to be electing people based on genetic traits that are poorly understood and don’t mean much.” Or we might start judging C.E.O.’s on whether or not they have a gene for dementia or schizophrenia, or potential mates or employees based on their genetic profiles. Does anyone remember the 1997 Andrew Niccol film Gattaca, in which the people of a future society don’t make love with or hire anyone without checking their DNA?Lest we forget, the genetic results we may sign up to collect do not

belong just to us—they also contribute to our children’s DNA. Five years ago, I had several of my own genes tested for an article and remember that when I came up positive for a gene associated with heart disease, I found myself placing my hand on the chest of my son, then seven years old, as he slept, to feel his heart beating. The thump thump was strong and healthy.We all will die of something, and in many cases our genes will play a role. Yet I still wonder what I have given to my son and how he will respond to information that until now has been secret to all of us. By the time he is my age, will people really “23andMe” information about their genes in the way we google the address of a restaurant? Or will we discover that there is some knowledge about ourselves that’s better left unknown? Think Simply.

Think Wisely. Curb Semantics. Speak the Truth.

Now you can chat without downloading messenger. Click here to know how.