(FORTUNE Magazine) – With genomics stocks deep in the tank, it seems fair to put a pointed question to Eric Lander, gene science's go-to guy for the big picture: Hasn't the value of his field been way overstated? Lander's answer includes a little gem of showmanship: He grabs a Diet Coke from his desk at the Whitehead Institute/MIT Center for Genome Research and begins reading the label out loud. It warns that one of the ingredients, phenylalanine, is toxic to people with a disease called PKU. Ingesting the chemical can cause brain damage in kids with the disorder--which they inherit from their parents. "To understand this disease," Lander observes, "we needed genetics."

Touche--it's hard to knock a field whose benefits appear on soda cans. This is standard Lander, "a ham with a lot to say," as a former colleague describes him. A math prodigy, he picked up biology in his spare time, then went on to become one of the biggest names in genetics. When the Clinton White House held a gabfest on genetics in 1999, Lander was summoned as the discussion leader--to no one's surprise. His passion is genetics' killer question: What DNA variants underlie disease, predisposing some of us to heart attacks before age 50, making it almost impossible for many of us to lose weight, or causing a few of us to get pancreatic cancer?

No one has attacked the question more forcefully than Lander. He pioneered concepts for mapping the human genome and built one of the world's top labs for studying DNA. In the race to sequence the three billion nucleotides, or chemical letters, that make up human DNA, Lander's lab helped lead the Human Genome Project's $300 million international effort against its commercial archrival, Celera Genomics. (The race, which ended two years ago, was deemed a tie.)

Now that the human genome is an open book, Lander's favorite question has become Topic A. Getting answers and turning them into new therapies will take years. But just when many investors seem to be giving up on the gene revolution, its pace of discovery is quickening, presaging advances that in the long run will make today's hype about genomics seem quaintly conservative. "The Genome Project was the groundwork to get us to the point where we can do serious work on human disease," says Lander. "Now we're ready. And boy," he adds in a conspiratorial undertone, "is it fun."

Next on the agenda is the biggest push since the Human Genome Project: a $100 million, two-year quest to speed research on which genes connect with which human traits. Variations in the DNA letters of the genome underlie individual differences from blue eyes to a tendency toward diabetes. The idea is to map out a master set of variations, or so-called haplotypes, that sum up most of what makes us genetically different from one another. This is possible because human DNA varies surprisingly little compared with the genomes of, say, worms (see box). Researchers would use the resulting "haplotype map," or hapmap, to help pinpoint genes linked to diseases--"the smoking guns" in the genome, as Lander puts it.

Genetic culprits have already been fingered in studies of many rare, inherited diseases. Such cases are typically due to misprints of a single DNA letter in a single gene--one such aberration is what causes PKU, the disorder that can make Diet Coke toxic. The most common killers, though--heart disease, cancer, diabetes--pose a far tougher challenge to geneticists. Such illnesses are thought to stem from multiple genes interacting with various nongenetic triggers, such as years of pigging out on junk food. The hapmap promises to help in the dauntingly complex quest to identify these insidious gangs of bad actors.

In theory, that goal is within reach, because researchers can now scan the entire genome to look for DNA variants of interest. Perlegen Sciences, a closely held company in Mountain View, Calif., recently announced that it can parse a person's genome in about ten days using so-called DNA chips--an astounding advance, given that it took an international army of scientists all of the 1990s to create the first draft of the human genome.

In practice, though, scanning an entire genome for disease-associated variants is prohibitively expensive--an estimated ten million of the 3.1 billion letters in human DNA are variable. Looking for risky variants amid all those candidates can require studies with many thousands of people.

That's where the hapmap comes in. By distilling humankind's genetic diversity into a relatively small set of widely shared DNA variants, it will greatly simplify the quest for smoking guns. Lander predicts that around mid-decade "we'll see a huge explosion" of gene discoveries thanks to the hapmap and other tools for mining the genome. As soon as risky genes are pinpointed, researchers can begin unraveling how they work and finding ways to neutralize their ill effects.

Even before novel drugs arrive, genetic tests may help prevent common diseases. For instance, a variant of the so-called Apo E gene has been found to increase the risk of Alzheimer's disease. Evidence is also growing that certain drugs, such as cholesterol medicines called statins, can slow down or block Alzheimer's. More data are needed for doctors to make use of these findings. But they raise the possibility of averting at least some cases of Alzheimer's that are spotted early with the aid of genetic tests.

Still, some experts question the value of the hapmap project, which will be funded by the U.S. National Human Genome Research Institute and allied agencies in Japan, China, Canada, and the U.K. Most attempts so far to mine the genome for common-disease genes have "led to a whole lot of nothing," says Columbia University geneticist Joseph Terwilliger. In his view, that's mainly because a mind-boggling multiplicity of genetic variants is probably involved in the most common diseases. Each variant may raise the risk of sickness too little to stand out in the "statistical noise" of clinical studies. (If statistics isn't your thing, imagine trying to pick out the sound of a pistol shot during a fireworks finale.) The hapmap won't solve the problem, he argues; it may only prompt gene hunters to waste time and money.

Another skeptic, Pennsylvania State University geneticist Kenneth Weiss, argues that hapmap proponents like Lander and Francis Collins, head of the federal genome institute, are trying to "geneticize" diseases that are due mainly to overeating, smoking, or other behavioral factors. "Because of mega-projects like the hapmap," Weiss says, "there's a big danger we're going to get distracted from what overwhelmingly causes disease."

Such gibes aren't new, but they have more bite after all the media hype about the Genome Project. There's a deeper reason for the criticism too: The new project hits the hot button of genetic determinism. If it does lead to the discovery of predisposing genes for common diseases, millions of people could wind up being classified as high-risk cases by health insurers and prospective employers. Further, the project may help find genes that influence traits such as IQ--knowledge that would likely be abused by divisive pundits, devaluing the idea that we can make of ourselves what we will.

Not surprisingly, Lander has a lot to say about the hapmap's value--we'll get to that. To distance himself from determinism, though, he needs only cite his own curvy, free-spirited career--his vita reads like an ode to the joy of flouting predetermined paths. "When I first showed his resume to my colleagues," says Howard Raiffa, an emeritus professor at Harvard Business School who lured Lander to teach there in 1981, "they thought it was almost a joke. They couldn't believe anyone could be that versatile."

The son of two lawyers, Lander, 45, first achieved national recognition at Stuyvesant High School in New York City, a mecca for gifted kids with a science bent. In his senior year, he was valedictorian, placed second in a national math test, earned a spot on the U.S. team in the math version of the Olympics and then, to top it all, won first place in the prestigious Westinghouse Science Talent Search for a paper on number theory.

At Princeton, he majored in math but blossomed into a polymath. "While the rest of us were groaning, 'How can we cope with this workload?' Eric took on one extracurricular thing after another," says his college roommate Cole Whiteman. "One day he came back to the dorm and told me, 'I think I'll try sportswriting for a while' " at the school newspaper. "I said 'Why?' And he said, 'I just think I should get good at it.' So he did." Lander's flair for writing even led to a summer internship at Business Week. After graduating as valedictorian at Princeton, he studied math as a Rhodes scholar at Oxford, earning a Ph.D. in less than three years.

A dazzling talker and natural leader, Lander shied away from the solitary life of the pure mathematician. "By a series of lucky accidents," he says, his resume landed on Raiffa's desk at Harvard Business School as he was finishing at Oxford. Raiffa, a mathematician by training, saw Lander as a kindred spirit and had him installed in 1981 as an assistant professor of economics--a subject Lander then learned on the fly while teaching. But he was turned off by what he calls the field's "overmathematicized model of homo economicus." Casting about for something else, he got interested in how the brain processes information, a subject recommended to him by his brother, Arthur, a neurobiologist.

"That led to a regress, where I had to learn about cell biology, then molecular biology, and then genetics," he says. "I got stuck on genetics because it was so interesting." Still teaching economics, Lander began hanging out at Harvard's biology department. One of his mentors there was Peter Cherbas, now at Indiana University, who recalls, "I'd give him this huge stack of papers, and he'd come back the next morning having read and understood it all. It was a strange experience having someone vacuum your brain like that. In a couple of weeks, I felt I'd taught him most of what I know about biology."

By 1985 Lander had joined the gene hunt, applying his math gift to its statistical challenges. That year he met David Botstein, a prominent geneticist then at MIT, who helped finagle a post for him at the MIT-affiliated Whitehead Institute for Biomedical Research in Cambridge. Says Lander: "He's from the Bronx, and I'm from Brooklyn, so when we were introduced one day in the hall at MIT, we immediately started arguing." The competitive brainstorming helped Lander win a MacArthur Foundation prize in 1987, establishing him as one of the young lions of genetics.

Two years later, he showed how sharp his claws had become in a landmark trial that turned on genetics. Lander was enlisted to consult with the defense team of a janitor accused of stabbing to death a woman and her two-year-old daughter in a Bronx apartment. The key evidence was a bloodstain found on the man's watch--DNA tests purportedly indicated that the blood came from the murdered woman. But in a pre-trial hearing, Lander ripped that claim to shreds, showing it was based on sloppy lab work. At the end of the hearing, two scientists appearing for the prosecution actually switched sides, joining with Lander to declare that the test was faulty. The judge threw the evidence out.

Ironically, the janitor ultimately confessed to the murders. Yet the case helped spur the National Academy of Sciences to convene a blue-ribbon committee on DNA fingerprinting. Lander served on the committee, playing a key role in hammering out its quality-control standards for forensic DNA tests.

In 1990 he formed the Cambridge genome center and launched himself into a Henry Ford shtick, spearheading technology to turn DNA research into a high-speed enterprise with biochemical assembly lines and economies of scale. Today, rows of factory-style robots at the center tirelessly squirt chemicals onto colonies of cells, cracking them open to extract DNA molecules. Nearby, 150 "DNA analyzers" coax the chain-like molecules through hair-thin tubes in a process that reads off their chemical letters--the center can parse more than 50 million DNA letters a day.

As the genome's letters fell into place, Lander's team and others began mapping out places in the sequence where one person's DNA typically differs from another's. That, in turn, enabled scientists to explore a burning issue: If you spell out the DNA of a particular gene in, say, a thousand people, do you find large numbers of different spellings or only a small number of widely shared ones? A closely related question is at the heart of the hapmap debate: Are there myriad variants of the genes that contribute to major diseases, as Columbia's Terwilliger suspects, or do such disease-predisposing genes have quite limited diversity, making the genome a friendly place for disease fighters?

"No one knows the answer to this question," says Lander. "But it's likely that common variants play at least some role" in high-prevalence diseases.

In the past two years, Lander has helped build the case for optimism--and for going ahead with the hapmap--by co-authoring studies linking common genetic variants to adult-onset diabetes and Crohn's disease, an inflammatory bowel disorder. The diabetes research, led by endocrinologist David Altshuler of Massachusetts General Hospital and Lander's center, showed that a whopping 85% of the patients studied carry a particular variant of a gene involved in the body's formation of fat cells. The variant apparently raises the risk of diabetes by only 25%, though--a subtle effect that was hard to detect. Indeed, four earlier studies examining the same gene's bearing on diabetes had found nothing. Altshuler's team succeeded by collecting data on a massive cohort of 3,000 patients.

Thus, the finding lent some support to Terwilliger's view--even if disease genes have limited diversity, their individual effects may be too small to register without very large studies. Still, to date about ten common gene variants have been convincingly linked to major diseases, says Altshuler. That's cause for hope, since the search for such genes is still at an early stage.

But what about the charge that geneticizing diseases will divert attention from lifestyle changes that could prevent them? Altshuler, who treats diabetic patients, says this objection overrates what can be achieved by urging people to change. "Go to any diabetes clinic and you'll find it hard to be sanguine about how much we can do by telling people to do things like keep their weight down," he says. "The drive to eat is extraordinarily strong." Current therapies aren't much help either, because they don't get at the roots of disease as gene-based medicine will.

Lander adds that "the media may be overhyping genes. But the solution to that isn't to stop looking for the genetic contributors to disease. The solution is to be very honest and admit that both genes and nongenetic factors work together.

"We stand a good chance of beginning to get at the genes over the next decade. Of course, it's nuts to say we're going to understand the genetic basis of all diseases in a time span relevant to Wall Street. But it's equally nuts to imagine that we won't over the next 30 to 40 years."

Hard-hit genomics investors may not find much to cheer about in that forecast. But here's some solace: The medical advances Lander envisions may let many of them live long enough to recoup their losses.

FEEDBACK: dstipp@fortunemail.com