There are many good reasons for bailing out of the first year of medical school: running out of money; hitting the wall in advanced biochem; deciding your talents would be better suited for the Peace Corps, etc. Eugene Chan (pictured left) had his own excuse. Once you've had one of the biggest, sexiest scientific ideas of the decade, it's hard to keep showing up for classes.

Chan's big idea--born in his dorm room at Harvard Medical School--was to build a machine that could decode an entire human genome in about 30 minutes. The device he envisioned could untangle the double helix of DNA and read each nucleotide in linear fashion, much like it happens in nature. Getting that M.D. suddenly didn't seem so important. "The idea was too good to lay dormant," Chan says in a phone interview.

He saw the implications immediately. Instead of having just a few sample genomes representing the entire human race, researchers could collect millions of variations on the theme, shedding light on all of the little genetic quirks that help to make us individuals. The machine could help unravel the genetic basis of many complex diseases, including diabetes, Alzheimer's disease, and some cancers. And doctors would suddenly have a much clearer picture of their patients ("I need a tox screen, a chem panel, and a full genome, stat!").

Gene Machine

The quick, cheap genome is still at least several years away, but Chan's big idea is already paying off. After dropping out of medical school at the age of 23, Chan turned his inspiration into a company called U.S. Genomics. Founded in 1997, the company has already brought in $57 million from investors and is poised to release its first product, the Trilogy Single Molecule Analyzer, later this year.

Trilogy can't read a complete genome, but it can scan single molecules of DNA or RNA--large or small--without the need for polymerase chain reaction (PCR) or any other kind of amplification. The machine's basics are essentially the same as those Chan envisioned 7 years ago. The first step involves tagging target sequences with fluorescent markers. Then comes the breakthrough: The sample runs through a nanochip that unfurls the tangle of nucleotides into a single long strand. An optical scanner then detects the fluorescent tags with microscopic lasers and feeds the information into a computer that creates genetic maps.

Molecules of DNA or RNA can move through Trilogy at a rate of 10 million to 30 million base pairs per second. At this point, the machine can't come anywhere close to reading each A, C, G, and T that whizzes by, but it can reliably detect tagged sequences. Trilogy can also count the levels of a given protein in any sample. Lab technicians who spend their days working with bulk fluorescence should note the key word: count. The machine physically tags, differentiates, and counts the molecules. The end result is an actual number, not a signal that suggests a number.

Counting proteins and analyzing snippets of DNA or RNA isn't quite as exciting as reading an entire human genome, but the research labs that are lining up for the machine at roughly $125,000 to $150,000 apiece don't seem too disappointed. They will soon be able to address directly many pressing questions: Is this virus associated with that disease? Are these two genes on the same chromosome? How many cytokines are in this tissue sample? The answers could help open many doors in medicine and greatly expand our understanding of the functions of genes.

Out of Place

Chan is accustomed to making big leaps. In his younger days, he successfully made the transition from a small public high school in Hopatcong, New Jersey, to Harvard University. He had been a physics prodigy in high school, the kind of kid who reads textbooks recreationally and wins statewide contests against much older competitors. After an advisor at Harvard told him that biology would drive the next wave of science, he switched gears to biochemistry. But physics was never far from his thinking. He still saw all of those organic molecules as physical entities, objects that could be unraveled, manipulated, and understood.

The small-town kid felt out of place in the Ivy League, but not for the reasons one might expect. "I got the top scores in every single class I took for the first 2 years," he said in a phone interview. "I knew it wasn't supposed to be like that." Surprise soon turned to frustration. For one thing, he didn't like dealing with grad-student TA's who weren't on his level. More significantly, he got tired of tackling problems that already had answers. "Brilliance comes in being able to answer unanswered questions," he says. "I realized I had something to contribute. I had a chance to create my own unique world."

Much of his world is centered at U.S. Genomics' headquarters in Woburn, Massachusetts. The company currently employs 55 people, including 40 on the tech side. The roster of Ph.D. scientists includes molecular biologists, biophysicists, computer engineers, mathematicians, and experts in bioinformatics, the science of finding patterns in complex biological data. Even as a 20-something dropout, Chan had little trouble luring top-notch researchers to his company. "He's a very charismatic guy, and his vision can be infectious," says Stephen DeFalco, a former president of PerkinElmer Instruments who replaced Chan as the CEO of U.S. Genomics earlier this year. "People knew they were going to be a part of something pretty damn interesting."

In his 6 years as CEO, Chan also won the confidence of many investors. He raised three rounds of venture capital, a great accomplishment for someone with no formal business training. He accomplished this partly by earning endorsements from scientists like J. Craig Venter, president of The Institute for Genomic Research and current board member of U.S. Genomics. (Venter, of course, was at the forefront of the first effort to decode the entire human genome.) "I rely on top-tier scientists to tell other people that Eugene has a great idea," Chan says. "Extremely good scientists recognize raw talent instead of resumes. If you're an A- person, you're not going to be able to spot A+ talent."

Over the years, even Chan's supreme confidence couldn't make up for his lack of business training. It eventually became clear that the company could go even farther with a businessperson at the helm. "There comes a time when investors don't want to see a technical founder [as the CEO]," DeFalco said in a phone interview. "In the old days, they would fund a great idea. Now they want to fund a business plan."

Chan is now the "vice president of advanced platforms," but, in many respects, U.S. Genomics is still his company. It still runs on his vision, and it still reflects his character, especially his commitment to openness and cooperation. Every Friday, company scientists get together to share their latest ideas. ("It's by far our most well-attended meeting," DeFalco says.) For DeFalco, it's a huge change from his days at PerkinElmer. "If I wanted to get four Ph.D.'s in a room, I'd have to schedule them 3 weeks out," he says. "And when they got there, they'd have to introduce themselves."

Open Books

While the scientists around him take incremental steps forward, Chan still has his sights on the big prize: the quick, cheap genome. He's confident that someone somewhere will make the necessary breakthroughs in tagging, optics, and computing that will help turn everyone's genes into an open book. He envisions an age of personalized medicine where doctors tailor their therapies and lifestyle recommendations to each patient's genetic profile. "Personal genetic information is going to drive health care," Chan says. "People are going to live longer, healthier lives."

Of course, there's always the danger of knowing too much about a person, even if that person is yourself. Chan acknowledges that his dream machine would open up many thorny ethical issues. It goes without saying that some genotypes would make better insurance risks or better employees than others. And although some genetic blueprints point to a life of health and happiness, others spell out illness and misery. If our genomes ever do become open books, who's in charge of the library? "There has to be a deep sense of ethics so information won't be used in the wrong direction," Chan says. He applauds the U.S. Senate's recent vote to approve a bill that prohibits genetic discrimination, and he trusts that the incredible gains in medicine will outweigh any potential harm.

Chan missed his chance to become a doctor, but he doesn't have any regrets. You don't need be an M.D. to revolutionize medicine. With a big idea, a lot of confidence, and plenty of nerve, even a medical school dropout can do all right.