Skip to Content

Consulting Biotech’s Oracle

The CEO of Human Genome Sciences, Bill Haseltine, has created a powerful new tool that he just knows will revolutionize the discovery of new medicines.
October 1, 2001

Sitting in the conference room of Human Genome Sciences on a Friday the 13th, Bill Haseltine has a grin stuck on his face that says, today is a lucky day. It is the smug grin of a boy showing off an incredibly cool and impossible-to-find new toy. And make no mistake, the CEO of Human Genome Sciences has an incredibly cool new contraption at his fingertips. He also has a visitor whom he would love to impress: AIDS researcher Anthony Fauci, director of one of the most influential branches of the National Institutes of Health.

Haseltine’s toy is actually one of biotech’s most comprehensive databases. By merging a crack team of molecular biologists with computer jocks, the company has generated over the past eight years a user-friendly database of human genes and the proteins they code for that-Haseltine asserts-contains more clues about how to treat and cure disease than all other related databases around the world, commercial and public, put together. It’s just the sort of outlandish boasting that has made Haseltine (pronounced “hazzle-teen”) one of biotech’s most controversial figures. But backing up his claim, the Rockville, MD, company already has moved five drugs into human trials, more than any other genomics-based biotech company.

“Tony, the feeling we have here is we’re doing what most of the world will be doing in 10 years,” says Haseltine, who at the age of 57 has slicked-back, thinning hair and wears owlish glasses. He then looks across the wide conference table that separates him from Fauci, pausing for effect, and says, “Maybe”-as in, maybe the clueless establishment will figure it out by then. Haseltine, who in his double-breasted suit looks more like a Wall Street moneyman than a former Harvard biology professor, follows this jab with a goose-honk laugh and rocks in his leather-lined chair.

Fauci is here to explore one of the potential gold nuggets found by Human Genome Sciences, a protein that stimulates antibody production. But while he’s visiting, Haseltine agrees to let him ask the database-which Haseltine has nicknamed the Oracle-any question that suits his fancy. Fauci says he’d like to search for the “elusive CD8 factor.”

Since 1986, AIDS researchers have known that HIV-infected people who defy the odds and suffer no immune damage spit out a mysterious chemical “factor” from a specific type of white blood cell dubbed CD8. Try as they might, AIDS researchers have failed to isolate this factor. “If we can’t get the factor out of this, we can’t get it,” says Haseltine, as a coworker with a laptop begins to search the Oracle for all proteins produced by CD8 cells.

A projector hooked to the laptop shines its display onto a screen for all to see. The Oracle reveals that so far Human Genome Sciences has found 64 different proteins secreted by CD8 cells. At the time of Fauci’s visit, a whopping 59 of these proteins had never been described in the medical literature or in any public database containing genetic and protein information on humans.

Fauci is beside himself. “That’s terrific. It’s amazing. It’s breathtaking,” he says. “I’m serious. There it is. This is fantastic.”

Temple of Genes

For three days, I watched visiting scientists drop their jaws to the floor as Haseltine and crew put on similar shows. The idea that everyone but Human Genome Sciences is panning the wrong river for genetic gold is a story that Haseltine shops expertly. It is a story that portrays the company as miles ahead of the many biotech and pharmaceutical firms that similarly are trying to make money from genes. Haseltine, both brilliant and brazen, is of course the hero, and there is an entire cast of antiheroes. The story ends with wondrous new drugs coming to market at lightning speed.

And it’s a story that many investors are buying: his company’s stock, of which he owns more than three million shares (after cashing in $56 million worth this spring), traded for between $35 and $107 in the past year. If Haseltine overinflates the tale, well, only one letter distinguishes hope from hype.

The other side of the story begins with the Human Genome Project, an international $3 billion effort largely funded by the U.S. government. In 1990, the project organized academics around the world to decode the entire sequence of human DNA. Separate from that project, Human Genome Sciences became one of dozens of biotech companies that sprung up in the early 1990s with their own fleets of machines working 24/7 to scour the raw sequences of As, Cs, Ts and Gs-the abbreviations used to designate the four chemical building blocks of a DNA molecule-for genes. All were racing to patent as many genes as possible before the data would go public.

Shortly after he cofounded Human Genome Sciences in 1992, Haseltine left Boston’s Dana-Farber Cancer Institute and tenure at Harvard University to become CEO of the new company. Within months of hanging up his lab coat, he signed a landmark $125 million deal with SmithKline Beecham (now GlaxoSmithKline) that gave the pharmaceutical giant exclusive rights to search the Oracle for leads on “small-molecule” drugs-the kind of pills that people swallow. Human Genome Sciences, however, retained the rights to develop treatments based on proteins-larger molecules, like insulin for diabetes or erythropoetin for anemia, that have to be injected. The deal blew minds in the biotech industry, which until then attracted investors based on the promise of bringing a drug to market, a process that could take over a decade and burn hundreds of millions of dollars. Human Genome Sciences showed that genomics companies could enjoy a steady revenue stream by selling information.

This bold move set off a cascade of events that changed what it meant to be a biotech company. Incyte Genomics, a Palo Alto, CA-based firm that sequences DNA, soon began selling access to its data and declared that it had no intention of even making treatments. Millennium Pharmaceuticals in Cambridge, MA, and Genset in Paris, France, soon cut multimillion-dollar deals of their own to help pharmaceutical companies find drugs by hunting through populations for disease genes. Next, a whole series of biotech firms sprung up around the idea of helping companies figure out the functions of genes, or the different proteins each gene instructs the body to make (see “The Proteomics Payoff,” ). “If all we did was be a catalyst for this change, which we already have done, we’d be a success,” says Haseltine.

Compared with other biotech startups, Human Genome Sciences had an inside edge as a result of its unusual relationship with a nonprofit outfit headed by J. Craig Venter. Venter left the National Institutes of Health because it refused to back a shortcut he had developed to sequence the genome, and in 1992 he signed a deal to join Haseltine in an elaborate business venture. Venter headed up the new Institute for Genomic Research, which sequenced DNA, while Haseltine ran for-profit Human Genome Sciences, which bought the institute’s data and marketed it to pharmaceutical companies.

However, the two men’s goals were as ill matched as their personalities. Venter wanted to publish data that Haseltine believed was proprietary. And soon Haseltine thought better of spending $10 million a year buying data from Venter’s firm; Human Genome Sciences, he decided, could set up its own in-house sequencing shop and do the job more cheaply. In 1997, Venter and Haseltine formally severed their business ties. To this day, the men continue to engage in what Haseltine’s sister Florence, herself an official at the National Institutes of Health, refers to as a “pissing match between alligators.”

While Venter went on to cofound Celera-a company that, in June 2000, at the same time as the Human Genome Project, completed its own draft of the sequence-Haseltine became an outspoken critic of these massive sequencing efforts. Indeed, he has taken a view contrary to much of the conventional wisdom surrounding the Human Genome Project, including the growing consensus that humans may only have 30,000 to 40,000 genes-not the 100,000 that most scientists had previously predicted.

Haseltine, true to form, insists that a serious case of groupthink plagues the field. He still maintains that humans have at least 100,000 genes and might even have as many as 120,000. Haseltine knows this because, he claims, his company  already has 90,000 distinct genes frozen away. “Why did they miss these?” Haseltine asks. “Because they decided genes have to have some similarity to known genes.” And the majority of the genes Human Genome Sciences has in its freezers, he says, “have virtually no similarity to anything found before.”

Human Genome Sciences has not published evidence to support these controversial claims, but the multitude and variety of whirring machines that continuously feed data into the Oracle make it difficult to dismiss Haseltine out of hand. His company has spent the last eight years sequencing genes, intensively studying the proteins they code for and simultaneously identifying potential drugs; other firms tend to have much more circumscribed goals. So Haseltine’s real redemption will come if he fulfills his promise to use the database to turn out actual, life-saving drugs. His company is focusing on the 10,000 genes that it knows code for proteins found on the outsides of cells, so-called secretory proteins that include hormones, receptors, immune-system messengers and enzymes.

So far, his company has moved five drugs into human trials that, if they work, may speed the healing of wounds, make cancer treatments less toxic, allow people with heart conditions to avoid bypass surgery, treat hepatitis C and spare the limbs of patients who otherwise would need amputations. By the end of the year, the company hopes to move at least three more new drugs into human clinical trials. And in July, the company reached the end of its commitment to provide GlaxoSmithKline access to the Oracle, creating even more exclusive opportunities for itself. “We are like kids in a candy store,” says Haseltine.

Outside of GlaxoSmithKline and Amgen, the world’s largest biotech company, no one else has yet used genomics to bring a drug into the clinic, says Haseltine. In his judgment, many scientists, in both industry and academia, simply don’t understand how to mine the human genome for drugs, wasting time on regions of the genetic map that Haseltine has already dismissed as worthless. He says he has “a deep concern” about what the Human Genome Project will mean for people. “So far, it’s a mixed blessing at best,” he says. “The gene, for my purposes, is part of an anatomy.Human Genome Sciences is going to redefine human anatomy. We’re going to take it to a new level of resolution.”

Barking Dog

To appreciate why Haseltine has the audacity to pronounce that his company sees the light while others continue to grope in darkness, consider that Human Genome Sciences, like most young biotechnology companies, has no products on the market and, so, must sell its vision. It also helps to know a few things about William Alan Haseltine.

Haseltine and his three siblings grew up on a naval base in China Lake, CA, a “secret city” in the Mojave Desert where their father and other scientists designed the Sidewinder missile and the ejection seat used in fighter jets. Their mother, Jean, who taught French on the base, required frequent hospitalization for manic depression and a series of serious physical ailments, including severe psoriasis and a myopia that stressed her eyeballs and made her retinas detach. At seven, Bill, too, became ill with a heart condition called pericarditis that kept him out of school for six months.

“I did not like being sick, and I hated my mother being sick,” he says. “I was terrified that she was going to die of blood poisoning. She had terrible psoriasis. I would actually go in and watch those red streaks go up her arm. And I knew if those red streaks went too far she would die. And that was a very upsetting thing. Kids are likely to feel responsible. It was a hopeless feeling.” The turmoil led the young Haseltine to medicine. “I wanted to be a doctor to cure these diseases,” he says.

After earning a BA in physical chemistry at the University of California, Berkeley, in 1966, Haseltine decided that his true love was research science, and he entered a PhD program at Harvard studying under Walter Gilbert (see “Bankrolling the Future,”). Gilbert, who shared the Nobel Prize in chemistry in 1980, remembers Bill as “a very lively student” who “alienated” some of the other grad students. Haseltine went on to do a postdoctoral fellowship at MIT in the lab of David Baltimore, who himself would win, in 1975, the Nobel Prize in medicine. “Bill was very smart and dominating,” says Baltimore, who is now the president of Caltech. “He did some wonderful work, and he didn’t make a lot of close friends. Bill is out to do as much as he possibly can in the world, and in some ways, it’s good for the world, but it doesn’t make him a beloved figure.”

Haseltine moved on to Dana-Farber, which is affiliated with Harvard Medical School, rising through the academic ranks to become a full professor. He accumulated a rsum that also includes starting several biotechs, raising two children, becoming a big-shot AIDS researcher who hobnobbed with the likes of Liz Taylor, getting divorced, and marrying Gale Hayman, cofounder of Giorgio in Beverly Hills, CA, inspiration for Judith Krantz’s Scruples and author herself of How Do I Look? The Complete Guide to Inner and Outer Beauty: From Cosmetics to Confidence. Along the way, Haseltine impressed colleagues with his polymathic, capacious mind and simultaneously irritated them with what critics like Leroy Hood (see “Under Biology’s Hood,” TR September 2001), a leading figure in the biological research community, call his “arrogance and infinite selfishness.” As Hood, who recently founded the Institute for Systems Biology in Seattle, puts it, “Bill raises as much animosity as admiration.”

When I visited, Haseltine’s office at Human Genome Sciences offered another clue as to why he rubs so many scientists the wrong way. A table framed by a long wall of windows looking into the Maryland woods had on it little plastic-coated wire stands displaying a few dozen scientific journals, books and popular magazines. The collection was, in its intellectual sweep, mesmerizing. But for all that gravitas, the display was obviously there for show-who puts what they’re reading on little stands?-and was meant to dazzle the visitor. Haseltine’s words, similarly, often aim to dazzle, which is antithetical to the belief of many scientists that data should speak for themselves.

Sitting in his office, I ask Haseltine about his legion of critics, and he smiles. “The dog barks and the caravan passes,” he says in French. Then, in English, he adds, “Who gives a good goddamn what people think? Do well and let them say what they want.”

There is substance behind Haseltine’s bravado, which is obvious to anyone who sees a demonstration of the Oracle-something that, remarkably, few leaders of the genomics revolution have done. “They have a very arrogant, self-centered view, which is they are the world, they are the heroes, they are the white knights,” says Haseltine. “I don’t think that people have any idea of the power of what it is that we do, because it’s two or three steps beyond what they can imagine.”

The back wall of the company’s conference room explains another reason that relatively few of the world’s scientists have had access to the Oracle. The wall is cluttered with dozens of bronzed versions of official documents from the U.S. Patent and Trademark Office. In fact, by July of this year, Human Genome Sciences had 179 “gene-based” patents and had filed patent applications on at least 7,500 other newly discovered genes for which it has declared medical utility. Anyone who wants to use the firm’s database must agree to give up any medical utility discovered. So researchers stay away.

The Answer

One morning, I observe a meeting with scientists from a biotech company that might form a joint venture with Human Genome Sciences. Haseltine takes them on an Oracle tour. “We have more answers than you have questions,” he says at the outset.

As much bluster as Bill Haseltine pumps out, as combative and grating and self-serving as he may be, his company’s database speaks for itself. Haseltine begins scrolling through the endless DNA samples in the Oracle. He can look at a particular gene and see how it’s expressed in a six-week-old embryo compared to, say, an embryo that is nine weeks old. Likewise, he can compare the genes expressed in a fetal kidney with those from an adult’s, a healthy ovary to one riddled with cancer. He randomly selects a gene, jumps to a public database run by the National Institutes of Health, and finds a similar gene present in worms. Another click of the computer shows that the protein is secreted. Human Genome Sciences has done 69 different biological tests with the protein, looking at how it relates to everything from myeloid leukemia to immune-system cells. A diagram even shows the biological pathway within which the protein operates. “We’ve already patented it,” says Haseltine.

Haseltine decides to look at levels of the protein in an adult kidney. His company has found 583 genes that are expressed in kidneys, a full 363 of which have not been described in public databases. Of these, 52 code for secreted proteins, and 27 of those Human Genome Sciences has filed patents on. Haseltine decides to swing the Oracle in another direction and look for the most abundant gene in this kidney sample. It’s septin. “Who knows what septin is?” he asks. “Not me. We’re about to find out.” Septin turns out to be involved with blood clotting.

What does all of this mean to the discovery of new medicines? It is plain that Human Genome Sciences has built a jazzy new type of microscope that, as Haseltine says, offers a view of human anatomy that we’ve never had before. What is much less clear is when this knowledge will help humans lead longer, healthier lives.
Francis Collins, the researcher who heads the Human Genome Project for the National Institutes of Health, says scientists have to strive not to oversell the promise of genomics. “I have no doubt that in 50 years, much of medicine will look entirely different, and much of that will be because of genomics,” he says. “It’s a revolution unlike almost any other that’s happened since the discovery of antibiotics. But we have to be honest with the press and the public and ourselves that the timeline is longer than we wish.”

Haseltine, in contrast, hears the wheels of a caravan turning and the faint sound of dogs barking in the distance.

Keep Reading

Most Popular

This new data poisoning tool lets artists fight back against generative AI

The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models. 

Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist

An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.

Data analytics reveal real business value

Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.

Driving companywide efficiencies with AI

Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.