Celera Genomics Group appears assured of its place in history with tomorrow's announcement that the assembly of the human genome - the genetic instructions for creating and running a human body - is largely complete. Now it faces what might be an even greater challenge: turning scientific triumph into profits.
Celera, which is based in Rockville, and the rival public Human Genome Project will jointly announce the milestone in Washington shortly after noon tomorrow in a public demonstration of harmony after months of mutual recrimination. But friction might quickly resurface, as Celera seeks to build the world's most comprehensive - and proprietary - repository of information on genes and proteins.
Celera plans to make much of its money by selling subscriptions to its databases - and access to software for searching them - to pharmaceutical companies, scientists and individuals wanting to use the new knowledge to treat or prevent diseases.
"The future will start soon," promises J. Craig Venter, Celera's president, of the company's plans to sell genetic information.
For now, however, 2-year-old Celera is losing money. It has just begun to build the many databases it promises will help identify genes that trigger disease. The computers that will help deliver its planned next generation of products - extremely high-speed protein sorters and analyzers - are still in development. Only five pharmaceutical companies and one academic institution, Vanderbilt University in Nashville, Tenn., have been identified as paying customers.
And, though Celera considers the mapping of genomes - fruit fly, human and, now, the mouse - the foundation of its efforts to build an unparalleled database of genetic information, some argue that such extended sequencing of the genome isn't very marketable.
About 5 percent of the genome contains genes, stretches of DNA that contain the instructions for making the hundreds of thousands of proteins that are the body's building blocks. The rest of the genome is padded with "junk DNA," which serves no known purpose.
The pharmaceutical industry is interested only in the genes, not the junk. And there are other ways to find genes besides sifting through the genome. What's more, Celera Policy Planning Director Paul Gilman concedes that the human genome it has assembled contains gaps, mostly in the regions of junk, though some genes might be missing as well. A truly complete genome, with all 3.1 billion chemical units assembled in the proper order, might not be finished for decades.
All of this leads Roy Whitfield, chief executive officer of Celera competitor Incyte Genomics, to sum up Celera's assembly of the human genome this way: "It's a commercial nonevent."
Celera hasn't needed a money-making product to be a hit with investors. Its image as a center for fast-paced, technology-driven science and its high-flying stock price were built on the drive and accomplishments of Venter, a scientist whose techniques for finding genes have helped revolutionize pharmaceutical research. Venter hasn't been shy about predicting his company's success.
In a recent interview, Venter claimed that Celera would have a facility capable of taking the next big step in biology, figuring out the chemical structure of the up to 1 million proteins produced by the body's 80,000 or so genes. To do this, Celera will need computers so fast they haven't been built.
Comments such as these helped propel Celera's shares toward the stratosphere early this year. The fledgling company took advantage of the rise to sell 3.8 million more shares at $225 each, raising hundreds of millions of dollars to supplement the deep pockets of its parent, PE Corp., which is based in Norwalk, Conn.
Poised to be a leader
Celera plans to plow the money back into the company, building its protein-sequencing laboratory and perhaps making more acquisitions such as its purchase of Paracel Inc., a maker of hardware and software for genomic computing.
As a result, admirers such as Wall Street analyst Winton Gibbons of William Blair & Co. say Celera is poised "to become the leading supplier of genomic information." Gibbons predicts Celera's revenue will grow rapidly over the next several years, increasing to $124 million in fiscal 2002 from $12.5 million in fiscal 1999.
Jim McCamant, editor of the newsletter Medical Technology Stock Letter, believes Celera could just as well be on the edge of a financial precipice.
"There are a lot of people who have been doing that for years," McCamant said about Celera's plans to provide information on proteins that genes set into motion. "No matter how much money you spend, it's not something where you can catch up."
Recent announcements indicate that Celera plans to build its databases partly through collaborations with other companies, meaning the databases could grow in value faster than some critics think. Celera and Geron Corp. of Menlo Park, Calif., for example, are working to identify genes at work in stem cells - the cells from which all other cells in the body evolve. The genes could give pharmaceutical companies important clues to Alzheimer's disease, aging and the regeneration of damaged organs.
Taking business elsewhere
Despite Celera's scientific accomplishments, many potential customers haven't signed on with the company and, in some cases, have signed up with competitors. One, Palo Alto, Calif.-based Incyte, has 18 pharmaceutical customers and competitor Gene Logic Corp., based in Gaithersburg, has at least a dozen.
Celera, however, has been focused on building databases and searching tools, not selling subscriptions. The company emphasizes that it signed up the big subscribers it has - heavy hitters Amgen Inc., Norvartis AG, Pharmacia Corp., Pfizer Inc. and Japan's Takeda Chemical Industries - without trying.
The five asked to sign with Celera early, giving them the first possible look at Celera's data while agreeing to help the company work out the bugs in it.
Knowing the location and function of genes is important to pharmaceutical companies researching ways to fight disease. Genes issue the directives to proteins that do the work in cells. Researchers who can identify a mutated gene that produces a mutated protein that results in cancer, for example, might be able to arrest the process.
Though nearly 99.9 percent of DNA is believed to be the same in everyone, scientists have discovered that even single-letter variations in DNA can make medicines that help one person but are toxic to another.
As a result, in addition to mapping the human genome, Celera is developing a database of such single-letter variations - single nucleotide polymorphisms - in hopes of selling the data to pharmaceutical companies. That database, Venter said, should contain about 6 million such SNPs (pronounced "snips") next month - far more than the 100,000 or so identified by a consortium of pharmaceutical companies.
Some other key Celera databases - such as the fruit fly genome and the yet-to-be-finished mouse genome - are designed so that researchers can compare DNA in creatures commonly used for laboratory research to human DNA. Humans share many genes with other creatures.
Celera, which generally strikes five-year deals with customers, charges university researchers $1,000 to $15,000 per lab per year for database access. It charges major pharmaceutical subscribers $5 million to $15 million annually. Overall, Venter says, the company has about $200 million in committed revenue over the next five years from subscribers.
Celera is burning up cash faster than subscriptions are bringing it in and expects to keep piling up losses. The company spent $43.5 million on research and development in the third fiscal quarter, which ended March 31, far outstripping revenue of $11.1 million.
Focusing on sales
Tony White, chief executive officer of Celera parent PE Corp., told analysts during the company's April 26 earnings conference call that research and development expenditures related to the human genome will be reduced during the next several months while sales efforts are increased. But the company's expenditures will grow again, as it begins heavy spending on protein-related systems. Meanwhile, many big subscribers that have deals with Celera also have them with Celera's competitors. Pfizer, for example, has agreements with Incyte and Gene Logic.
Venter estimates that the database portion of Celera's business should break even by 2002, but he declines to disclose revenue and profit targets. Gilman said the company has none.
One challenge Celera has in turning losses to profits involves convincing smaller companies - which can't afford to simultaneously contract with numerous genomic information companies - that Celera has what they need. Corixa Corp., for example, signed up with competitor Incyte because Corixa executives were convinced that Celera couldn't help them develop experimental vaccines.
The Seattle-based company is trying to find and manufacture proteins that coax patients' immune systems to attack cancer cells. Corixa begins by using Incyte's database of expressed sequence tags, bits of genes synthesized in a laboratory.
Dr. Mark Magnuson, assistant vice chancellor for research at Vanderbilt University in Nashville, Tenn., believes that Celera's broad offerings give researchers incredibly powerful tools. Vanderbilt just became the first university to sign up to use the Celera Discovery System database via a secure Internet connection.
In addition to gaining access to a computerized collection of gene bits, or expressed sequence tags, like those Corixa is using, Magnuson is excited about the fly, mouse and human genomes that the deal includes. He plans to electronically align the segments of mouse DNA with those of human DNA as he researches what triggers a particular gene - the glucokinase gene - to cause adult-onset diabetes.
"You could just say, 'Show me glucokinase for human,' and it's there. 'Show me glucokinase for mouse,' and it's there," he said.
"Regions that are similar will just pop right out, and we can say, 'Oh, here it is,'" Magnuson said.
The degree of Celera's success depends not just on how much more useful its databases will be than others available, but also on how long they'll be more useful and to whom.
The public project, sponsored by the National Institutes of Health and the Department of Energy, has been putting its data on the Internet, open to all, as it has been gathered.
About 35,000 sites tap into the free public data daily. Anecdotal evidence shows that scientists looking for a specific gene are by far the most frequent users.
Private vs. public
For that kind of research, said David Lipman, director of the National Center for Biotechnology Information, the agency coordinating data for the public project, there is little advantage to using Celera's "finished" data.
"For most people, there's not much of a difference," Lipman said. But, he noted, if a researcher wants to compare, say, mouse and human data, Celera's database would be superior.
Others say the public data aren't superior in any way. William Blair analyst Gibbons says the data are so filled with errors and so difficult to search that Celera - with its super-fast computers - has an undeniable advantage.
Lincoln Stein, a researcher at Cold Spring Harbor Laboratory in New York who develops software for searching genomic databases, agrees that the public data are flawed. "It would require a massive bioinformatics effort to use it for anything," he said.
Stein believes there is a market for a private company to make money by cleaning up the public data and making it easier to search, but whether Celera will achieve that remains to be seen.
"It depends on what they've got, how they execute it and how they're going to deal with the volatility of their data," Stein said. "The entire scientific community is going to be working on a publicly available sequence and publishing the results. Eventually, [it will be] Celera's biologists and bioinformatics staff against the rest of the community."