NEW YORK, Dec. 26 - Sequencing the human genome was widely expected to produce a wealth of new drug targets, and the genomics startups that burst onto the scene in recent years have largely delivered. As a result, the research market for genomics companies is vibrant: Worth just $2 billion two years ago, the market is poised to reach $8 billion by 2004, the need for better technology and marketable drugs and diagnostics fueling the growth.
Today, drug companies sift through hundreds of newly identified gene targets and proteins amassed through collaborations with genomics companies and academic institutions. Delivering these targets, however, proved to be bittersweet for the genomics sector in 2001. The once-heady prices paid for genomic databases have slumped as researchers began to sort out data deemed irrelevant and turned their attention to systems biology and characterizing critical proteins.
The genomics sector also has been bruised by technology made too fast too quickly, and is poised to enter a marketplace next year in which more and more traditional tool and tech shops set their sights downstream as more and more of them, flush from the boom times, bend to consolidation.
Leading the curve too quickly
The genomics sector, now entering adolescence, is having to accept this frustrating development: Their pharma and biotech clientele are beginning to suspect that access to too much genomics data can be as confounding as having too little.
As one pharma R&D exec bemoaned, losing sleep worrying about where his next new target would come from is one thing. Laying awake pondering how his unit will process the 10 new targets it will receive in the next week is another.
UBS Warburg analyst Michael Clulow believes some of the technology offered by bioinformatics and genomics companies was simply two or three years ahead of what most pharmaceutical companies could use.
At the same time, however, pharma executives grumble that recent technology and scientific advances are too slow to let them efficiently prioritize their vast combinatorial chemistry libraries and novel drug targets.
"Genomics databases offered [drug researchers] the promise of forward integration," says Signature Bioscience CEO Mark McDade. This appealed to the drug industry not only for the precision of the science, but also because it promised to assemble all the disaggregated targets it produced (think ligands, antisense drugs, monoclonal antibodies.) It’s assembling the pieces that’s proven formidable for the drug industry, says McDade. Consequently, the plodding tempo of Big Pharma research will likely spur a realignment of sorts in the genomics industry next year, executives and analysts say.
Most predict that within 18 months many genomics technologies, largely the laggards in the sector, will be strategically merged or acquired. Furthermore, if the stock market turns around—which is widely expected in the second half of 2002—analysts say that money will begin flowing toward such relatively underfunded genomics tributaries as chemogenomics, proteomics databases, and patient-based pharmacogenomics.
Genomics companies have another advantage: Many analysts have calculated that, on average, genomics companies boast cash reserves of $350 million. So many have enough cash to live on their own through 2002 if revenues continue contracting.
Not all companies in the sector can claim such resources. For example, many content and bioinformatics providers born as late as 2000 are already re-examining how well their business models will hold up in an upcoming season that, like the broader biotech sector, will likely be marked by mergers and acquisitions.
"Companies that have recently discovered they can't make a go of it with their platform are contemplating buying another business," notes Winton Gibbons, an analyst with William Blair & Co., adding that many companies have the cash to pursue such deals. "They expect an acquisition to turbo them to another level."
Gibbons believes there is no single reason for recent acquisitions of genomics companies. Through its $650 million purchase of Packard Biosciences, for example, Perkin-Elmer obtained liquid-handling and gene-expression biochip businesses, and Sequenom's $230 million stock snag of Gemini essentially paid for Gemini's tissue sample database.
I wish to defect.
Perhaps the biggest decision facing many genomics companies next year, says Gibbons, is whether "to jump the fence" from humble tool manufacturing to lofty—and inherently risky—drug development.
The notion is complex. For a company like Celera Genomics, which led the drive downstream with its Axys acquisition, the shift was always part of the grand plan. But for a company like Incyte—which ironically began life as a pharma shop but later morphed into a genomics powerhouse—the decision to "go drug" has been more gradual as growth of its core business first slowed and then stalled.
By comparison, companies like Perkin, Applied Biosystems, and Waters "will never make the move," Gibbons predicts. "Same with Invitrogen. Their businesses are fine, they won't face the product issues like they did this year. There is no reason to shift."
Younger companies, like Applied Molecular Evolution, deCode Genetics, Deltagen, Genaissance Pharmaceuticals, GeneLogic, and Variagenics—which are all on the fence—face tougher choices.
For example, is a business model based on population studies sufficient for deCode Genetics? Is there a big enough big-pharma base for, say, DeltaGen's functional genomics platform? If the answers turn out to be ‘no’ then a few leaders will emerge in each market sub-segment while pieces of the remaining companies will be either sold off or absorbed. One only needs to think of Merck, which bought Rosetta Inpharmatics this year principally for its bioinformatics and gene-expression talent, and Exelixis, which last month paid $110 million in stock to acquire Genomica pretty much exclusively for its cash.
One unit of the sector, though, faced considerable pressure in 2001. Bioinformatics companies had difficulty distinguishing themselves while technology vendors swiftly developed usable interfaces that can link public and private genomic data. Gibbons thinks that consolidation or integration in that sector—anticipated to grow to $3.8 billion by 2005—will even touch leaders like Lion Bioscience and Compugen. Lion CEO Friedrich von Bohlen agreed as much, commenting in an interview with GenomeWeb in early autumn that he would consider being acquired if the terms were right—an uncharacteristic revelation by the brash executive.
The consensus seems to be that while tools and tech shops are fine to start out in life sciences, the long-term end game is all about the drug. "Agricultural work, diagnostics, outsourced research like [the kind] Millennium and Incyte did in their early days, those businesses pay the bills," said an analyst who asked not to be named. "The dual stories fuel near-term growth and help build the infrastructure for drug development. Investors still need to see a drug at the end of the road."
(Amgen's recent $16 billion purchase of Immunex, the largest biotech acquisition in history that came on the heels of Millennium's bold nab of COR Therapeutics, demonstrates the premium put on marketed drugs, even if the drugs were not internally developed.)
Yet with all the scientific risks, costs, and time pressures involved in drug development, few genomics executives will admit they’re prepared to jump the fence. Frank Gleeson, CEO of MDS Proteomics, argues scientists and executives understand that the pieces of the puzzle—genes, SNPs, proteins—all connect at a meta-level. However, at the commercial and development level—where Big Pharma has staked its claim long ago—the connections are not necessarily in an orderly sequence.
People were attempting to advance high-throughput biology 10 years ago, but the genomics revolution, ironically, stalled further development there, said Gleeson. While the genomics revolution opened the door for proteomics companies like MDS, Gleeson, not surprisingly, believes drug companies are now looking at genomics data merely as "confirmatory information."
New mass spec technologies allow drug companies to zero in on the structure and activity of effective compounds. Data gleaned from the gene sequences, therefore, can confirm, rather than point to, the structure and function of proteins.
William Baker, an analyst with GARP Research, sees the structure of proteins and gene-expression data as so much math. Many genomics companies have been launched on the premise that they can aggregate the math differently, better, or faster than their competitors. Baker, who follows GeneLogic and its aggregation of expression data, suggests that large biotechs and pharma shops may move toward a model in which they share those data.
Drug companies are already taking advantage of GeneLogic's novel toxicology screening platform, which is based on gene-expression databases compiled from the many research organizations with whom it has research collaborations.
After all, "if it's all just a pile of math, why not just buy it cheaper? And from fewer sources?" says Baker.