ATLANTIC CITY, N.J. - Cheerful hordes of excited people converged on this gambling city over the weekend for the annual Miss America pageant, with all of its royally sequined pomp and circumstance.

But a few days earlier the tone was different at the convention center, where staid members of the biotechnology industry met to ponder (among other things) the wisdom of placing their bets on genomics.

All agreed the sector's glamour has faded, at least in the eyes of Wall Street.

At the World Genomics Symposium and Exhibition that ended Friday, analyst Charles Duncan, most recently with Dresdner Kleinwort Wasserstein, sought opinions from panel members and attendees.

"Why do you think the bubble has burst?" Duncan asked. "Why aren't we having so much fun anymore?"

The day before, William Haseltine - president and CEO of Human Genome Sciences Inc. and a leader in the genomics charge - acknowledged to the symposium in his keynote address that "this is a difficult time in the United States and a very difficult time for biotechnology and genomics companies in general."

Genomics, he said, is still "transforming the scientific and pharmaceutical industries. I'm aware that's not the reigning and current perception," he added, noting that many have declared genomics' promises "have been much greater than the reality. That's certainly not my belief, but it's an understandable belief."

He blamed a general misunderstanding of what genomics can do, and how fast. For example, single nucleotide polymorphism mapping and other technologies are "meant to be applied to populations where there is a clear linear relationship," such as families. Investors, though, naively hoped that patients in outbred populations could be analyzed so that drug response (or lack of it) might be predicted.

Similar unrealistic optimism attached itself to the idea of using genomics tools as diagnostics, he said.

"It was a popular idea three or four years ago," Haseltine said, before researchers came to realize that patients are "very reluctant to undergo testing unless there is a treatment," which made companies focus their efforts toward the latter instead.

Focus turned out to be the subject of Duncan's panel, titled "Genomic and Proteomic Business Models For Success in the Coming Decade" - a pretty tall order, he said.

"I don't think we've given enough time to these companies to really decide which business models will be successful," he said, echoing Haseltine's belief that genomics' practical potential will be realized only slowly.

"We told institutional investors that they should just focus on drug developers" in the genomics realm, Duncan said, "but I think [they] have taken it to an extreme now." He said it's "important to make a distinction between publicly traded companies and those that are invested in by private investors right now. There is a broad range of business models out there that are interesting, and it really depends on the risk tolerance and patience of the investor base."

Institutional investors, he added, "at least in terms of public equity, are very impatient and not risk-tolerant at all."

Panelist Norman Russell, former president and CEO of Lynx Therapeutics Inc. and now president and CEO of privately held Aviva Biosciences Corp., said he had gone to "lengths to try and uncouple Lynx from its dependency on the service model."

In April, Lynx added itself to the legion of biotechnology firms cutting staff to concentrate efforts and raised $22.6 million at the same time. Specifically, the firm got rid of 30 percent of its employees, and said it aims to expand the user base for its Massively Parallel Signature Sequence technology, which the company said is designed to identify almost all the DNA molecules in a given sample by deploying Lynx's Megaclone system - a library of about 16.7 million short synthetic DNA sequences and their complementary anti-tags for marking and processing each molecule.

Aviva, which combines biochips with cell biology, last year spun off an antibody company, Aviva Antibody Corp., in an effort to focus its efforts on its core business, which it described as developing multiple-force active biological chips.

Russell urged companies and their investors to think about where an enterprise lies in the value chain, whether near the top with products or near the bottom with instruments, reagents and the like.

Also on hand for the panel was Carl Foster, senior vice president of business development for Oxford GlycoSciences plc, a proteomics firm that "moved into the proteomics space about six years ago before proteomics was even really a term." Formerly named Oxford GlycoSystems, the company was a reagent and instrument manufacturer.

"Now we're even making one more step up the value chain, taking some of the results of our proteomics research and putting those into development" through collaborations, Foster said.

"We're doing exactly what [Russell] has described, but it does take a lot of research funding," he said. "Fortunately we have that but many companies do not, and it's not a good time to go out and try to raise that."

Foster warned about "the traditional platform technology, let's go get big pharma to give us $2 million and we'll do a program for them, and try to get some milestones and royalties." But that model yields fewer customers as more firms compete for the business.

Foster said backers of OGS don't want the company to retreat to the old ways.

"If you're in the fee-for-service business, you're hopefully going to be marginally profitable but it's very low risk and you're never going to hit the home run," he said. "Our view is that we're a biotech company, and it's a risky industry and the investors know that. They would like to see us swing for the fences, and that's what we're trying to do."

Risk, however, can be diversified, he said.

"We have a collaboration with Medarex Inc., for instance, where we are providing up to 30 protein targets," Foster said. "They are making human monoclonal antibodies against those targets, and then we're jointly developing those all the way to the market. Obviously, OGS could not develop 30 targets and products on our own, but we can share the risk with Medarex. We're also doing that with several other companies."

Before genomics, he said, pharmaceutical companies said they had only a few bad targets to work with. Now they have many targets to work with, he said - still "bad" ones, requiring validation to make them worthwhile. But providing a service won't make the kind of money that genomics firms will need to survive, Foster said, although consolidation has made whopper deals between genomics firms and pharmaceutical companies much less likely.

"It was over a year and a half ago when we had the last uber-deal' that came out," he said, referring to the $1.34 billion agreement between Bayer AG and CuraGen Corp., disclosed in January 2001 and focused on obesity and adult-onset diabetes.

"You just don't see those anymore," Foster said.

Russell agreed. Pharmaceutical companies are not overburdened with targets; they're just bereft of the technology to sort them out, he said. "Working with pharma to bring forth better validated targets in partnership is the way forward for biotech. Genomics and proteomics tools companies have really got to face up to that. It's unlikely that pharma companies now are going to pay a $10 million access fee and $5 million in subscription fees for your services. These days are past."