By Randall Osborne

Editor

Over the years, biotech has seen plenty of Holy Grails come and go so many that the phrase has become a weary cliche in the industry. When researchers began finding single nucleotide polymorphisms (SNPs), and linking these DNA variations to the propensity for disease, the science of doing so became yet another so-called key to drug design.

Maybe SNPs work really will prove to merit the ballyhoo. But, with the oft-bemoaned logjam of genomics data, the question could become whether investment in firms based on SNPs efforts will pan out, or pan out soon enough to make enough difference for those who are banking on it.

Many are banking on it, and some in a big way. Recent headlines include the $330 million merger of PE Biosystems with Third Wave Technologies Inc., which brings to the table its Invader SNPs analysis method.

Third Wave also has a deal with the Sanger Centre. A month after the center disclosed its completion of the chromosome 22 DNA sequence, it made known its collaboration with Third Wave to build the first high-density SNP map of the chromosome, using Invader technology.

PE Biosystems also has a deal with Illumina Corp., in which the latter will use its BeadArray technology, which deploys fiber optic bundles for SNP analysis. Illumina recently raised $28 million to advance the method, and company officials figure to have products on the market in a year or less.

Sequenom Inc., focused on industrial genomics, said late last year it would seek $70 million in an initial public offering (IPO) to expand its SNP research, and LJL BioSystems Inc. formed a Genomics Science Group to extend the firm's high-throughput screening strategies to the genotyping of SNPs.

"Everybody's going to have to move in this direction," said Paul Kelley, an analyst focused on genomics in drug discovery with ING Barings LLC in New York.

SNPs findings, Kelley said, ultimately may fix the most expensive problem that bedevils drug discovery: the fact that more than 80 percent of compounds optimized to some degree fail to pass muster in the clinic. One study calculated a 30 percent dropout rate as late as Phase II, he noted.

Of these, "a large percentage have activity in some people, but fail because they're not effective in the population at large, or are highly toxic in a small sub-group of patients," Kelley said, and even prescription drugs on the market are estimated to have limited efficacy in about a third of the population they are approved to treat.

SNPs research, some day, could change that. Kelley said it's not unreasonable to question whether investors ought to jump immediately aboard firms based on SNPs technology, since they are adding to the genomics-information glut. But, he said, complementary methods are being devised to analyze SNPs data. Furthermore, the whole enterprise will move no faster, or slower, than the market permits or demands.

"It's not going to go slow," Kelley said. "There's [already] a mad rush into the area. But you're going to see the technology first used in the most tractable fashion that is, in a limited set of target genes, and a limited set of genes that will bear on [drug] toxicity."

Major pharmaceutical companies are being forced to invest in SNPs work, or (more commonly) ally themselves with biotech firms following that line of research, he said, mainly because it promises to solve problems that once seemed an ordinary, and very costly, part of doing business.

"The companies wouldn't go down this route, if they had any other choice," he said. "Sheer economic exigencies have pushed the industries to this position." Giants such as SmithKline Beecham plc have made known they are placing SNPs in a high priority position, he noted, and Glaxo Wellcome plc also has a strong program in place.

Pharmaceutical companies, Kelley said, have been accustomed to the "one drug fits all" approach, with its attendant expenses. With SNPs comes the possibility of "atomized" drug design, which means fitting the drug to the patient, as nearly as might be done.

"We're already seeing that in HIV, not [matching therapy to] the patient necessarily, but to the virus," Kelley said. "That presages what we'll see in human organic disease, down the line."

As more pharmaceutical firms travel the SNPs road, more must do the same in order to compete.

"It's a limited tool at the moment," Kelley allowed. "Genetic variation, or variation at the level of the gene sequence, does not automatically inform whether that gene product has a functional difference."

Again arises the problem in genomics, which is making sense of data that has been gleaned from emerging technologies.

"The market for SNP genotyping tools will ultimately be over $500 million," Kelley said, adding that these are "back-of-the-envelope calculations. We're just peering over the horizon, and you can come up with numbers that are fantastic."

In other words, the investment potential in SNPs, counting what's required to analyze the findings that come out of such research, is immense.

"If you're doing a SNPs study in the clinic, and you find out the drug doesn't work in a certain percentage of the population, then you need to figure out what the causes are, and whether you can develop a predictive test," Kelley said. "That's like drug discovery in the rear-view mirror."

He said the way around such circuitous methods is typified in the deal between Structural Bioinformatics Inc. (SBI) and Quest Diagnostics.

"Quest is the biggest developer of HIV gene sequence data," Kelley said. "What [the company] will do is feed raw sequenced data from HIV genotyping to SBI, which will develop structures from those sequences, and the two will sell those to drug discovery companies to help them develop second-generation drugs. It's the first ever of that kind of study that I know. Ultimately, this will happen with everything."

Illumina President and CEO Jay Flatley said the field of SNPs is chaotic and confusing now, but will standardize to some degree, and already is more "binary" than the challenging realm of gene expression.

"It will be less intimidating than other bioinformatics problems," he said. "With SNPs, you generate lots of data when you have lots of patients. But there will be some standard SNPs that [researchers] use, and they'll run them against groupings of populations. There will be some period where that gets sorted out," he said, probably a stretch of several years or so.

"People think there's somewhere between 3 million and 6 million SNPs in total," Flatley said. "That's not a huge number, and we're going to find out where most all of them are. Some will be in genes, and some won't, but it's likely the ones in genes will be more relevant. We don't know that for sure, but we suspect that. And these studies will be done at different companies," so the cost will be distributed, he said.

Like gene sequencing, SNPs-finding will evolve into a hybrid public and private effort, Flatley predicted. At least for now, The SNP Consortium a $45 million project begun last spring is putting its discoveries in the public domain, while firms such as CuraGen Corp. and Incyte Pharmaceuticals Inc. are developing proprietary SNPs.

"The value [of those will] decrease over time," he said. "The information companies will be motivated to make that information available, maybe selling it in platforms like ours."

Information gleaned by SNPs work will be plentiful, but not altogether overwhelming, Flatley said.

"Certainly, it's more data, another layer of information," he said. "The big problem with data coming out now, is that people don't know what [the data] mean. SNPs themselves are only the beginning. If you find 20 SNPs together that are somehow related to heart disease, for example, which ones are they?"

Kelley, the analyst, said everybody including investors is likely to benefit from the pharmaceutical industry's facing of hurdles presented by SNPs technology. More and more, the industry will turn to biotech firms to help develop effective, "tailored" therapies, which "should defray 'service-related' costs [brought on by therapies that miss the mark], which are far and away the biggest costs in the industry."

In any case, he said, now that SNPs have been found and researchers have begun, at least, to explore what they mean, the widening, deepening project must continue.

"It's a tide that can't be turned back," Kelley said.