But the technologies in place will surely buckle under the weight of too many A, C, T, G sequences, experts say.

Bioinformatics -- the mining of biological data using information technology -- is the buzzword du jour in biotech. Bioinformatics companies, however, are using the brute force of old technologies to crunch their data.

Unless someone introduces a new and unique technology to process all this information efficiently, experts say, the decade-long, $3-billion Human Genome Project could be nothing but an expensive science project.

"Computers can solve things, but they can't predict anything new," said Dr. Alan Walton, chairman of Oxford Bioscience. "We're going to have to have a breakthrough here."

If researchers are going to extract drug discoveries out of GenBank the Human Genome Project's free database, and Celera the company that created its own genome map, without getting buried in information, biotech will need revolutionary, eureka-style discoveries.

Jeremy Levin, CEO of Physiome Sciences in Princeton, New Jersey, is shouting "eureka!" and some experts say he's worth a listen.

Physiome has developed technology called "In Silico Cell" that helps researchers create models of living systems to predict which drugs will work before they begin clinical trials.

Using an XML-based language called CellML, scientists can create a mathematical representation of any type of cell -- from heart, to lung, to kidney -- and perform simulations to test drugs.

"Here you've got a situation where I believe you can predict things the pharmaceutical industry would not know by any other method," Walton said.

The program can tell researchers that a certain heart drug, for example, will cause the left ventricle to expand.

If researchers can test drugs "in silico" and get meaningful results before they go into clinical trials, they could save millions of dollars and possibly save lives.

"For the first time in the history of biology, we have the capability to reduce biological functions to mathematics and compute them in a cost-efficient fashion," Levin said.

Money spent on research and development at pharmaceutical companies has increased at a staggering rate. In 1980, researchers spent $3 billion, according to the Pharmaceutical Research and Manufacturers of America. This year they spent nearly $23 billion. Even more surprising is the fact that only $2.4 billion of that was spent on drugs that made it successfully through Food and Drug Administration clinical trials.

"This is a major crisis," Levin said. "Preclinical expense -- unless pharmaceuticals improve all their systems -- is going to drown this industry."

Drug discovery is an imprecise process that relies heavily on serendipity. Pharmaceutical companies come up with about 140 new drug candidates per year, but only about seven are marketed. Big pharmaceuticals typically aim to develop two new drugs per year, and most have failed.

"The industry is built on failure," Levin said.

With $50 million in venture capital money in its coffers, Physiome hopes to garner revenue by helping pharmaceuticals know what will be a success before they even begin clinical trials, which can take more than a decade and about $800 million for just one drug.

Scientists tested the In Silico Cell on a heart drug originally developed by Bristol-Myers Squibb. The drug made it to Phase Three clinical trials in 1995 where, unexpectedly, it killed several women.

In 1999, Physiome used the drug to test In Silico Cell. Researchers mathematically coded biological models of men and women with different genetic backgrounds, and simulated a dosage of the drug.

"The results were striking," Levin said. "You could see that the female was three times more sensitive to this drug. It could have saved lives."

Physiome scientists also are using In Silico Cell to examine the FDA's entire database of would-be heart drugs that never made it through clinical trials.

By choosing trial subjects they know are more likely to respond to a drug, some of these failed drugs could actually get approval for specific populations.

"The FDA has unofficially said they love it," Levin said.

Of course other companies hope they'll be the ones with a revolutionary solution for the information bottleneck for pharmaceuticals.

Stephen Friend, CEO of Rosetta Inpharmatics says his company has built a system that allows researchers to use the entire genome of a human or other organism as a "sensor pad."

While Physiome's system is focused on modeling the biochemistry of cells, Rosetta's technology takes a snapshot of an entire genome and shows researchers what changes a drug makes.

The company's core product, the Resolver System, is a "decoder" that can find patterns in gene sequences and identifies protein function.

"Whether it's a cardiac or central nervous system cell, we figure out what the changes are in all of its proteins," Friend said.

Rosetta hopes to first supply tools to pharmaceutical companies and then to work with them as collaborators.

"We picture that within the next two to three years we'll primarily be working as collaborators with big pharmas doing drug discovery," Friend said.

Putting biology "in silico" could also help doctors know what drugs are best for individual patients after they're on the market.

Scientists recently used a computer model to understand how the HIV virus mutates to make itself resistant to the AIDS drug stavudine, which could help doctors tailor treatments more precisely. Scientists at Belgian company Virco designed a neural network to replicate the mutations.

Ultimately, these companies hope that the FDA will accept results produced by their technologies to provide safety data for the first phases of clinical trials. Although that may seem like a pipe dream now, experts say changes that big need to happen if the drug-discovery industry is to stay afloat.

"My dream is in ten to twenty years, rather than biotech companies having three hundred employees, a high burn rate, and the public getting turned off every two or three years," Walton said, "it will be a couple people and a great computer, and they'll go from zero to Phase Two (clinical trials) in a matter of hours rather than the time and cost now."

Kristen Philipkoski covers the science and ethics of genetic research and medical technology for Wired News. Before joining Wired News, Kristen wrote on a freelance basis for the print magazines Wired, Health, People, and Life.