In 1997, the Internet domain name business.com was sold for $150,000 . . . in 1997, the Human Genome Project was less than halfway to completion . . . in 1997, the genome of E. coli was sequenced . . . in 1997, Dolly the sheep was born . . . in 1997, HMS Beagle, went live . . .

Scientific discoveries come thick and fast these days. Fifty years ago, a single postgraduate thesis might have been dedicated to fully elucidating a single molecular structure, for instance. Today, while not a trivial task, technology readily allows a crystallographer to pack in at least a handful of structures in a single year. Even protein structures are revealed at a rate of two or so per Ph.D.

"These days very large structures can be determined, a task that was almost unimaginable half a decade ago," says molecular biologist Nenad Ban of the Swiss Federal Institute of Technology at Zurich.

If scientific practice is changing, so too is the way scientists operate. Robert Ubell, president of BioMedNet USA in 1997, described in an early issue of HMS Beagle how he periodically liked to ask groups of scientists about their library habits. At the time, he found just one scientist in a discussion group of biologists in New York who confessed to a visit to a library in the preceding week; some hadn't even been that year. Even then it was more likely you would see a scientist in a discotheque than a bibliotheque. Well, maybe not. It seems the situation has now swung still further away from the library turnstile despite the attractions of hardwired Internet cubicles and caffe latte.

According to bioinformatician Matteo di Tommaso, the general manager of Wisconsin-based Genetics Computer Group, the Internet has changed everything "by shrinking the world so that ideas and information can be easily and quickly exchanged." Allan Jordan, a medicinal chemist at the drug discovery company RiboTargets, based in Cambridge, England, certainly sees a persistent turning away from the traditional library visit. "I get title pages in my mailbox on a daily basis, access to the latest articles before they go to print, back issues searchable in numerous ways, and multijournal searches with articles relevant to my work and interests flagged up and mailed to me directly," he explains. "Why would anyone want to spend hours scouring dusty shelves for months-old papers?"

It is only ten years or so since Tim Berners-Lee's World Wide Web hit the streets and less time than that since the emergence of the first graphical Web browser. Yet you are reading a magazine that allows you to search millions of words in seconds for topics and people of interest. HMS Beagle's parent site BioMedNet can connect you directly to other online periodicals and databases with a few clicks of a plastic mouse. You can then watch other researchers develop their ideas through online forums and pick up and comment on their research "publications" before they even hit a printing press.

Think about it. Gutenberg's work brought words to the masses, but it took centuries before we threw out the trays of metal letters. Essentially within the last four years, we have witnessed the virulent spread of an information tool. It allows almost anyone, anywhere in the world, from a cybercafe in the Australian outback to a delegate en train to a conference in Paris, to access instantaneously (bandwidth permitting) the very details that reside in every cell of our bodies and help make us what we are.

There are still storage issues to be addressed, and there is still a need for journal back catalogs to go further back in time, prior to about 1997. A scientist's need to visit the library will decrease drastically, saving time and to a degree reducing overhead.

Matt Tudor is another of the thrusting new breed of scientists looking to the net for inspiration. He began his biology Ph.D. at the Massachusetts Institute of Technology at Cambridge around the time of Beagle's launch and is working on the modulation of gene transcription by DNA methylation. He touts his biological expertise on the Askme.com discussion board, although he is not yet convinced of the value of such schemes. "The chat boards like AskMe, Biosupplynet, Biowire.com, and DoubleTwist are not really useful resources for practicing scientists," he says, "but the multi-hundred-thousand-dollar-a-year subscription services are outside the reach of even the moneyed labs I work with."

Tudor recognizes, even in his short scientific career, some major changes in the last four years in information technology, the primary development being "progress in, not completion of, the sequencing of the human and other genomes." He points out that traditional questions can now be answered more completely, and new questions can be asked, for example, about the identification and delimitation of previously unknown genes, the identification of control elements, and the ability to theorize about the molecular evolutionary processes that brought us here.

According to Ormond MacDougald of the University of Michigan at Ann Arbor, "The way my lab does business now is very different from four years ago." He points out that the development of analysis by microarray gene chips and the widespread availability of knockout and/or transgenic animals have altered entirely the way he does research. "Knockouts," he says, "even alter how we do our in vitro analyses because mouse embryonic fibroblasts (MEFS) allow us to do more mechanistic work in vitro without messing around with antisense or other methods of killing a gene's expression." The microarray work is opening the team's eyes to how activation of signaling pathways and/or transcription factors coordinate regulation of gene expression during cell development.

The microarray is certainly a major technological development. "It was," says Tudor, "a trivial matter to predict that microarrays would impact research like mine, which deals with gene transcription. But microarrays are also being used for genotyping in disease gene hunts, for functional characterization, for diagnosis, and increasingly for protein applications."

Microarrays, popularly known as biochips or DNA chips, consist of an array of biomolecules on either a glass slide or a silicon chip, which allows one to quantitate ten thousand or more genes in a hundred-milligram sample in a few days. Previously, this took several years and kilograms of material. Such technology facilitates gene expression profiles of different tissues and is leading to a better understanding of biological processes. Ultimately, it will allow better drug targets with controlled side effects to be developed and tailored treatments to be prescribed.

Postdoc Bryan Crawford, who works on developmental biology at the University of Washington at Seattle, has also seen the benefits of the information revolution over the last few years. "There's definitely been a quantitative if not qualitative change in the ease with which I can find information," he says. He points out that "relevant references are frequently online within days of their submission, and sequence data in the various databases is easily searchable and much more easily interpreted than a few years ago." He feels that the human genome is not as significant to his line of research at this stage, but that any completed genome will be interesting. "I'm really looking forward to the day when we have completed genomes for representatives of all phyla, and maybe even most of the major classes," he explains. "That will be a database worth searching!"

"Just as computing and biology merge into bioinformatics, structural biology transcends several fields, e.g., computing, biology, NMR, crystallography, and chemistry," adds Jordan. "Furthermore, it's structural biology that will bridge the gap between genome and medicine," he claims, "providing structural data for pharma companies to use in the pursuit of new drugs, although this is not as trivial as most commentators seem to imply!"

The level of automation has also rapidly increased in chemical synthesis, analysis, and biological screening. It is making scientists' lives easier but at the same time increasing productivity. Recently merged pharmaceutical giant GlaxoSmithkline, for instance, has the technology at its center in Harlow, England, to design a chemical series, plan the synthesis, buy the materials, do a quality control assessment, weigh out the reagents, do the reaction, purify the products, analyze them, and test them for biological activity - all automatically.

Such a system simply did not exist four years ago. What will chemists and biologists actually do when such systems are widespread?

Functional genomics, proteomics, pharmacogenomics, combinatorial chemistry, and high-throughput screening have all fallen to the Web, with distributed databases and Web applets providing researchers with access to the attendant tools from their desktops, laptops, or, in many cases, Palms. These advances are pushing forward the generally maturing biotechnology industry - despite financial turmoil on NASDAQ. Where the process of drug discovery was once a singular art, the computer-assisted conversion of genomic information to proteome and metabolome brings mass to the scientific masses.

The new technologies are already forcing researchers to rethink the direction of their projects. After all, once you have the wheel in hand, it is not such a long journey to a transport system. With new systems comes the need for new generalists in such diverse and relatively young fields as chemo- and bioinformatics, combinatorial chemistry, genomics, and all the other "omics." Fields that were tiny niches within the broader church just half a decade ago are now the mainstay of countless research projects.

Areas such as bioinformatics are not simply computer support for the particular branch of molecular biology that is looking for drug candidates. It works down the discovery pipeline in a way that did not exist four years ago. It can validate drug targets, unravel mechanisms and metabolic routes, and even offer a toxicity profile. As such, some observers are describing it as the hot field for the post-genomic scientist. Bioinformatics allows teams to integrate vast quantities of data from countless sources. They - the data and the teams - can be scattered around the world. With the Internet, it is irrelevant where you or your data are.

If bioinformatics hints at a fusion of two disparate sectors of science - biology and computing - then it is the archetype of the new interdisciplinarity that is knocking down walls and opening up borders between scientists. Where many of our scientific ancestors usually saw themselves as polymaths - think Faraday, da Vinci, et al. - convenient boundaries meant grant-awarding bodies knew exactly what they were paying for, once science was running at the hot pace of the twentieth century. But net savvy, computer literacy, chemical know-how, and some vital biology now mean that specialists who keep to themselves are not so much a dying breed as a dwindling group of outsiders. Multidisciplinarity is the cross talk of science.

"Certainly, science has become more integrated. During my Ph.D., I seemed to be among a small number of scientists in my willingness to bridge chemistry and biology and actually learn some new skills by working with biologists and actively interacting. Now this is not only accepted more, but in most cases it is expected," says Jordan.

There is one aspect that sets the modern scientist apart from the old image of a solitary figure sweating over a Bunsen burner: with interdisciplinarity comes an increasing need for team workers. There are too many threads to even the simplest of genomics projects. Just look at the author lists on the Science and Nature papers announcing the sequencing of the human genome!

Yet with the emergence of increasingly powerful computer technologies, we are truly only at the beginning of developments; their greatest impact on scientists is yet to be seen. Many buzzwords have been coined in the last four years, but one word that persists is "adaptation." To survive and thrive, science and scientists must continue to adapt.