Bioinformatics Backgrounder

One day, two patients with the same disease but different colored eyes may get different drugs. Children may be treated for Alzheimer's disease even though its symptoms are decades away. And drugs will be more effective, with fewer side-effects, because they will be tailor-made for patients.

This is the dream offered by genomics, currently one of the hottest areas of research and investment. Genomics is driven in part by the Human Genome Project, which has proposed to identify and map the location of every human gene by 2003. Now, lured by the opportunity to treat diseases at their genetic origins, pharmaceutical companies are snapping up genomic data and research programs. The feeding frenzy began in 1993, when SmithKline-Beecham handed $125 million to Human Genome Sciences (the commercial arm of the nonprofit The Institute for Genomic Research, or TIGR) for exclusive rights to its genomic information. The pharmaceutical industry has since invested over $1 billion in genomics companies, and most of the largest players, including Merck & Company, Inc., Glaxo-Wellcome plc, and Hoffman-La Roche, have launched their own genomics efforts. "All major pharmaceutical companies are getting into genomics," observes Chris Fields, vice president of scientific affairs and chief scientific officer at Molecular Informatics, Inc. (MII). "Their lives as competitors depend on it."

Why are pharmaceutical and some agrochemical companies "betting on the genome" (Science, February 7, 1997)? Because they believe genomics has the potential to unlock new "targets," the protein engines that run living organisms. Most drugs work by inhibiting or initiating the action of proteins. Genomics offers the chance to move beyond proteins to learn about the genes that produce them (see "What's in a Genome"). "Rather than fixing broken protein machinery, which has been the focus of 20th century medicine, we are starting to look for ways to treat diseases genetically," says Elbert Branscomb of the Lawrence Livermore National Laboratory.

But 21st century therapies will remain a dream if scientists are unable to interpret genomic data. It's not so much the size of the genome that stymies scientists -- at three billion nucleotide base pairs, the genome is equivalent in volume to the hard drive on a well-equipped PC. The difficulty comes in exploring what each portion of DNA does. Drug-hunters need to know how DNA differs among individuals. What does our DNA say about the way we work, the diseases we are susceptible to, and the drugs to which we will respond? To answer these questions, scientists need ways to easily collect sequence data, analyze it, and store it for future reference.

Enter bioinformatics, broadly defined as the role of computers in the capture, storage, and manipulation of biological information. Like genomics, bioinformatics is a nascent industry. Genomics researchers discovered early that they could do more faster by creating software to analyze and store genome data. Today, the Internet and other public-domain areas are awash with software and databases that assist with all aspects of gene research. The result has been likened to the Tower of Babel -- software has been written in every possible programming language, making it impossible to use analysis packages together and requiring biologists to master computer science as well as biology.

While many organizations are enthusiastic about bioinformatics, only a few are organizing themselves to define the next generation of bioinformatics systems. "There's a difference between a serious software company creating a product and a bunch of researchers sitting down to write some code," MII's Fields points out. In fact, the presence of dedicated bioinformatics organizations makes genomics a more viable field and has the potential to sweeten its appeal on Wall Street. As pointed out in Science recently, one of the selling points of genomics is that these companies have an immediate, valuable product: the genomic data they sell to the pharmaceutical and agrochemical giants. For these customers, success lies in how well, and how quickly, they are able to turn that data into the knowledge that supports their discovery programs. Sophisticated bioinformatics systems can help them make better decisions about how to take advantage of genomic data.

Companies like MDL Information Systems, Inc. and MII, along with the National Center for Supercomputing Applications (NCSA) at the University of Illinois in Urbana, are supporting modern discovery objectives with robust, enterprise-wide bioinformatics systems. These systems take advantage of advances in modern computing while preserving the grassroots, collaborative atmosphere associated with bioinformatics and genomic research. Consider the Bioinformatics Workbench, developed by the NCSA and marketed by MDL. The NCSA could have created a single software package to replace the reams of analysis packages available on the World Wide Web. But rather than reinventing the wheel, the NCSA built a system that links existing analysis software through an easy-to-use Web browser. The Bioinformatics Workbench allows an organization to bring all the genomic databases and analysis programs its scientists need onto its corporate intranet, where they can queried safely and securely using familiar software such as Netscape Navigator.

Scientists also need ways to manage genomic information, which is not only complex, but widely dispersed. DNA sequences can be catalogued in-house, acquired from collaborators or partners, or downloaded from public-domain databases available on the Internet. To maintain their competitive edge, organizations demand industrial-strength systems that can hold all necessary genomic data securely and make it available for future reference. The BioMerge System, developed by MII and marketed by MDL, brings together public, third-party, and proprietary genomic data in a single, fully functional relational database management system. Scientists are able to register new sequences, annotate existing ones, and search, browse, and report on genomic data.

The industrialization of bioinformatics offers the chance to make genomic information an integral part of the discovery process. Along with chemical synthesis and assay development, genome sequences are important tools for finding the next therapeutic agent, plant growth regulator, or chemical innovation. By integrating bioinformatics systems with existing systems for managing scientific information, scientists can more efficiently identify fruitful biological targets and promising lead compounds. For MDL, bioinformatics is a logical extension of its core business-supporting the discovery process. Through its agreements with MII and the NCSA and the innovation of its customers, MDL is laying the foundation for 21st century biological and chemical research.

<Sidebar>What's in a Genome?

Call it an instruction manual for life, a standard-issue document for every cell. The genome explains how to build and maintain an organism. And, as with other important messages, genomic instructions are encrypted -- into coded strands called DNA.

A DNA sequence is simply a string of letters, a list of As, Cs, Gs, and Ts representing DNA's four basic nucleotide building blocks. The human genome comprises about three billion of these letters, which are portioned out in 24 packages (chromosomes) in each human cell. Certain combinations of letters form "words," or genes, which explain how to build the proteins that initiate and sustain life.

The trouble is, the DNA recording medium is as old as life itself. A few of the instructions applicable to bacteria, amazingly, still apply to humans. The rest are obsolete or have been updated. And yet, these ancestral codes have not been dropped from the sequence. Only 3% of all human DNA codes for proteins; the rest is currently considered gibberish.

The first challenge for scientists is to separate meaningful genes from junk DNA. Then comes the more daunting task-determining the message encoded in each gene. Subtle differences in the encoded message are responsible for the many differences between individual organisms, including an organism's susceptibility to disease and responsiveness to drug therapies. Bioinformatics assists with both tasks, providing software tools to help investigate DNA sequences and discover the hidden message carried by our DNA.

Suggested Reading

Science recently reviewed the genomics revolution in a a special biotechnology section titled "Betting on the Genome" (275, February, 1997, pp. 767-782). The New York Times also published an overview of pathogenic genomics on February 3, 1997. Another general write-up appeared in the May 8, 1995 issue of Business Week ("The Gene Kings," pp. 73-78). A primer on genetics, molecular biology, and bioinformatics is available on the Johns Hopkins University Web site. For more technical perspectives on genomics and bioinformatics, look for recent issues of Genetic Engineering News or Nature Biotechnology, both of which frequently publish articles and news stories covering these areas.