Life's Secrets on an Excel Spreadsheet

More than 3,000 delegates and 100 exhibitors attended the inaugural BioIT World Conference and Exhibition, held March 12-14 at the Boston World Trade Center. BIO-IT World editors present some of the show's highlights.

Life's Secrets on an Excel Spreadsheet

By Kevin Davies

Fortunes told: "Pharma should develop the drugs before you know you need them," said Eric Lander about the power of gene expression tools.

"There is nothing boring in the human genome," said Eric Lander, something of an expert on these matters. In his inaugural keynote address at the BioIT World Conference and Expo, the founder and director of the Whitehead Institute Center for Genome Research acted as the tour guide of "the most remarkable library on this planet" — the human genome.

A recurring theme was the critical importance of computational methods to unlock the secrets of the sequence. "The promise of genomics depends on bio-IT," said Lander. "This depends tremendously on information sciences marrying with life sciences."

As Lander and his consortium colleagues reported in a landmark article in Nature last year, the total number of human genes, about 32,000, surprised everyone. (Lander joked he would be writing letters to all his former MIT freshman biology students apologizing for the years he told them the total was 100,000.) As of now, about 70 percent of the human genome exists in finished form, with the consortium aiming to complete the sequence (give or take 1 percent) in April 2003 — the 50th anniversary of the double helix. With a few simple keystrokes, anyone can access the DNA sequence via the Internet. As memories recede of the dark ages when researchers toiled for years in search of a single gene, Lander suspected it would not be long before his freshman class "will assume the sequence was always available."

Lander devoted most of his address to the comparison and extraction of genome information across species, individuals and tissues. Sequence comparisons between mouse and human are facilitating the identification of human genes, as well as revealing a surprising number of conserved sequences with no known function. Lander called it the "genome thumbing its nose at us," but "because of bio-IT, we know it is there." By systematically searching for conserved gene regulatory sequences in yeast, Lander's team is now "reading out regulation from [sequence] information."

The very slight genetic variation between humans — about one common variant every 1,300 bases — is a testament to the youth of the human race, which is only about 3,000 generations old. By comparison, chimpanzees and orangutans have two and 10 times as much genetic variation, respectively. The punch line proved irresistible: "We think chimpanzees all look alike — to chimps, we all look alike."

A handful of these specific DNA variants — single nucleotide polymorphisms, or SNPs — have been associated with susceptibility to common diseases, including Alzheimer's, heart disease, and diabetes. The present tally of more than 2 million SNPs lays the foundation for direct testing of disease associations, an era when "human genetics will be reduced to a very big Excel spreadsheet." Lander expressed optimism that an international consortium would soon begin work on a so-called haplotype map of the human genome, cataloguing patterns of SNPs in different populations to aid the search for common disease genes.

Aside from annotating the genome, much of Lander's current research focuses on the molecular profiling of cancers using DNA microarrays. Such studies are producing breakthroughs in cancer diagnosis, in some cases redefining disease and treatment regimens. By comparing gene expression patterns, new computer programs can accurately categorize some cancers with virtually 100 percent success. Lander compared the pioneering studies of Sidney Farber, spanning four decades on classifying and treating the leukemias, to the present: "Forty years of research condensed into the time it takes to press 'return.'"

One of the most striking examples is a form of acute lymphoblastic leukemia called MLL, which Lander, together with Todd Golub, Stanley Korsmeyer and colleagues, has shown is a distinct entity at the molecular level. In the wake of last year's highly publicized success of Gleevec, the Novartis drug used to treat chronic myelogenous leukemia, it turns out that Novartis has a related inhibitor that targets the best MLL marker. A similar scenario is playing out with lymphoma and an existing Eli Lilly drug. Lander's prescription: "Pharma should develop the drugs before you know you need them."

Lander concluded that "Genomic information is now driving a revolution in biomedicine," while noting the serious shortage of bioinformaticians, scientists who can read and interpret the sequence. He estimated that the field might need 50 times as many people as are currently qualified. "We're just kindergartners reading this information," he said — it is up to the next generation to extract the secrets of the sequence.

The Pharma Prophecies


Click here to login and leave a comment.  


Add Comment

Text Only 2000 character limit

Page 1 of 1