Brenner discusses turning data into knowledge
Posted May 1, 2002; 02:33 p.m.
The next great task of biology will be to convert its quickly accumulating supply of data into a broad understanding of the way organisms work, biologist Sydney Brenner told an audience at Princeton Tuesday.
Brenner opened a three-day series of lectures by laying out a vision of biology as a highly mechanistic discipline that, despite its complexity, will ultimately answer questions with the same authority with which physics predicts the behavior of non-living systems.
"Science is not just a recorder of everything that is," said Brenner, who has pioneered many of the great advances of biology over the last half century. "It actually claims to predict the future."
Brenner, the founder and director of the Molecular Sciences Institute in Berkeley, Calif., is presenting three lectures that address the theme of "Biology after the Genome Project." The talks continue Wednesday and Thursday with lectures titled "Does E. Coli Understand Itself?" and "The Architecture of Biological Complexity." They will begin at 8 p.m. in McCosh 10. The talks are free and open to the public.
In his opening talk, "From Data to Knowledge," Brenner argued that the vast amount of information collected from the human genome and other projects does not amount to an understanding of biological organisms. The true test of understanding, said Brenner, is whether scientists can develop theories that they can use to predict the outcome of experiments that they have never performed.
He said biologists should be able to ask such questions as, "If I made these genetic changes, what would this organism look like?"
A general examination question for a graduate student in 2070, he said, "might be 'Describe the genetic program to make a centaur, or prove the impossibility of a genetic program that would create a centaur,'" Brenner said. "There are a lot of problems, when you think about it," he quipped. "They have two torsos ... the colon of the man enters the esophagus of the horse." We are only now beginning to develop the understanding we need to say why such a configuration can or cannot work, he said.
One of the chief challenges in developing such a deep understanding, Brenner said, is to distinguish pertinent data from the "noise" of biological systems. While some genes will prove crucial for causing or preventing a given biological process, others will turn out not to matter; it will make no difference if they are "on" or "off." The presence of such "don't care" genes is one reason why the genome data alone can never lead to complete understanding and why biologists must develop overarching theories of how organisms work, he said. These theories, he predicted, will grow from the mathematics of computing. Biological organisms should be thought of as being "computed" from their underlying genetic information, he said.
Brenner said he plans to discuss examples of such genetic programming in the remaining lectures.
Brenner's talks are designated as Louis Clark Vanuxem Lectures and are part of the University's Public Lectures Series.
Contact: Marilyn Marks (609) 258-3601