Driven By Data

Slice of life: laser-scanned image of Arabidopsis root.

Philip Benfey's research using mutant strains of the plant Arabidopsis thaliana holds great promise for discovering how living tissue develops from a single cell. Its reliance on advanced research techniques and sophisticated instruments provides a good example of the ways in which biology has increasingly become a "data-driven science."

A recent experiment brings Nori Matsumoto, a postdoctoral fellow working with Benfey, into a small, darkened room in the lab complex. In one hand, he holds a plastic tray filled with gray gelatinous growth medium supporting rows of tiny, sprouting Arabidopsis plants no larger than a comma. With a sharp scalpel, a steady hand, and a practiced eye, he slices away a sprouting root, deftly transferring it onto a microscope slide. He places the slide beneath the stage of a laser-scanning microscope--a machine that fills the room and that is "more expensive than my car," he jokes.

After a few clicks of a computer mouse button, the image of an Arabidopsis root tip appears on the microscope's video display. It's a radiant scarlet structure, like the fiery nose cone of some alien spaceship entering the atmosphere. Matsumoto is not especially interested in the root tip itself, though, but in the swath of fluorescent green speckles that lines its inner cell rows.

This green fluorescence signals the activation of a marker he has attached to a gene segment called a "promoter," which activates one of the plant's development genes. Successfully manipulating such promoter segments is key to tracing the control machinery of the gene--just as a homeowner flips the breaker switches on an electrical junction box one by one to discover which house lights they control. Satisfied that the image is a good one, with a few more clicks of the mouse, Matsumoto saves the image to a computer elsewhere in the building.

According to Benfey, machinery such as laser microscopes and high-powered computers represent the future of biology. They have, in fact, changed the very face of biology to more resemble its formerly distant cousin, physics. "The pervasive use of computational approaches for addressing cell- and molecular-biology problems is something new," he says. "It's certainly very different from when I was a graduate student, where we used computers to make figures, or do word processing, or look up the occasional number in a data base. I have members of my lab who spend more than 50 percent of their time on computers now, doing data analysis and computer modeling."