Faculty Research Update - Physicist as Brain Scientist

Wed, 2009-12-16 16:40 -- Anonymous (not verified)

 Duke Professor Henry Greenside is a theoretical physicist who studies how the brain works. He says, “The brain is so complicated we need to approach it from various angles. The thing a physicist brings to bear is the belief that if you push hard enough there will be some understanding. Physicists use theoretical and mathematical tools and we’ve repeatedly found concise mathematical descriptions of very complicated things.” Greenside spent the fall term on sabbatical at Janelia Farm, the research campus of Howard Hughes Medical Institute in Virginia. There, he worked with theoretical physicist Dmitri Chklovskii, who is mapping out the connections among neurons in a fly brain using mathematical models, computers and extremely high-resolution images of slices though fly brains. Greenside eventually intends to use that wiring diagram, which is not yet complete, to study how information is processed in various regions of the brain. In the meantime, he spent the fall working on the neuron level rather than the circuit level. “I’ve been testing a theory of how the shape of a neuron plays a role in the kind of information it processes,” he says. The theory, developed by Chklovskii and others, involves balancing the need for a neuron to minimize energy consumption (which requires being small) with the need to maximize connections to other neurons (which requires being big). Greenside has been using a computer program he wrote to analyze the diversity of shapes of more than 2,000 mouse neurons that come from different brain regions and from mice of different ages. It’s sort of a warm-up exercise to what he will be doing with Chklovskii’s fly brain database, although in that case he’ll be looking for how information flows through many neurons rather than at the function and shape of individual neurons. The fly brain contains about 100,000 neurons, making the wiring diagram incredibly complicated. “Even just to look at the data you need a computer to hide information that might not be interesting,” he says. “We’ll be looking for what details we can ignore. One of the first steps is to use the wiring diagram to find out which neurons are not connected.” (Just imagine trying to unravel the workings of the human brain, which has about 100 trillion connections among 100 billion neurons!) Greenside hopes his work will help bridge the gap between experiment and theory. “Brain science right now is very much like alchemy—you can observe all kinds of effects but there is virtually no infrastructure to interpret the results,” he says. “Neurobiologists and others do great experiments in the brain but often have little idea why the experiments produce the results they do. Theorists speculate why things are happening but they can’t prove it. We’re mounting a systematic quantitative effort to get data of a novel kind that will let us test whether an idea is right or wrong.”