Most neurological disorders affect multiple aspects of behavior, indicating they are likely to be brain-wide phenomena. Scientists from Rice University, Baylor College of Medicine, the University of Rochester, and the University of Minnesota are embarking on new research that will lead to a better understanding of information processing in the brain, and how failures in networks across the brain may result in disease.
Dr. Xaq Pitkow, assistant professor of electrical and computer engineering, Rice University and assistant professor of computational neuroscience, Baylor College of Medicine; Primary Investigator Dr. Dora Angelaki, Wilhelmina Robertson Professor & Chair, Department of Psychology, Baylor College of Medicine, and professor, electrical and computer engineering, Rice University; and Dr. Gregory DeAngelis, chair and professor, brain and cognitive sciences, and professor, biomedical engineering, neurobiology and anatomy, Center for Navigation and Communication Sciences, and Center for Visual Science at the University of Rochester, have received a three-year, $600,000 grant from the Simons Foundation. In addition, a two-year, $300,000 NSF BRAIN EAGER award will also support a related project and will add Paul R. Schrater, associate professor of psychology and computer science at the University of Minnesota.
“Neuroscience traditionally measures how the brain solves simple, repetitive tasks, which is useful and illuminating, but it misses some of the more interesting and important computations that the brain performs,” Pitkow explained. “In this new project, we will substantially increase the complexity and naturalism of the experimental task, and hope to make it engaging for the animal subjects. We will be training subjects to perform a task, and then we will record activity in multiple brain regions to see how the animal chooses its actions,” Pitkow said.
Increasing the complexity of the task means the team also has to increase the sophistication of its analyses. Their approach is a novel one – developing a mathematical model of the subject’s behavior to use as a guide for what they expect its neurons to be doing. The math will tell the team what the subject is thinking about things it cannot directly observe. The team’s mathematical model aims to decode the subject’s internal model of the world, the model that decisions are based on. The interactions between the mathematical variables give predictions about the interactions between corresponding neural representations, and those predictions can then be directly tested.
“By directly addressing the question of how internal models are implemented in the brain, this research should bring us a step closer to understanding the brain’s fluid and adaptive behavior in a changing world, which I view as one of the great mysteries of cognition,” Pitkow stated.
The interdisciplinary, collaborative nature of the work is essential to the project’s success. The Rice-Baylor partnership is the project’s cornerstone; the research team will benefit from their experimental and quantitative expertise, and will build upon that with the knowledge in behavioral neuroscience and behavioral modeling that experts from the University of Rochester and University of Minnesota will contribute. The team’s young scientists will gain interdisciplinary training and experience they need to understand how the brain works and apply it to their future research.
“The work could not be done by any single one of our groups. There are big experimental and theoretical parts, and the teams really have to be able to talk a shared language,” Pitkow said. “These days it is possible to work closely with people at great distances, and we regularly enjoy having lab meetings with participants from across the globe.”