Damon A. Clark

Scholar: 2014

Awarded Institution
Assistant Professor
Yale University
Department of Molecular, Cellular, and Developmental Biology


Research Interests

How do Individual Neurons Contribute to Circuit Computations?


The brain is fundamentally a computational organ, performing complicated operations that combine sensory inputs and internal states in order to produce appropriate behaviors. My lab aims to understand how networks of neurons — their connections and their individual properties — perform the computations that transform sensory inputs into behaviors. We focus on motion perception in fruit flies, which have powerful genetic tools that allow us to manipulate their small visual circuits and then connect those manipulations to changes in circuit function. For any neural circuit to perceive motion, it must perform specific mathematical operations, like multiplication and temporal filtering. Which mathematical operations are being performed, exactly, and what purpose do they serve? How do neurons actually implement these mathematical operations? We use behavioral and neural measurements, sophisticated visual stimuli, genetic tools, and mathematical models to answer these questions. At the algorithmic level, we investigate the set of mathematical operations that neural circuits use to extract information from visual environments. By defining the algorithms in a circuit, we can begin to assign computational roles to neurons within it. At the implementation level, we ask how neurons use their connections and biophysical properties to implement specific mathematical operations. By manipulating individual neurons and their properties, we can discover how single neuron types contribute to circuit computations. Since the mathematical operations we study appear in many neural circuits and models of neural computation, a detailed understanding of this small circuit will apply to many other systems. Comparisons between how Drosophila and vertebrates each detect motion will illuminate potential algorithms for estimating motion and the diverse ways these can be implemented.