Cortical activity underpins perception. This activity is often sparse, with a minority of neurons responding to specific stimuli. How do sparse groups of neurons influence perception and behavior? How does cortical computation impact perception? We use the mouse whisker system to tackle these questions.

Large-scale recording of cortical activity in animals performing controlled behaviors allows us to identify populations that may underpin perception. Such correlative experiments are an essential prerequisite to experiments that probe circuit computation and the perceptual role of specific populations.

Recurrence is a defining feature of cortical circuitry, and is believed to underpin many computations. We use perturbation in conjunction with network models to probe how recurrence among specific populations influences cortical computation.

Identifying the cortical populations that correlate with perception is not enough to establish causality. We use focal perturbation to understand how specific populations of neurons contribute to behavior.

Combining high speed videography with modern image processing techniques allows for precise measurement of the behavioral state of the animal. Tasks are designed so that even subtle changes in the animal's behavior can be detected.

Two-photon microscopy using modern calcium indicators allows us to record the activity of thousands of cortical neurons during behavior. Neurons can be tracked over months, allowing us to study how cortical dynamics evolve over time.

Our experiments depend on rapid turnaround of terabyte-scale datasets. We are developing pipelines to process data acquired during calcium imaging and behavioral videography, and to relate neural activity to behavior.

Neurons encoding particular features are often intermingled with other neuron types. Approaches like multiphoton ablation, microlesions, and two-photon optogenetics allow us to test the hypotheses originating from our large scale mapping experiments.