Skip to content


Our goal is to uncover the neural circuit mechanisms giving rise to the symptoms of neurological and psychiatric disease. To do so, we take direct inspiration from clinical observations to form hypotheses about the neural cell types and circuits involved in producing cognitive and behavioral symptoms. We then train mice to play games that require them to think and move in ways that patients struggle. Then, to understand how neural circuits are involved in these cognitive processes, we observe and manipulate genetically-defined neurons as mice play these games. We simulate the perturbations of these circuits in the setting of disease using a toolkit of multisite electrophysiology, fiber photometry and optogenetics. We then apply advanced statistical methods, machine learning and probabilistic models to understand how these neural circuits affect the mouse’s behavior and relate these findings to what we see in the clinic.

Deciding when we can’t know the right answer

How do we choose when we can’t know the right answer or the right answer might not even exist? The philosopher al¬Ghazali reasoned free will lies in our ability to act under uncertainty, in situations where we can only pick “randomly” at best. What neural mechanisms enable us to behave stochastically? To answer these questions, we are elucidating the neural circuits involved in stochastic behavior by examining their behavior as mice play games that force them to make decisions under uncertainty.

The ability to move when we want

Patients with Parkinson’s disease classically feel “slowed down” and experience difficulty initiating movements. But surprisingly, it’s not all movements that are affected. Patients particularly struggle to initiate movements “at will,” but they can move in reaction to events in their environments with relative ease. How does the brain decide when to move? We are chasing this question down with a combination of optogenetics and multi-site electrophysiology and photometry experiments in mice trained to play a self-timed movement game.

Recovery from neurotrauma

What determines who recovers from neurotrauma and who does not? We are leveraging deep learning and bioinformatics to identify the genetic and biological signaling pathway determinants of resilience to neurotrauma with the goal of developing targeted therapies to improve recovery after traumatic brain injury.

Linking human and animal cognition

We are studying the similarities and differences between humans and mice as they learn to make stochastic decisions under uncertainty. By studying the behavior of both species as they play the same games, we can leverage the unique advantages of each to translate findings obtained with the circuit dissection tools available in mice to better understand the human brain in the settings of health and disease.

A striatal neuron recorded during the self-timed movement task. This quiet little cell ramps up slowly and starts firing like wild right before the “self-willed” movement occurs! We train mice to play games and record and manipulate their neural circuits to understand how they make decisions.