from behavior to theory to neural circuits
We aim to understand how context shapes perception, cognition and learning. To do this we start with behavioral experiments to isolate key variables. We then build conceptual and theoretical models to determine the algorithms the brain might use to execute these behaviors. Finally, we aim to understand how the brain implements these algorithms by monitoring the neural dynamics during behavior and dissecting the underlying circuitry.
The long-term goal is not only to understand normal brain function but to use these insights to make headway into treating intractable neurological disorders.
A still-frame set of images showing a mouse executing a licking response to a sensory cue. The ability to monitor detailed motor responses, including response latencies, coupled with the ability to control task parameters allows us to probe behavior with high precision.
behavior
A major goal of modern neuroscience is to understand how humans and other animals perceive, think, interact and engage in a complex world. We aim to link behaviors to the underlying neural implementations but the work must start with behavior itself. A core strength of our department is that we put behavior first across model systems, from humans to bats to owls to rodents. This allows faculty, students and postdocs in our department to draw connections in ways that are truly unique.
theory
Just as in the physical sciences, neuroscience benefits from a strong conceptual and theoretical framework. To that end, we build models within the lab to understand behavior and network dynamics. Moreover, we have forged active collaborations with theorists including Adam Charles, Josh Vogelstein, Andrew Saxe, and Mate Lengyel. We tend to use theory in two ways: 1) to make testable predictions of behavioral outputs and network dynamics, and 2) to fill gaps in our experimental arsenal and make mechanistic insights when the tools are not yet available.
neural dynamics and circuitry
We now have the tools to explore neural circuits with unprecedented spatiotemporal resolution. In the lab, we observe neural dynamics with two-photon calcium imaging, monitor synaptic inputs with whole-cell recordings, and manipulate neural circuits with opto- and chemo-genetics. All of these are done in the intact brain during behavior. Our philosophy is observation-first followed by targeted manipulations. We benefit from interactions with neuroscience labs across Hopkins who use complementary tools and methodologies.
We regularly use multi-color, two-photon imaging to observe the activity of multiple circuit elements simultaneously. Above is an image of cholinergic axons projecting from the basal forebrain (green, axon-GCaMP6s) and auditory cortical neurons (red, RGECO1a).
We exploit two-photon imaging to allow us to ‘track’ the same neurons throughout learning. Here, is an example of excitatory neurons (CaMK2-GCaMP6f) that we tracked over thousands of trials across weeks in Layer 2/3 of auditory cortex . This allows us to observe changes in neural dynamics across trials and across days, revealing rapid emergence (within just tens of trials) of a robust reward prediction signal in a sub-population of AC neurons.