Research Interests
The Problem
One of the key features of brain function is to extract patterns from noisy sensory input, update expectations based on this experience, and then use that knowledge to guide behavior adaptively. We call these processes Perception, Learning, and Memory. In the Franks Lab, we study how these processes are implemented in the brain—and what happens when they break down in disease. Using the mouse olfactory system, we investigate how neural circuits transform raw sensory input into stable odor representations, and how these representations are modified by experience, stored over time, and used to support adaptive behavior.
In fact, deficits in odor identification and recognition are among the earliest and strongest predictors of neurodegenerative diseases like Alzheimer’s and Parkinson’s. By uncovering the basic principles of how the brain learns, stores, and retrieves sensory information, our work provides a foundation for understanding what goes wrong in disease and how we might intervene.
Our Approach
We integrate large-scale electrophysiology, optogenetics, in vitro slice physiology, behavior, and computational modeling to causally link behavior to circuit dynamics to synaptic mechanisms. The olfactory system offers a highly tractable circuit for this approach because we can directly and precisely control the sensory input to cortical circuits. This allows us to characterize how information is transformed within the piriform cortex, and to uncover the neural mechanisms supporting these transformations by manipulating and measuring activity in defined neuron populations.
Our goals are to elucidate the piriform cortex’s role in odor-guided behavior and to identify fundamental principles of cortical computation that apply across sensory modalities and behavioral states—including those disrupted in neurological and neurodegenerative disease.
Innovations and Discoveries
We have extensively characterized the piriform cortex's excitatory and inhibitory synaptic organization, contributing to a quantitative wiring diagram of this evolutionarily conserved circuit. By combining in vivo recordings, optogenetics, and modeling, we revealed how input from the olfactory bulb is integrated and shaped by local inhibition and recurrent connectivity, providing a mechanistic foundation for how sensory representations are formed and refined in the cortex.
We discovered that distinct features of the population response encode odor identity and odor intensity. While odor identity is determined by which neurons are activated, intensity is encoded in the temporal synchrony of the response. Importantly, we showed that piriform cortex circuitry transforms these features into concentration-invariant representations, ensuring that the identity of an odor remains stable despite changes in stimulus strength.
Most recently, we’ve shown that the piriform cortex implements a phase-to-rate transformation: converting temporal input from the olfactory bulb, which is aligned to the respiration cycle, into reliable ensemble codes. This transformation supports rapid and robust odor recognition and may reflect a more general strategy by which cortical circuits stabilize temporally structured input across sensory systems.