From Algorithms to Cells: The Neurophysiology of Emotional States

Objectif

Emotional states such as aggression and fear are vital for survival and mental health, yet we still lack clear, causal rules for how the brain builds, sustains, and reads them out. I propose that emotional states are generated by algorithms represented as stable patterns of population activity in specific brain circuits. Inputs, neuromodulators, and body signals tune the strength and duration of the emotional state, and downstream pathways read these population patterns to guide actions.
I will test this hypothesis using a deep-brain two-photon microscope that can both record and control the activity of specific neurons in awake mice, guided by testable models that predict how the population pattern evolves over time. With this approach, I will causally connect the emotion-state algorithms inferred from neural activity to the underlying circuits that generate them.
Building on my recent findings, expertise, and preliminary data, I propose an ambitious yet feasible plan to (i) define how inputs, neuromodulators, and body states set the intensity and persistence of emotional states by measuring and controlling the stability of population-activity patterns and behaviors; (ii) identify and compare organizing rules across different emotional states to distinguish shared and state-specific algorithmic and circuit architectures constructing them; and (iii) determine how downstream pathways interpret low-dimensional population dynamics to produce scalable, persistent behavioral outputs.
By using model-guided causal experiments, integrating theoretical modeling with circuit dissection, these efforts will establish core principles and mechanisms for examining how emotional states are implemented in the brain. This will open the path to facilitate comparative research across circuits and species, and lay the groundwork for closed-loop control of dysregulated affect.