Goal learning, memory, and drift in the Drosophila head direction system
Flores-Valle, A.; Honnef, R.; Seelig, J. D.
Show abstract
Selecting and memorizing goal direction are essential for navigation behavior. Heading information is represented in the head direction systems across species, including Drosophila. However, how navigation decisions are made and how goal memories are represented in these systems is little understood. Here, using a navigation learning assay for flies walking in virtual reality during two-photon imaging, we describe neural dynamics for direction selection and memory. We find that neurons which encode walking direction in the fan-shaped body, a navigation and learning related area in the center of the fly brain, show continuing autonomous activity or directional drift when the animal is at rest. Drift during rest centers around opposite directions to activity during walking, suggesting different computations between these two behavioral states. Targeted optogenetic activation of these neurons during rest is sufficient to induce a subsequent directional navigation preference. Learning leads to changes in drift distributions during rest depending on goal direction, revealing a memory in the network. The flys head direction system thus offers a compact architecture for direction selection, learning, and memory. Changes in neural representations due to goal learning and between rest and walking suggest similarities in navigation circuits across species.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Walking strides direct rapid and flexible recruitment of visual circuits for course control in <Drosophila> 98%
- Voltage imaging reveals circuit computations in the raphe underlying serotonin-mediated motor vigor learning 98%
- A cortical circuit mechanism for coding and updating task structural knowledge in inference-based decision-making 97%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.