Multiple tracking and machine learning reveal dopamine modulation for area-restricted foraging behaviors via velocity change in Caenorhabditis elegans.

Affiliation

Department of Bioscience and Informatics, Faculty of Science and Technology, Keio University, Yokohama 223-8522, Japan; Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung City, 80708, Taiwan; Waseda Research Institute for Science and Engineering, Waseda University, 2-2 Wakamatsucho, Shinjuku, Tokyo 162-8480, Japan. Electronic address: [Email]

Abstract

Food exploration is an essential survival behavior in organisms. To find food efficiently, many organisms use a foraging strategy called area-restricted search (ARS) wherein individuals first turn more frequently, restricting their search to one area, then turn less frequently, moving along a straight path to widen the search area. Previous research suggests that the nematode Caenorhabditis elegans shows ARS behavior by changing turn frequency, and that dopamine is a crucial determinant. However, the effects of dopamine on multiple behavioral parameters have remained unknown. Here, we evaluated turn (pirouette) frequency, moving velocity, and specific area occupancy (cell occupancy) over time by using a multiple-worms tracking system. In the control (mock) experiments, all parameters changed over time, but no changes were observed in experiments with dopamine pre-exposed and dopamine-deficient animals. In inverse reinforcement learning analysis, the value function for specific velocity was found to modulate over time in mock animals only. These results demonstrate that dopamine regulates ARS via changes not only to pirouette frequency change but also to velocity.

Keywords

Area-restricted search,Behavioral assay,Dopamine,Foraging,Inverse reinforcement learning,Machine learning,

OUR Recent Articles