FB2026_03 , released September 17, 2026
Reference Report
Open Close
Reference
Citation
Duarte, T., Blache, D., Corte, C.L.D. (2026). A centroid-based video-tracking framework for integrated analysis of sleep, locomotor activity, and spatial behavior in Drosophila melanogaster.  J. Neurosci. Methods 435(): 110868.
FlyBase ID
FBrf0266081
Publication Type
Research paper
Abstract
Sleep in Drosophila melanogaster is commonly quantified using infrared beam-break systems, which are robust for high-throughput monitoring but provide limited information about the spatial and temporal organization of behavior. We proposed that continuous centroid-based video tracking could provide an accessible, no-code framework that simultaneously extracts sleep-related immobility, locomotor activity, temporal movement profiles, and spatial behavioral readouts. A centroid-based video-tracking framework was developed in ANY-maze for integrated analysis of sleep, locomotor, temporal, and spatial occupancy in D. melanogaster. Individual flies were recorded in horizontal tubes, and the workflow extracted immobility-defined sleep, sleep architecture, locomotor output, activity distribution, and spatial occupancy. Analytical reproducibility was assessed through repeated analysis of identical recordings, and sleep-related outputs were compared with DAM2 measures under light-dark (LD) and constant-light (LL) conditions. The framework showed high analytical reproducibility across outputs. Under LL, canonical sleep phenotypes were detected, including reduced total sleep and fragmented sleep architecture. Continuous tracking also detected changes in spatial occupancy and temporal movement organization that were not captured by cumulative activity measures. Biological outcomes for LL-induced sleep disruption were similar between ANY-maze and DAM2. We discuss methodological differences between continuous centroid tracking and event-based beam-break monitoring. Unlike beam-based systems, the ANY-maze workflow preserves spatial and temporal information without custom programming. The no-code, video-based ANY-maze framework provides an accessible, reproducible strategy for multidimensional behavioral phenotyping in fruit flies and extends behavioral interpretation beyond total sleep or activity measures while remaining compatible with standard laboratory setups.
PubMed ID
PubMed Central ID
Associated Information
Comments
Associated Files
Other Information
Secondary IDs
    Language of Publication
    English
    Additional Languages of Abstract
    Parent Publication
    Publication Type
    Journal
    Abbreviation
    J. Neurosci. Methods
    Title
    Journal of Neuroscience Methods
    Publication Year
    1979-
    ISBN/ISSN
    0165-0270
    Data From Reference