High-resolution, high-throughput analysis of Drosophila geotactic behavior

Abstract

Drosophila innate response to gravity, geotaxis, has been previously used to assess the impact of aging and disease on motor performance. Despite its rich history, fly geotaxis continues to be largely measured manually and assessed through simplistic metrics. The manual nature of this assay introduces substantial experimental variability while simplistic metrics provide limited analytic insights into the behavior. To address these shortcomings, we have constructed a fully automated, programable apparatus, and developed a multi-object tracking software capable of following sub-second movements of individual flies, thus allowing reproducible, detailed, and quantitative analysis of geotactic behavior. The apparatus triggers and monitors geotaxis of 10 fly cohorts simultaneously, with each cohort consisting of up to 7 flies. The tracking program isolates cohorts and records individual fly coordinate outputs allowing for simultaneous multi-group, multifly tracks per experiment, greatly improving throughput and resolution. The algorithm tracks individual flies during the entire run with [~]97% accuracy, yielding detailed climbing curve, speed, and movement direction with 1/30 second resolution. Our tracking also allows the construction of multi-variable metrics and the detection of transitory movement phenotypes, such as slips and falls, which have thus far been neglected in geotaxis studies due to limited spatio-temporal resolution. Through a combination of automation and robust tracking, the platform is therefore poised to advance Drosophila geotaxis assay into a comprehensive assessment of locomotor behavior.