![](/wp-content/uploads/2021/03/abstractheadergraphic2012.gif)
Meeting Abstract
S1-1.3 Wednesday, Jan. 4 Making Automated Tracking and Behavior Analysis High Throughput in Practice ROBIE, Alice A.; KABRA, Mayank; BRANSON, Steven; HIROKAWA, Jonathan; KORFF, Wyatt L.; BRANSON, Kristin*; HHMI Janelia Farm; HHMI Janelia Farm; HHMI Janelia Farm; HHMI Janelia Farm; HHMI Janelia Farm; HHMI Janelia Farm bransonk@janelia.hhmi.org
As part of the large scale effort at Janelia to understand the function of the Drosophila melanogaster nervous system through correlation of high-throughput behavioral and neuroanatomical studies, we have combined the tools of FlyBowl, Ctrax, and new machine learning-based behavior classifiers to create a high-throughput behavioral screen of fruit fly locomotor and social behaviors. FlyBowl, a chamber developed to facilitate automated tracking, has been modified to increase throughput and image quality consistency, allowing unsupervised use of an updated Ctrax tracking algorithm. We developed new behavior learning tools that can generalize behavior definitions across over a thousand GAL4 lines. We are currently screening lines from the Rubin GAL4 collection at a rate of 75 lines per week, requiring processing of 400 16-minute videos of 10 male and 10 female flies per week. To provide oversight and visualize behavioral effects in such a large data set, we have been developing visualization tools for examining the stability of experimental conditions, detecting errors in the data collection or analysis, and finding new and interesting behavioral phenotypes. In our TRPA1 screen, we saw significant differences in our automatic metrics for locomotor and social behaviors which recapitulate human annotation. The development of this assay pipeline from data collection through automated analysis allows for the rapid generation of detailed, quantitative descriptions of behavior changes due to sparse neural activation.