We have developed Autobehaver, a novel high-throughput platform that integrates scalable recording hardware with state-of-the-art Transformer-based machine learning to decode Drosophila behavior at a granular level. By transforming raw video into rich, quantifiable profiles of movement and orientation, we can now identify subtle phenotypic signatures of human disease models that were previously invisible to the naked eye. This technology not only empowers us to dissect the neural and genetic underpinnings of complex disorders but establishes a scalable foundation for the future: performing high-volume genetic and chemical suppressor screens to identify novel therapeutic targets that can restore healthy behavior.
