Protocol for quantitative analysis of adult zebrafish swimming behavior using DeepLabCut

Here, we present a protocol for quantifying adult zebrafish swimming behavior using open-source software. We describe steps for recording and preprocessing swimming behavior videos, labeling and training a DeepLabCut network, and applying the trained model to analyze behavior. We then detail procedures for processing and visualizing behavioral data. This protocol is optimized for single-fish tracking under standardized recording conditions and can be completed using beginner-level computational skills with basic laboratory hardware.