Zebrafish Dataset Published in Nature Scientific Data

How do zebrafish establish social dominance? Our new dataset published in Nature Scientific Data provides unprecedented insight into this question through high-resolution 3D behavioral tracking. We captured 173 five-hour experiments at 140 frames per second, tracking not just the position but the full posture of each fish—head, pectoral fins, and tail—while maintaining individual identity throughout extended social encounters. The dataset includes wild-type zebrafish alongside disease models for Rett syndrome and aggression-boldness syndrome, revealing how behavioral dynamics differ in health and disease.

This openly available resource enables researchers to investigate the physical principles underlying social behavior at scales previously impossible. From initial aggressive confrontations to the establishment of stable hierarchies, the fine-grained temporal resolution captures behavioral motifs that were previously invisible. The dataset serves multiple communities: biologists studying the neural basis of social behavior, machine learning researchers developing better tracking algorithms, and physicists exploring emergent phenomena in animal groups. Access the complete dataset, code, and documentation through our Zenodo repository and GitHub.

[Read paper] [Download dataset] [View code]

Gaze angle asymetry arising in wile-type, but not mecp2 mutant zebrafish fights. More info in the published paper.