PyRAT

PyRAT analyzes rodent tracking data to classify behaviors, estimate movement metrics, and associate behaviors with synchronized neural recordings.

Key Features:

• Behavior Classification: Employs unsupervised machine learning, including hierarchical agglomerative clustering and t-distributed stochastic neighbor embedding (t-SNE), to identify distinct behaviors from tracking data without prior labels.
• Movement Metrics Estimation: Calculates traveled distance, speed, and area occupancy from tracking data.
• Integration with Neural Data: Associates detected behaviors with synchronized neural recordings to enable correlation analyses between actions and brain activity.
• Visualization Tools: Visualizes behavior–neural associations in pixel space to facilitate comparison of behavioral patterns and neural signals.

Scientific Applications:

• Neuroscience research: Supports studies linking animal behavior and brain activity using synchronized behavioral tracking and neural recordings.
• Rodent experimental paradigms: Enables detailed analysis of rodent movements and behaviors for experiments on neurological conditions, behavioral responses to stimuli, and pharmacological interventions.

Methodology:

Processes raw tracking data using unsupervised clustering algorithms; applies hierarchical agglomerative clustering to group behaviors and t-SNE to visualize high-dimensional features, and uses algorithms to associate detected behaviors with synchronized neural data.