Statistical Analysis of Zebrafish Locomotor Response

Area of Science:

• Neurobiology and Pharmacology
• Biostatistics
• Animal Behavior

Background:

• Zebrafish larvae exhibit complex locomotor behaviors in response to stimuli.
• Analyzing large-scale time-series locomotor data presents statistical challenges due to multiple influencing factors.
• Existing methods are insufficient for robustly comparing larval activity profiles.

Purpose of the Study:

• To establish a statistical framework for analyzing zebrafish visual motor response (VMR) data.
• To compare locomotor responses across different wild-type (WT) zebrafish strains and developmental stages.
• To investigate factors influencing VMR and assess the statistical power of the proposed methods.

Main Methods:

• Application of Hotelling's T-squared test for comparing locomotor profiles over time.
• Multivariate analysis of variance (MANOVA) to identify factors affecting VMR.
• Dynamical analysis of larval activity at one-second intervals and power analysis for test performance evaluation.

Main Results:

• Hotelling's T-squared test effectively differentiated VMR between WT strains and developmental stages.
• MANOVA revealed that larval activity is significantly influenced by developmental stage, light stimulus, their interaction, and plate location.
• Biological and technical repeats showed negligible effects, suggesting data pooling can enhance statistical power.

Conclusions:

• A robust statistical framework using Hotelling's T-squared test and MANOVA is established for VMR data analysis.
• The framework demonstrates sensitivity in detecting differences in larval behavior and identifies key influencing factors.
• The findings support combining experimental repeats to improve statistical power in zebrafish locomotor studies.