Normalization of large-scale behavioural data collected from zebrafish.

Many contemporary neuroscience experiments utilize high-throughput approaches to simultaneously collect behavioural data from many animals. The resulting data are often complex in structure and are subjected to systematic biases, which require new approaches for analysis and normalization. This study addressed the normalization need by establishing an approach based on linear-regression modeling. The model was established using a dataset of visual motor response (VMR) obtained from several strains of wild-type (WT) zebrafish collected at multiple stages of development. The VMR is a locomotor response triggered by drastic light change, and is commonly measured repeatedly from multiple larvae arrayed in 96-well plates. This assay is subjected to several systematic variations. For example, the light emitted by the machine varies slightly from well to well. In addition to the light-intensity variation, biological replication also created batch-batch variation. These systematic variations may result in differences in the VMR and must be normalized. Our normalization approach explicitly modeled the effect of these systematic variations on VMR. It also normalized the activity profiles of different conditions to a common baseline. Our approach is versatile, as it can incorporate different normalization needs as separate factors. The versatility was demonstrated by an integrated normalization of three factors: light-intensity variation, batch-batch variation and baseline. After normalization, new biological insights were revealed from the data. For example, we found larvae of TL strain at 6 days post-fertilization (dpf) responded to light onset much stronger than the 9-dpf larvae, whereas previous analysis without normalization shows that their responses were relatively comparable. By removing systematic variations, our model-based normalization can facilitate downstream statistical comparisons and aid detecting true biological differences in high-throughput studies of neurobehaviour.

Statistical Analysis of Zebrafish Locomotor Behaviour by Generalized Linear Mixed Models.

Upon a drastic change in environmental illumination, zebrafish larvae display a rapid locomotor response. This response can be simultaneously tracked from larvae arranged in multi-well plates. The resulting data have provided new insights into neuro-behaviour. The features of these data, however, present a challenge to traditional statistical tests. For example, many larvae display little or no movement. Thus, the larval responses have many zero values and are imbalanced. These responses are also measured repeatedly from the same well, which results in correlated observations. These analytical issues were addressed in this study by the generalized linear mixed model (GLMM). This approach deals with binary responses and characterizes the correlation of observations in the same group. It was used to analyze a previously reported dataset. Before applying the GLMM, the activity values were transformed to binary responses (movement vs. no movement) to reduce data imbalance. Moreover, the GLMM estimated the variations among the effects of different well locations, which would eliminate the location effects when two biological groups or conditions were compared. By addressing the data-imbalance and location-correlation issues, the GLMM effectively quantified true biological effects on zebrafish locomotor response.

Computational classification of different wild-type zebrafish strains based on their variation in light-induced locomotor response.

Zebrafish larvae display a rapid and characteristic swimming behaviour after abrupt light onset or offset. This light-induced locomotor response (LLR) has been widely used for behavioural research and drug screening. However, the locomotor responses have long been shown to be different between different wild-type (WT) strains. Thus, it is critical to define the differences in the WT LLR to facilitate accurate interpretation of behavioural data. In this investigation, we used support vector machine (SVM) models to classify LLR data collected from three WT strains: AB, TL and TLAB (a hybrid of AB and TL), during early embryogenesis, from 3 to 9 days post-fertilisation (dpf). We analysed both the complete dataset and a subset of the data during the first 30after light change. This initial period of activity is substantially driven by vision, and is also known as the visual motor response (VMR). The analyses have resulted in three major conclusions: First, the LLR is different between the three WT strains, and at different developmental stages. Second, the distinguishable information in the VMR is comparable to, if not better than, the full dataset for classification purposes. Third, the distinguishable information of WT strains in the light-onset response differs from that in the light-offset response. While the classification accuracies were higher for the light-offset than light-onset response when using the complete LLR dataset, a reverse trend was observed when using a shorter VMR dataset. Together, our results indicate that one should use caution when extrapolating interpretations of LLR/VMR obtained from one WT strain to another.

Statistical Analysis of Zebrafish Locomotor Response.

Zebrafish larvae display rich locomotor behaviour upon external stimulation. The movement can be simultaneously tracked from many larvae arranged in multi-well plates. The resulting time-series locomotor data have been used to reveal new insights into neurobiology and pharmacology. However, the data are of large scale, and the corresponding locomotor behavior is affected by multiple factors. These issues pose a statistical challenge for comparing larval activities. To address this gap, this study has analyzed a visually-driven locomotor behaviour named the visual motor response (VMR) by the Hotelling's T-squared test. This test is congruent with comparing locomotor profiles from a time period. Different wild-type (WT) strains were compared using the test, which shows that they responded differently to light change at different developmental stages. The performance of this test was evaluated by a power analysis, which shows that the test was sensitive for detecting differences between experimental groups with sample numbers that were commonly used in various studies. In addition, this study investigated the effects of various factors that might affect the VMR by multivariate analysis of variance (MANOVA). The results indicate that the larval activity was generally affected by stage, light stimulus, their interaction, and location in the plate. Nonetheless, different factors affected larval activity differently over time, as indicated by a dynamical analysis of the activity at each second. Intriguingly, this analysis also shows that biological and technical repeats had negligible effect on larval activity. This finding is consistent with that from the Hotelling's T-squared test, and suggests that experimental repeats can be combined to enhance statistical power. Together, these investigations have established a statistical framework for analyzing VMR data, a framework that should be generally applicable to other locomotor data with similar structure.