Data-driven modeling of zebrafish individual and collective behavior
Zebrafish have recently emerged as an important animal model in preclinical studies due to their genetic similarity with humans and ease of use in laboratory studies. Along with this growing interest, experimentation with zebrafish poses ethical issues regarding animal use, thereby requiring the exploration of alternative testing methods. In this context, the advantages of computer simulations, or in-silico experiments, include the capability to pre-test hypotheses, validate existing observations, and inform data collection efforts with far fewer animal experiments. In the first part of my talk, I will present our data-driven modeling framework whose parameters are calibrated on experimental data of zebrafish swimming and validated by comparing traditional behavioral observable obtained from in-silico experiments to those from real data. In the second part of the talk, I will demonstrate how it can be included in traditional models of collective behavior to provide a more realistic computational framework to study zebrafish shoaling with application to the design of control algorithms for multi-agent systems. In the last part of the talk, I will show how it can be leveraged to generate synthetic data-sets to study leader-follower relationships and animal-robot interactions through the informational theoretical construct of transfer entropy.