The heart and the mind: an integrative approach to brain-body interactions in the zebrafish Our current U19 has focused primarily on Exteroception, which can be defined as the accumulated sensory experience originating from events in the outside world. However, all neural computation takes place in the context of the body, which is subject to the dynamics of hunger, fatigue, motivation and diurnal cycles. The information that reaches the brain from the organs and receptors inside of the body, summarized under the umbrella term of Interoception, is therefore overlaid onto exteroceptive computations. This adds behavioral variability to any animal task, from changing vigor to altering valence of rewards to vetoing responses altogether. To better understand such embodied computation, we will first incorporate cardiac and respiratory variables that dynamically report basic internal states to our standard behavioral classification methods, and we will include detailed characterization and modeling of the autonomic and intracardiac nervous system in our functional imaging studies. Finally, as a critical resource for the description and validation of the generated neural circuit models, we will create a complete body-and-brain connectome of the larval zebrafish. This research plan is aligned with the three levels of understanding articulated by David Marr more than thirty years ago, who emphasized the importance of considering Evolutionary Principles, Algorithmic Representation and Hardware Implementation as a unified and interconnected approach for understanding the brain. Following this framework, we first try to isolate aspects of fish behavior that are adapted to their native environment and context. We next use a variety of controlled behavioral assays to extract the algorithmic rules that govern the dynamics of modular sensorimotor transformations, and that are augmented by knowledge about internal state changes as reflected by observed modulation of cardiac and respiratory activity. To characterize the hardware implementation of these explicit and latent behavioral algorithms, we propose to first measure cellular activity using brain-wide functional light imaging of the central, the autonomic and the intracardiac nervous systems (CNS, ANS and ICNS). We next will use these comprehensive datasets to generate realistic circuit models of the observed dynamics. These models will then guide a series of circuit dissection experiments, such as optogenetic perturbations of specific cell types, targeted patch-clamp physiology and sparse connectomics tracing, allowing us to validate, extend, and constrain our models. We note that this approach also affords us the opportunity to discover new circuit elements and circuit motifs, and novel ways to implement computational algorithms by the brain. Another extension of our current U19 is the addition of a dynamical modeling framework for these behaviors, where animals can interact flexibly with the environment, an...