Computational Neurophysiology Core

COMPUTATIONAL NEUROPHYSIOLOGY CORE - PROJECT SUMMARY The purpose of the COMPUTATIONAL NEUROPHYSIOLOGY CORE is to support the scientific translation across species, experimental modalities, and neural system scales that underlies our Center. To do this, the Core will serve as the central hub for all Projects, providing analytical support for the projects as well as carrying out its own computational modeling aims. The Core will ensure that trial-by-trial behavioral and neurophysiological data are analyzed in a uniform and compatible manner, so that findings can be compared and integrated across the five Projects. It will also provide expertise in identifying computational validity between instantiations of tasks, homologies between brain systems across species, and crossing abstract levels of analysis to maximize interpretability. Our fundamental hypothesis is that psychosis is characterized by computational changes in state representations — how the brain estimates, stores, and processes information about the state of the world. To test this hypothesis, our Center will perform two parallel translational tasks across all Projects and species: a cognitive control Translational Orientation Pattern EXpectancy Task (TOPX, derived from the well-studied DPX task), and a reward-based Translational restless Bandit Task (TBT, derived from well-studied probabilistic reinforcement learning paradigms). Each task allows us to probe different aspects of state representation (state estimation, learning, and maintenance), facilitating the translation of experimental findings across mice, macaques, and humans. This translation will be facilitated by similar analytics pipelines across Projects, guided by the expertise and efficiency of our central Core. We will also ensure that all analyses are conducted using advanced techniques such as machine learning and mixed models. In addition, and as a foundational component of our Center’s scientific agenda, our Core will pursue specific computational modeling aims, with the goal of applying current theories of state representation and attractor dynamics to the specific experiments done in the five Projects. In Aim 1, we will develop integrated analysis strategies to assess state representation disruptions within and across the TOPX and TBT tasks. In Aim 2, we will integrate analyses of TOPX and TBT behavioral data with computational neural hypotheses. TOPX and TBT parameters will be integrated with neurophysiologic findings across species, and translation will be facilitated by using connectivity data (combination of diffusion and resting-state fMRI) to identify homologies across nonhuman animals and humans. We will focus particularly on salience, default mode, and central executive networks, along with their shared and divergent connectivity with the striatum.