Computational Modeling Core

点击翻译按钮获取中文摘要

Project Summary: Computational Modeling Core Our team for this BRAIN Initiative U19 proposal on Oxytocin Modulation of Neural Circuit Function and Behavior is located at NYU Grossman School of Medicine, and staff for the Computational Modeling Core are located in the NYU Neuroscience Institute alongside the Oxytocin U19 Project and other Core labs. This proximity facilitates their interactions, which has produced a number of ongoing collaborations and will promote the development, validation, implementation, standardization, and dissemination of computational analyses of oxytocin function that are the foundation of this proposed Computational Modeling Core. Core Director Dr. Kenway Louie is a computational neuroscientist with extensive experience in network modeling and behavioral analysis, who will lead a team of postdoctoral-level staff with expertise in the proposed modeling techniques. This Computational Modeling Core will serve as a centralized resource for computational modeling relevant to oxytocin mechanisms in Project team lab research, coordinating theoretical, analytic, and simulation approaches across the Project labs. This research support will enable theoretically-motivated collaborations between the four Project labs, standardize computational modeling approaches to predict behavioral effects of cellular, synaptic, and circuit changes, provide an integrated, hierarchical modeling framework for different social behaviors, and provide straightforward and robust access to the computational examination of oxytocin modulation of circuit function and behavior for all U19 team members. Aim 1 of the Computational Modeling Core is to develop circuit-based models of oxytocin function and individual animal behavior. Using established dynamical rate models customized to circuits of interest, this approach will identify specific oxytocin-related experimental manipulations relevant to the brain areas under study, allowing comparison to experimental data and hypothesis testing about circuit-specific oxytocin function. Aim 2 is to develop agent-based models (ABM) of multi-agent social behaviors, which will examine how oxytocin contributes to the relationship between neural computations, individual animal behavior, and emergent multi-agent social behaviors. This work will make predictions about the effect of specific manipulations (e.g., oxytocin receptor knockout) on specific behaviors (e.g., social hierarchy formation), test ABM predictions in empirical behavior, and refine our understanding of the computational and cognitive role of oxytocin in different social behaviors. Aim 3 is to develop a unified theoretical framework that synthesizes modeling approaches for diverse datasets, specifically drawing on both the dynamical modeling of Aim 1 and ABM frameworks of Aim 2, to capture multiple timescales of social behavior. We will work closely with the Data Science and Behavior Cores to standardize model structures, simulated behavioral and neural data, and metadata curation. We will also ensure accessibility of our work by maintaining a dedicated U19 Computational Modeling Core source code repository.