RI: Medium: Collaborative Research: Through synapses to spatial learning--a topological approach

在神经科学中，在感兴趣的涌现现象（如学习和记忆）与获取大多数数据的水平之间存在紧张关系。例如，许多实验室通过建立体外电生理学测量与动物行为实验中收集的数据之间的经验相关性来研究突触连接的强度及其动力学如何影响认知。然而，这些相关性缺乏因果解释：到目前为止，还不存在将单个神经元和突触中的记录与认知学习动力学联系起来的机制。这个问题并不是因为缺乏对神经元或系统水平的观察;相反，它反映了我们将这两个尺度联系起来的能力的主要差距。即使能够对大脑中的每一个神经元进行完整的描述，我们仍然有能力从局部数据过渡到对它如何结合产生系统认知结果做出定性结论。解决这个问题需要一个概念框架，包括一个计算模型，将实验得出的个别细胞的特性与这些特性在合奏水平的影响。拟议的研究旨在提供一种方法来建立这样的连接：开发一个数据驱动的，系统的海马空间学习模型的基础上的参数的海马突触结构，包括参数的突触可塑性，使用新的拓扑和几何技术。代数拓扑学的最新发展将用于整合突触连接性和突触可塑性的参数（例如，长期和短期的增强和抑郁症），研究这个地图的结构，它的形成和恶化的机制，并评估所需的时间来产生一个给定的环境空间地图等。这个项目是一个自然的演变以前的工作所做的Dabaghian组建模的机制，空间学习，代数拓扑方法的基础上开发的M？莫利集团对学习现象的基于理论的洞察将使我们更好地理解如何解释数据，如何设计新的实验，测量中应该针对哪些变量，以及如何最大限度地减少动物的使用，该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准。

There is a tension in neuroscience between the emergent phenomena of interested, such as learning and memory, and the level at which most data are acquired. For example, numerous experimental labs study how the strengths of synaptic connections and their dynamics affect cognition by establishing empirical correlations between in vitro electrophysiology measurements and data collected in animal behavioral experiments. However, these correlations fall short of causal explanations: to date, there exist no mechanisms connecting recordings in individual neurons and synapses with cognitive learning dynamics. The problem is not due to a lack of observations at either the neuronal or the systemic level; rather, it reflects a principal gap in our ability to link these two scales. Even if a full description of every neuron in the brain could be produced, there would still be a gap in our ability to transition from local data to making qualitative conclusions about how it combines to produce systemic cognitive outcomes. Addressing this problem requires a conceptual framework encompassing a computational model that would link the experimentally derived characteristics of individual cells with effects of those characteristics at the ensemble level. The proposed research aims to provide a way to establish such a connection: developing a data-driven, systemic model of hippocampal spatial learning based on the parameters of the hippocampal synaptic architecture, including the parameters of synaptic plasticity, using novel topological and geometric techniques. Recent developments in Algebraic Topology will be used to integrate the parameters of synaptic connectivity and synaptic plasticity (e.g., long- and short-term potentiation and depression), to study structure of this map, the mechanisms of its formation and deterioration, and to evaluate the time required to produce a spatial map of a given environment, etc. This project is a natural evolution of prior work done by the Dabaghian group on modeling the mechanisms of spatial learning, based on algebraic topology methods developed by the M?moli group. The theory-based insight into learning phenomena will produce a qualitatively better understanding of how to interpret data, how to design new experiments, what variables should be targeted in measurements, as well as how to minimize use of animals, and in general how to optimize use physical and intellectual resources.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.