BRAIN EAGER: Discovery and characterization of neural circuitry from behavior, connectivity patterns and activity patterns

BRAIN EAGER：从行为、连接模式和活动模式中发现和表征神经回路

• 批准号: 1451081
• 负责人: Carey Priebe
• 金额: $ 30万
• 依托单位: Johns Hopkins University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-09-01 至 2016-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1451081&HistoricalAwards=false
• 关键词: BRAIN; EAGER; Discovery; characterization; neural

Johns Hopkins University is awarded a grant for research leading to an improved understanding of how the brain is connected. Drosophila larvae, with 10000 neurons (and about 1000 neuron types), offer an opportunity to determine how an entire nervous system generates behavior. The research combines information from three sources: a neuron activity map of the entire Drosphila larval nervous system; a library of neuronal lines yielding a neuron behavior map; and a "wiring diagram" or connectome for the entire larval nervous system. Together, the neuron-behavior map, the neuron-activity map, and the connectome complement one another, laying the groundwork for a brain-wide understanding of the principles by which brains generate behavior.The technical goals for this project will be to develop principled statistical pattern recognition & machine learning methods for clustering neurons based on three different data sets, both individually and jointly. The extent to which clusters obtained from the three datasets agree, and the manner in which they disagree, will reveal how the structure of neural circuits relates to their function and how brains generate behavior. Current methods for discovery and characterization of neural circuitry from behavior, connectivity patterns and activity patterns - fusion and inference from multiple disparate data sources- are insufficient; the approach developed in this project will yield improved methods developed in conjunction with neuroscientists.

约翰霍普金斯大学被授予一项研究经费，以提高对大脑如何连接的理解。果蝇幼虫有10000个神经元（大约1000种神经元类型），这为我们提供了一个确定整个神经系统如何产生行为的机会。 该研究结合了来自三个来源的信息：整个果蝇幼虫神经系统的神经元活动图;产生神经元行为图的神经元线库;以及整个幼虫神经系统的“接线图”或连接体。神经元行为图、神经元活动图和连接体相互补充，为全脑范围内理解大脑产生行为的原理奠定了基础。该项目的技术目标是开发基于三个不同数据集（单独和联合）的原则性统计模式识别机器学习方法，用于聚类神经元。从三个数据集获得的聚类一致的程度，以及它们不一致的方式，将揭示神经回路的结构如何与它们的功能相关，以及大脑如何产生行为。目前从行为、连接模式和活动模式中发现和表征神经回路的方法--从多个不同数据源中融合和推断--是不够的;本项目中开发的方法将产生与神经科学家一起开发的改进方法。