CIF: BCSP: Large: Connectivity and Information Flow in a Complex Brain

大脑回路运作背后的网络特性基本上还不清楚。没有一个复杂的大脑可以实时观察到整体的活动模式。该项目将绘制一个小而复杂的大脑中信息流的主要模式，并分析其功能网络架构。以一种完全不偏不倚的方式对待网络，它将通过人工刺激大脑电路产生数据，旨在调查信息流的路径，而不考虑大脑为什么在做任何事情或输出是什么。这项任务将在实验和计算神经学家以及网络理论家之间不同寻常的合作下进行，以绘制通过果蝇复杂但相对较小的大脑的信息流，然后分析其揭示的网络功能原理。关注果蝇大脑的理由是，它足够小，可以完全监控它的活动，但也足够复杂(15万个神经细胞)，可以为复杂的大脑网络的主要问题提供解决方案。此外，其神经细胞连接的整个布线模式正在顺利进行，很快就会完成。本项目的目标是从其底层功能和架构的角度来描述网络的特征，这可能与其实际物理布局不同。有了手头的数据，我们将对包含在大脑网络中的设计原理进行工程搜索。根据这些发现，基于第一次广泛的全网络分析，将形成大脑中信息流的一般理论。智力价值：该项目旨在从生物系统中挖掘可应用于计算和工程问题的原理和策略。同时，它以一种技术上可行的方式解决了神经科学中的一个主要问题。要理解大脑使用的整体策略面临的挑战是，在大脑的运作中，整体大于部分之和。这一特性被称为“紧急”特性，它源于数千个神经细胞的信号活动的时间和模式。如果人们能够获得这些数据，并将这种活动与大脑神经细胞连接的模式联系起来，那么就有可能为大规模神经元编码制定一个理论基础。了解这一过程，反过来又为计算和工程中的应用带来了巨大的希望。广泛的影响：教育影响：除了为研究生、博士后研究员和本科生提供培训机会外，了解大脑中信息处理的基本策略将使我们能够设计最大限度地利用这种理解的教育策略。技术影响：该项目将产生的原则和理论公式也有可能为计算机网络设计和信号的交通流动提供新的战略，并为涉及网络和信息流的工程战略提供新战略。为该项目概述的实验构成了一种新的方法，以解决是否存在大脑网络操作的基本基本原理的问题，如果识别出这些基本原理，将对应用程序产生广泛的影响，例如为计算、工程设备和通信而构建的人工网络的设计和实现。此外，人们热衷于在这些领域寻求适应性、多功能性和健壮性等特性，并致力于能够捕捉到基因网络执行这些特性的能力。

The network properties underlying the circuit operations of the brain are still basically unknown. There is no complex brain in which the overall activity patterns have been observed in real time. This project will chart the major patterns of information flow in a small, complex brain, and analyze its functional network architecture. Treating the network in a totally unbiased way, it will generate data from artificial stimulation of brain circuits aimed at surveying the paths of information flow, without regard to why the brain is doing anything or what the output is. The task will be carried out in an unusual collaboration between experimental and computational neuroscientists and network theorists to chart information flow through the complex, but relatively small, brain of the fruit fly Drosophila and then to analyze the principles of network function that it reveals. The rationale for focusing on the fruit fly brain is that it is small enough to be possible to monitor its activity completely, but also complex enough (150,000 nerve cells) to offer solutions to the major problems of complex brain networks. In addition, the entire wiring pattern of its nerve cell connections is well under way and will soon be completed. The goal of this project is to characterize the network from the perspective of its underlying function and architecture, which not may not be identical with its actual physical layout. With the data in hand, a search will be conducted for the engineering the design principles embodied in the brain's networks. From these findings, a general theory of information flow in the brain will be formulated, based on the first broad-based, whole network analysis.Intellectual Merit: The project aims at mining biological systems for principles and strategies that can be applied to problems in computing and engineering. At the same time, it tackles a major problem in neuroscience in a technically feasible fashion. The challenge to understanding the overall strategy used by the brain is the fact that, in its operations, the whole is greater than the sum of the parts. This feature, known as an "emergent" property, arises from the timing and patterns of signaling activity of many thousands of nerve cells. If one had access to that data, and could correlate the activity with the pattern of nerve cell connections brain, it would be possible to formulate a theoretical basis for large-scale neuronal coding. Understanding this process, in turn, holds substantial promise for applications in computing and engineering.Broader Impact: Educational Impact: In addition to providing the opportunity for training of graduate students, postdoctoral fellows, and undergraduates, a knowledge of the fundamental strategies of information processing in the brain would allow us to design educational strategies that take maximum advantage of that understanding. Technological Impact: The principles and theoretical formulation that will emerge from this project also have the potential for providing novel strategies for computer network design and traffic flow of signals, and for engineering strategies that involve networks and information flow. The experiments outlined for this project constitute a new approach to the question of whether there are fundamental underlying principles of brain network operation which, if discerned, would have wide-ranging implications for applications such as the design and implementation of artificial networks constructed for computing, engineered devices, and communications. In addition, properties such as adaptability, versatility, and robustness are keenly sought in these areas, and efforts are aimed at being able to capture the ability of gene networks to carry them out.