BIC: From Cellular and Gene Networks To Principles of Robust Communication and Distributed Systems Design

提出了一种跨学科的方法，涉及计算机科学、数学、统计物理和通信科学的概念，以开发一个综合的分析和软件平台，同时有助于分析和理解几种类型的生物和社会网络，包括DNA-蛋白质、蛋白质-蛋白质、代谢和细胞间信号网络，以及信息和计算网络的设计，包括容错和抗攻击的联邦数据库和自组织通信网络，以及分布式系统，如关键基础设施网络(如国家电网)。复杂网络领域以高度协同的方式发展，涉及物理和生命科学、计算和工程学科。这项提议旨在为不同的社区开发一套共同的工具，从而促进思想的交叉授粉，并刺激综合技术和创新的发展。过去，关于复杂网络的大多数工作都是面向分析的；即识别和测量现有系统的不同特征(例如，互联网、蛋白质相互作用、经验社会网络)，并提出可能导致此类系统出现的潜在动力学。该提案侧重于一种更平衡的办法，其中除了建模工作外，还涉及如何利用建模工作来设计信息网络，以及如何制定系统的方法来从复杂网络中挖掘可用的信息。具体的主题和问题包括：(1)网络断层摄影术--从观测数据推断隐藏的结构：生物网络的结构特征是什么，它们是如何进化和形成的？最近关于大规模生物网络的大量数据才刚刚开始分析，PI建议设计一套复杂的网络工具来解决这些问题。(2)复杂网络中的信息动力学：复杂网络中的信息是如何处理和交流的？在连通性本身是时变的网络中，节点之间能否出现同步或稳定的通信？(3)Net-Modeler：开发一个全面的复杂网络建模器，用于生物网络的静态和动态探索；(4)设计者复杂系统-设计动态信息和计算网络：如何利用生物和其他复杂网络的结构和功能来设计分布式和自组织网络系统的本地协议，从而使系统具有所需的全局属性，如低延迟和健壮的功能？提议中开发的方法和软件工具将整合到一个基于网络的虚拟实验室中。这将导致(I)生物和社会网络分析的全国性资源，以及(Ii)关于复杂网络及其应用的跨学科本科生和研究生课程和培训计划。

An interdisciplinary approach involving concepts from computer science, mathematics, statistical physics, and communication sciences, is proposed to develop an integrated analytical and software platform that would simultaneously facilitate the analysis and understanding of several types of biological and social networks, including DNA-protein, protein-protein, metabolic, and inter-cellular signaling networks, and the design of information and computational networks, including fault-tolerant and attack-resistant federated databases and ad hoc communication networks, and distributed systems such as critical infrastructure networks (e.g., national power grids). The field of complex networks has evolved in a highly synergistic fashion, involving the physical and life sciences, computing and engineering disciplines. This proposal aims to develop a common set of tools for the diverse communities, thus promoting cross-pollination of ideas, and spurring the development of integrated technologies and innovations. In the past, most work on complex networks has been analysis-oriented; that is, identifying and measuring different characteristics of existing systems (e.g., the Internet, protein-protein interactions, empirical social networks) and proposing potential dynamics that could have led to the emergence of such systems. This proposal, focuses on a more balanced approach, where in addition to the modeling work, it also addresses the issue of how the modeling work can be harnessed to design information networks, and how to formulate systematic methods for mining the information available from complex networks. Specific topics and issues pursued include: (1) Network Tomography - Inferring hidden structures from observed data: What are the structural characteristics of biological networks and how have they evolved and formed? The recent flood of data concerning large-scale biological networks are just beginning to be analyzed, and the PI proposes to design a set of complex network tools to address these issues. (2) Information Dynamics In Complex Networks: How is information processed and communicated in complex networks? Can synchronization or stable communication among nodes arise in networks where the connectivity is itself time varying and the same pair of nodes are only intermittently connected? (3) Net-Modeler: development of a comprehensive complex network modeler to be used for both static and dynamical explorations of biological networks; (4) Designer Complex Systems - Designing Dynamic Information and Computing Networks: How to harness the structure and functionalities of the biological and other complex networks to design local protocols of a distributed and ad hoc networked system, so that the system has the desired global properties, such as low-latency and robust functionality? The methodologies and software tools developed in the proposal will be integrated into a virtual web-based laboratory. This will lead to (i) a nationwide resource for the analysis of biological and social networks, and (ii) interdisciplinary undergraduate and graduate courses and training programs on complex networks and their applications.