Computational Models of Infectious Disease Threats

描述（由申请人提供）：微生物威胁，包括生物恐怖主义和自然出现的传染病，对美国的国家安全和全世界的健康构成严重挑战。该提案描述了在约翰霍普金斯大学彭博公共卫生学院与布鲁金斯学会、美国国家航空航天局、马里兰大学和帝国理工学院（伦敦）的主要专家合作建立一个传染病计算建模中心。该项目的总体目标是将最先进和最强大的流行病学数据分析技术与计算机模拟技术（基于主体的建模）相结合，以产生科学合理、高度可视化和用户友好的统一计算流行病学，并响应生物安全和公共卫生政策要求。数据分析将以这样一种认识为指导，即流行病在空间和时间上的模式可以作为几乎可分解的系统来处理，在这种系统中，发病率信号的频率成分可以被分离和研究。小波变换和使用Hilbert-Huang变换的经验模态分解将用于筛选非线性，非平稳的流行病学数据，允许定义频带模式。然后，孤立的频率模式将与外部强迫（天气、社会接触模式）和内部动力（Kermack-McKendrick捕食-猎物模型）联系起来。流行病学数据分解分析的结果以及传染病专家的知识将指导基于主体的模型的创建和发展。这些模型的特点是在人工社会中流动的个体群体与其他个体在当地互动。基本模型的特征包括可变的社会网络结构、个体易感性和免疫力、潜伏期、传播率、接触率和其他可选参数。在对基于主体的模型进行校准以生成与现实世界流行病学一致的流行病模式之后，将系统地引入预防策略，包括接种疫苗、接触者追踪、隔离、检疫和其他公共卫生措施，并评估其影响。将开发方法来评估单个模型的效用，并根据多个模型的综合结果做出决策。最初研究的传染病包括天花、SARS、登革热、西尼罗河和未知但假设合理的病原体。作为合作协议的一部分，该中心将与其他研究小组、生物信息学核心小组和NIGMS合作开发数据集、软件和方法、基于代理的模型和可视化工具。在传染病流行的紧急情况下，该中心将在国家安全管理系统的指导下重新调整其活动，以服务于国家安全。

DESCRIPTION (provided by applicant): Microbial threats, including bioterrorism and naturally emerging infectious diseases, pose a serious challenge to national security in the United States and to health worldwide. This proposal describes the creation of a center for computational modeling of infectious diseases at the Johns Hopkins Bloomberg School of Public Health, with the collaboration of key experts at the Brookings Institution, the National Aeronautic and Space Administration, the University of Maryland, and Imperial College (London). The overarching aim of this project is to integrate the most advanced and powerful techniques of epidemiological data analysis with those of computer simulation (agent-based modeling) to produce a unified computational epidemiology that is scientifically sound, highly visual and user-friendly, and responsive to biosecurity and public health policy requirements. Data analysis will be guided by the insight that epidemic patterns over space and time can be approached as nearly decomposable systems, in which frequency components of the incidence signal can be isolated and studied. Wavelet transforms, and empiric mode decomposition using Hilbert-Huang Transforms, will be used to sift nonlinear, nonstationary epidemiological data, allowing frequency band patterns to be defined. Isolated frequency modes will then be associated with external forcing (weather, social contact patterns) and internal dynamics (Kermack-McKendrick predator-prey models). Results of the epidemiological data decomposition analysis, along with the knowledge of infectious disease experts, will instruct the creation and development of agent-based models. Such models feature populations of mobile individuals in artificial societies that interact locally with other individuals. Features of the basic model include variable social network structures, individual susceptibility and immunity, incubation periods, transmission rates, contact rates, and other selectable parameters. After the agent-based model is calibrated to generate epidemic patterns consistent with real world epidemiology, preventive strategies including vaccination, contact tracing, isolation, quarantine, and other public health measures will be systematically introduced and their impact evaluated. Methods will be developed for assessing the utility of individual models, and for making decisions based on combined results from more than one model. Infectious diseases to be studied initially include smallpox, SARS, dengue, West Nile, and unknown but hypothetically plausible agents. As part of a Cooperative Agreement, the Center will work with other research groups, a bioinformatics core group, and the NIGMS to develop data sets, software and methods, agent-based models, and visualization tools. In an infectious disease epidemic emergency the Center will redirect its activities to serve the nation's security, as guided by the NIGMS.