Integrating Bioinformatics and Clustering Analysis for Disease Surveillance

 描述(由申请人提供)：在生物信息学中，有一个巨大的重点是将数据从实验台转化为临床决策的信息和知识。这包括对个性化药物和治疗的人类遗传学分析。然而，对用于公共卫生实践的转译生物信息学的关注要少得多，例如监测新出现的/重新出现的病毒。这涉及到病毒遗传学的数据获取、整合和分析，以推断新毒株出现时的起源、传播和进化。这一做法的相关科学领域包括分子流行病学和系统地理学的某些方面。最近的注意力集中在人畜共患病的病毒上，这种病毒被定义为可以在动物和人之间传播的病原体。除了季节性流感和西尼罗河病毒外，这种病原体分类还包括中东呼吸综合征和甲型H7N9流感等新型病毒。尽管文献中强调了这些成功，但在国家公共卫生、农业和野生动物机构中，几乎没有利用生物信息学资源和工具进行人畜共患病监测。以前，这种类型的资源主要限于学术界的人。 虽然生物信息学很少用于人畜共患病病毒的监测，但其他应用程序，如地理空间信息系统(GIS)，已被国家卫生机构用于分析感染的空间模式。这包括使用一系列数据类型(如临床、地理或人员流动性数据)生成疾病地图的软件，用于地理编码、集群或疫情检测等任务。此外，地理空间统计方面的进步使卫生机构能够进行更强大的时空分析，以推断时空模式。然而，这些地理信息系统只考虑传统的流行病学数据，如报告病例的位置和时间，而不考虑导致疾病的病毒的基因。这使卫生机构无法了解病毒基因组及其传播环境的变化如何影响疾病风险。 这项建议的长期目标是通过应用生物信息学原理来访问、集成和分析病毒遗传学和时空可报告的疾病数据，从而加强对人畜共患病病毒地理空间热点的识别。该项目将包括生物信息学、遗传学、空间统计学、地理信息系统和流行病学的方法。为此，我将首先衡量生物信息学资源和工具的利用，以及国家公共卫生、农业和野生动物机构为检测和预测人畜共患病病毒热点(群)而确定的当前方法和限制(目标1)。然后，我将使用这个反馈来开发一个空间决策支持系统，用于检测和预测人畜共患病热点，应用生物信息学原理来访问、集成和分析病毒遗传学、环境和时空可报告的疾病数据(目标2)。在目标3中，我将评估我的集群检测和预测系统，对照一个不考虑病毒遗传学和依赖传统时空数据的系统，并执行预测能力的验证。其他内容 将对用户满意度和系统可用性进行评估。

 DESCRIPTION (provided by applicant): There has been a tremendous focus in bioinformatics on translation of data from the bench into information and knowledge for clinical decision-making. This includes analysis of human genetics for personalized medicine and treatment. However, there has been much less attention on translational bioinformatics for public health practice such as surveillance of emerging/re-emerging viruses. This involves data acquisition, integration, and analyses of viral genetics to infer origin, spread, and evolution suc as the emergence of new strains. The relevant scientific fields for this practice include certain aspects of molecular epidemiology and phylogeography. Recent attention has focused on viruses of zoonotic origin, which are defined as pathogens that are transmittable between animals and humans. In addition to seasonal influenza and West Nile virus, this classification of pathogens includes novel viruses such as Middle Eastern Respiratory Syndrome and influenza A H7N9. Despite the successes highlighted in the literature, there has been little utilization of bioinformatics resources and tools among state public health, agriculture, and wildlife agencies for zoonotic surveillance. Previously this type of resource has been restricted primarily to those in academia. While bioinformatics has been sparsely used for surveillance of zoonotic viruses, other applications such as Geospatial Information Systems (GIS) have been employed by state health agencies to analyze spatial patterns of infection. This includes software to produce disease maps using an array of data types such as clinical, geographical, or human mobility data for tasks such as, geocoding, clustering, or outbreak detection. In addition, advances in geospatial statistics have enabled health agencies to perform more powerful space-time analyses to infer spatiotemporal patterns. However, these GIS consider only traditional epidemiological data such as location and timing of reported cases and not the genetics of the virus that causes the disease. This prevents health agencies from understanding how changes in the genome of the virus and the associated environment in which it disseminates impacts disease risk. The long-term goal of this proposal is to enhance the identification of geospatial hotspots of zoonotic viruses by applying bioinformatics principles to access, integrate, and analyze viral genetics and spatiotemporal reportable disease data. This project will include approaches from bioinformatics, genetics, spatial statistics, GIS, and epidemiology. To do this, I will first measue the utilization of bioinformatics resources and tools as well as the current approaches and limitations identified by state agencies of public health, agriculture, and wildlife for detecting nd predicting hotspots (clusters) of zoonotic viruses (Aim 1). I will then use this feedback to develo a spatial decision support system for detecting and predicting zoonotic hotspots that applies bioinformatics principles to access, integrate, and analyze viral genetics, environmental, and spatiotemporal reportable disease data (Aim 2). In Aim 3, I will then evaluate my system for cluster detection and prediction against a system that does not consider viral genetics and relies on traditional spatiotemporal data, and perform validation of the predictive capability. Additional evaluation of the user's satisfaction and system usability will be evaluated.