Regression, Phylogenetics, and Study Design in Infectious Disease Epidemiology

 描述（由申请人提供）：从约翰·斯诺对霍乱流行的调查开始，了解和预防传染病传播一直是流行病学的基本目标之一。病毒和细菌的全基因组序列是一个很有前途的新的疾病传播信息来源，但目前的统计方法无法将这些数据纳入家庭和其他密切接触群体的传播分析。长期目标是开发统计和流行病学方法，利用高分辨率传播数据和基因序列数据，为快速有效地应对新出现的感染提供信息。拟议研究的目标是为传染病传播数据开发可整合病原体基因序列的灵活和稳健的回归模型。这些将以最近开发的半参数回归模型为基础，该模型可以估计流行病数学模型和干预措施设计的关键参数，包括协变量对传染性和易感性的影响的风险比，以及感染-易感对中传播的基线风险。为了使其成为传染病流行病学的更实用的工具，该模型将被扩展到考虑外部感染源、缺失数据和小样本。该模型的偏似然是与人、地点和时间的流行病学数据一致的传播树集合的总和。由于将来自受感染个体的病原体样本联系起来的遗传关系限制了可能的传播树的集合，因此病原体遗传序列数据可以与流行病学数据相结合，以获得更有效的传播参数估计。流行病学和遗传学数据将通过开发算法结合起来，以同时找到与两者一致的传播树集。这些算法将被纳入马尔可夫链蒙特卡罗或序贯蒙特卡罗估计程序，将考虑缺失数据和系统发育的不确定性。这些方法将作为为疫情调查和疫苗试验开发有效的病例对照和病例队列研究设计的理论基础。这项研究具有创新性，因为它综合了生存分析和统计遗传学来分析传染病传播数据。它之所以重要，是因为它将改善数据的收集和分析以及对流行病干预措施的评价，从而能够更有效地控制新出现的感染。

 DESCRIPTION (provided by applicant): Beginning with John Snow's investigations of cholera epidemics, understanding and preventing infectious disease transmission has been one of the fundamental goals of epidemiology. Whole-genome sequences from viruses and bacteria are a promising new source of information about disease transmission, but current statistical methods are unable to incorporate these data into the analysis of transmission in households and other close-contact groups. The long-term goal is to develop statistical and epidemiologic methods that use high-resolution transmission data and genetic sequence data to inform rapid and effective public health responses to emerging infections. The goal of the proposed research is to develop flexible and robust regression models for infectious disease transmission data that can incorporate pathogen genetic sequences. These will be based on a recently-developed semiparametric regression model that can estimate parameters crucial to mathematical models of epidemics and the design of interventions, including hazard ratios for covariate effects on infectiousness and susceptibility and baseline hazards of transmission in infectious-susceptible pairs. To make it a more practical tool for infectious disease epidemiology, this model will be extended to account for external sources of infection, missing data, and small samples. The partial likelihood for this model is a sum over the set of transmission trees consistent with the epidemiologic data on person, place, and time. Since a phylogeny linking pathogen samples from infected individuals constrains the set of possible transmission trees, pathogen genetic sequence data can be combined with epidemiologic data to obtain more efficient estimates of transmission parameters. Epidemiologic and genetic data will be combined by developing algorithms to find the set of transmission trees simultaneously consistent with both. These algorithms will be incorporated into Markov chain Monte Carlo or sequential Monte Carlo estimation procedures that will account for missing data and phylogenetic uncertainty. These methods will serve as a theoretical basis for the development of efficient case-control and case-cohort study designs for outbreak investigations and vaccine trials. The proposed research is innovative because it synthesizes survival analysis and statistical genetics to analyze infectious disease transmission data. It is significant because it will improve the collection and analysis o data and the evaluation of interventions in epidemics, allowing more effective control of emerging infections.