COVID-19 transmission in Alaska: geographic and genomic modeling

项目总结 疫情传播的有效建模取决于对这两个问题最重要的数学描述 疾病传播和进展的各个方面，以及从数据中估计参数。收集遗传基因 新冠肺炎大流行期间SARS-CoV-2样本的序列产生了大量新的数据 来源，以补充更传统的公共卫生。阿拉斯加的地理人口结构， 以具有定义良好的互连的子群为特征的，为使用此类数据提供了独特的机会 开发新的模型和参数估计方法。此外，除了地理上的异质性， 大流行的进展涉及冠状病毒变种流行率的变化以及缓解 策略，因此参数也具有时间依赖性。 该项目利用公共卫生和冠状病毒大流行的测序数据，通过三个主要 模型推力。变量分布的统计模型，它既能捕捉地理结构，又能捕捉变化 将开发不同的比例，并将其应用于阿拉斯加基因组数据集。传输估算 将使用系统动力学方法调查区域内和区域之间的速率。这包括开发 一种新的统计方法，用于快速估计随时间变化的参数，以及专注于 努力研究亚群之间的传播模式。最后，一个传统的隔室模型 将开发阿拉斯加的新冠肺炎，它捕捉到影响疾病动态的关键地理要素 在这个州。虽然最初是使用医院发病率和检测结果等公共卫生数据进行拟合，但它 然后将进行修改，以更好地反映前两次推进所产生的见解和估计。一个目标是 通过将测序数据整合到建模中来促进和展示改进的预测能力 输油管道。 尽管该项目侧重于阿拉斯加的建模和数据，但其方法和结论将会更多 适用范围广泛。人口结构与城市/郊区核心相似且较小的地区，主要 农村地区具有明确的互联互通，在欠发达国家很常见，但它们可能缺乏 有效建模所需的足够数据。对这种结构产生的疾病动力学的洞察也将得到改善 了解具有更广泛相互联系的区域的疫情发展的各个方面。

PROJECT SUMMARY Effective modeling of epidemic transmission depends on both a mathematical description of the most important aspects of disease transmission and progression, and estimation of parameters from data. Collection of genetic sequences of SARS-CoV-2 samples throughout the COVID-19 pandemic has produced an extensive new data source to supplement more traditional public health ones. The geographical population structure in Alaska, characterized by subpopulations with well-defined interconnections, offers a unique opportunity to use such data to develop new models and parameter estimation methods. Moreover, in addition to geographic heterogeneity, the progression of pandemic has involved changes in prevalence of coronavirus variants as well as mitigation strategies, so that parameters have a temporal dependence as well. This project leverages public health and sequencing data from the coronavirus pandemic through three main modeling thrusts. A statistical model of variant spread, which captures both geographic structure and changing variant proportions will be developed and applied to the Alaskan genomic data set. Estimation of transmission rates in and between regions using phylodynamic approaches will be investigated. This includes development of a new statistical approach for rapid estimation of parameters that change through time, as well as a focused effort to study transmission patterns between subpopulations. Finally, a traditional compartment model of COVID-19 in Alaska will be developed, which captures the key geographic features affecting disease dynamics in the state. While initially fitted using public health data such as hospital incidence rates and testing results, it will then be modified to better reflect insights and estimates emerging from the first two thrusts. A goal is to promote and demonstrate improved predictive ability from incorporation of sequencing data into the modeling pipeline. Although the project focuses on modeling and data from Alaska, its methods and conclusions will be more broadly applicable. Areas exhibiting similar population structure of an urban/suburban core and smaller, largely rural regions with well-defined interconnections, are common in less developed countries, but they may lack sufficient data for effective modeling. Insights into disease dynamics arising from this structure will also improve understanding of aspects of epidemic progression in regions with more diffuse interconnections.