Statistical innovation to integrate sequences and phenotypes for scalable phylodynamic inference

项目摘要/摘要 本提案的目标是设计、开发和分发贝叶斯统计方法和软件 为了研究快速演变的病原体的历史和实时出现，如埃博拉、人类免疫- OdefiCency，在非典型肺炎、拉萨、fl-CoV-2、西尼罗河、黄热病和寨卡病毒中。这项提议充分利用了 新颖的可扩展数据集成，为我们应对大规模流行病和流行病提供装备，并帮助为行动提供信息- 有能力的公共卫生政策。我们的多学科团队拥有横跨统计思维、数据科学、 进化生物学和传染病利用先进的测序技术和高通量 生物实验，可以表征数千个病原体基因组，表型测量，生态 来自单一暴发的逻辑和临床信息。我们的主要创新有三个方面。首先，我们将发明 并实施可扩展的贝叶斯系统动力学技术，以集成表型测量和研究 它们与疾病传播相关的进化。第二，我们将培育生物丰富的进化模式 通过新的有效算法绘制和了解疾病进化中的异质性。三是大力发展 高维和混合型表型模型，用来连接一致的病毒基因型/表型变化 大规模并行计算。尽管没有竞争对手的软件来整合表型和序列数据 在这种规模下，我们将比较我们的数据集减少的模型的受限情况与当前最先进的情况 使用实数来评估这些限制注入的计算性能改进和偏差的方法 世界榜样。该提案将提供用于部署的低级工具箱库和用户友好型软件 涉及统计和医学领域中迅速扩大的一系列大规模问题。

PROJECT SUMMARY/ABSTRACT This proposal targets the design, development and distribution of Bayesian statistical methods and software to study the historical and real-time emergence of rapidly evolving pathogens, such as Ebola, human immun- odeficiency, influenza, Lassa, SARS-CoV-2, West Nile, yellow fever and Zika viruses. The proposal exploits novel scalable data integration to equip us for large-scale epidemics and pandemics and help inform action- able public health policy. Our multidisciplinary team carries expertise across statistical thinking, data science, evolutionary biology and infectious diseases to leverage advancing sequencing technology and high-throughput biological experimentation that can characterize 1000s of pathogen genomes, phenotype measurements, eco- logical and clinical information from a single outbreak. Our chief innovations are three-fold. First, we will invent and implement scalable Bayesian phylodynamic techniques to integrate phenotypic measurements and study their correlated evolution with disease spread. Second, we will foster biologically-rich evolutionary models to map and understand heterogeneity in disease evolution through new efficient algorithms. Third, we will develop high-dimensional and mixed-type phenotype models to link concerted viral genotype / phenotype changes using massively parallel computing. Although no competing software exists to integrate phenotype and sequence data at this scale, we will compare restricted cases of our models with reduced datasets to current state-of-the-art approaches to evaluate computational performance improvement and bias that these limitations inject using real- world examples. This proposal will deliver low-level toolbox libraries and user-friendly software for deployment across a rapidly expanding range of large-scale problems in statistics and medicine.