CCF-BSF: AF: Small: Collaborative Research: Algorithmic Techniques for Inferring Transmission Networks from Noisy Sequencing Data

许多病毒用RNA编码其基因组，并在宿主体内表现出高度的基因组多样性。测序技术的进步使得在全球范围内跟踪病毒传播和及时发现疫情成为可能。该项目的目标是开发一套全面的预测数学模型和精确的计算方法，用于综合分析新兴分子监测程序产生的大量流行病学和测序数据集。研究成果将通过期刊出版物和在国际会议上的报告广泛传播，包括由pi组织的分子流行病学计算进展讲习班。开发算法的原型实现将作为开源软件包分发，并纳入疾病预防控制中心开发的基于云的全球肝炎暴发和监测工具包（GHOST）。该项目将为促进妇女和代表性不足的群体参与GSU、UCONN、Georgia Tech和特拉维夫大学的生物信息学和分子流行病学研究提供充足的机会。该项目的一个重要方面是将计算机科学和计算生物学的核心概念和思想传播给广泛的目标受众，包括：(1)在非正式环境中向初高中学生教授计算机科学，以及(2)将计算思维纳入大学本科阶段的生命科学课程。拟议的研究和教育活动将利用一个由计算机科学家、数学家和分子流行病学家组成的跨学科团队的广泛专业知识，为分子流行病学的关键问题开发精确的数学模型和计算方法，包括从容易出错的汇集测序数据中反卷积和推断病毒变异，推断病毒样本和传播网络之间的相关性，传输事件时间和网络参数的推断，以及传输网络动力学的预测建模。该团队将在CDC生成的大量分子监测数据集上进行广泛的算法验证，并开发健壮的原型软件实现。

Many viruses encode their genome in RNA and exhibit high genomic diversity within their hosts. Advances in sequencing technologies have made it feasible to track viral transmissions and timely detect outbreaks on a global scale. The goal of this project is to develop a comprehensive set of predictive mathematical models and accurate computational methods for integrated analysis of the massive epidemiological and sequencing datasets generated by emerging molecular surveillance programs. Research results will be broadly disseminated via journal publications and presentations at international conferences, including the Workshop on Computational Advances in Molecular Epidemiology organized by the PIs. Prototype implementations of developed algorithms will be distributed as open-source packages and incorporated in the cloud-based Global Hepatitis Outbreak and Surveillance Toolkit (GHOST) developed at CDC. The project will provide ample opportunities for promoting participation of women and underrepresented groups in bioinformatics and molecular epidemiology research at GSU, UCONN, Georgia Tech and Tel Aviv University. An important aspect of the project is to disseminate core concepts and ideas from Computer Science and Computational Biology to wide target audiences including: (1) teaching Computer Science, in an informal setting, to middle and high school students, and (2) incorporating computational thinking into Life Science curriculum at the undergraduate university level. The proposed research and education activities will leverage the extensive expertise of an interdisciplinary team comprised of computer scientists, mathematicians, and molecular epidemiologists to develop accurate mathematical models and computational methods for key problems in molecular epidemiology including deconvolution and inference of viral variants from error-prone pooled sequencing data, inference of relatedness between viral samples and transmission networks, inference of transmission event times and network parameters, as well as predictive modeling of transmission network dynamics. The team will carry out extensive algorithm validation on massive molecular surveillance datasets generated at CDC and develop robust prototype software implementations.