IIBR Informatics: Innovative Software and Databases to Leverage RNA Polymerase as a Phylogenetic Marker in Metagenomic Data

提高我们对地球上生物多样性的程度和性质的认识是生物学的一个基本挑战。地球上最大的未表征的系统发育多样性储存库存在于细菌和微生物的神秘谱系中，通常被称为“微生物暗物质”，无法在实验室中培养，只能通过不依赖于培养的分子方法知道存在。了解这些群体的生理学、进化历史和环境影响是未来研究的主要前沿，这一领域的研究已经导致了地球化学、进化生物学和细胞生理学的重要发现。高通量DNA测序的进展以及宏基因组学领域的发展提供了前所未有的宏基因组数据量，现在可以挖掘这些数据来发现新的微生物谱系，但需要新的计算方法来利用这些“大数据”来实现这些生物学见解。该项目的重点是开发计算工具，通过分析RNA聚合酶（RNAP）的序列来评估宏基因组数据中的微生物多样性。RNAP是一种高分辨率的系统发育标记基因，可为细菌和微生物提供准确的分类学分配，并可用于识别和分类生物圈中的隐蔽微生物谱系。这项工作还将涉及对研究生和本科生进行计算生物学和生物信息学方面的培训，以及将这些生物信息学方法纳入大学课程。推进我们对地球上生物多样性的程度和性质的认识是生物学的一个基本目标，目前地球上最大的未表征的系统发育多样性库由细菌和微生物的神秘谱系组成，它们通常被称为“微生物暗物质”。高通量DNA测序和宏基因组学的最新进展提供了前所未有的大量“组学”数据，这些数据可用于评估这些谱系在环境中的多样性和分布，但依赖于小亚基rRNA基因的传统方法由于序列组装引起的复杂性而与这些数据不兼容。鉴于目前可用的宏基因组数据量巨大，迫切需要开发与当前宏基因组数据集兼容的用于分析微生物多样性的准确和标准化的生物信息学工作流程。这项工作的目标是整合研究和教学活动，重点是开发软件，用于宏基因组数据集中RNAP序列的预测和系统发育特征。这将通过以下方式解决：1）开发和验证开源软件，该软件使用自定义隐马尔可夫模型，基因邻域评估和机器学习算法从宏基因组数据准确预测RNAP基因，以及2）编译RNAP数据库，用于快速准确地对新RNAP序列进行系统发育分类。RNAP是宏基因组数据中最好的系统发育标记，因此这项工作将为评估微生物系统发育多样性和进化关系提供一个重要的框架。拟议工作的影响将通过在GitHub和弗吉尼亚理工大学图书馆开放获取软件和数据产品来最大化。该软件的实施将被整合到弗吉尼亚理工大学的研究生和本科生课程中，以增加学生的参与和参与。此外，这项研究将由博士生和本科生进行，其中一些人是代表性不足的少数民族。这些学生的专业发展将通过PI的密切指导和科学会议和职业建设研讨会的出席优先。这项研究的结果将在艾尔沃德实验室网站（www.aylwardlab.com）和GitHub网站（https：//github.com/faylward）上提供。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的知识价值和更广泛的影响审查标准进行评估来支持。

Advancing our knowledge of the extent and nature of biodiversity on Earth is a fundamental challenge in Biology. The largest uncharacterized reservoir of phylogenetic diversity on the planet resides in cryptic lineages of Bacteria and Archaea, often referred to as "microbial dark matter", that cannot be cultivated in the laboratory and are only known to exist from cultivation-independent molecular methods. Understanding the physiology, evolutionary history, and environmental impact of these groups is a major frontier for future research, and studies in this area have already led to important discoveries in biogeochemistry, evolutionary biology, and cellular physiology. Advances in high-throughput DNA sequencing together with developments in the field of metagenomics have provided an unprecedented amount of metagenomic data that can now be mined to discover novel microbial lineages, but new computational methods are required for leveraging this "big data" to achieve these biological insights. This project focuses on developing computational tools to assess microbial diversity in metagenomic data through analysis of sequences of RNA polymerase (RNAP). RNAP is a high-resolution phylogenetic marker gene that provides accurate taxonomic assignments for Bacteria and Archaea and can be leveraged to identify and classify cryptic microbial lineages in the biosphere. This work will also involve the training of graduate and undergraduate students in computational biology and bioinformatics as well as the integration of these bioinformatic approaches into university curricula. Advancing our knowledge of the extent and nature of biodiversity on Earth is a fundamental goal in biology, and currently the largest uncharacterized reservoir of phylogenetic diversity on the planet consists of cryptic lineages of Bacteria and Archaea that are often referred to as "microbial dark matter". Recent advances in high-throughput DNA sequencing and metagenomics provide an unprecedented amount of "'-omic" data that can be used to assess the diversity and distributions of these lineages in the environment, but traditional approaches relying on small subunit rRNA genes are not compatible with these data due to complications that arise from sequence assembly. Given the massive quantity of metagenomic data currently available, there is an urgent need for the development of accurate and standardized bioinformatic workflows for the analysis of microbial diversity that are compatible with current metagenomic datasets. The goal of this work is to integrate research and teaching activities focused on developing software for the prediction and phylogenetic characterization of RNAP sequences in metagenomic datasets. This will be addressed this by 1) development and validation of open source software that uses custom Hidden Markov Models, gene neighborhood assessment, and machine learning algorithms to accurately predict RNAP genes from metagenomic data, and 2) compilation of an RNAP database for the fast and accurate phylogenetic classification of novel RNAP sequences. RNAP is the best available phylogenetic marker to use in metagenomic data, and this work will therefore provide a critical framework for the assessment of microbial phylogenetic diversity and evolutionary relationships. The impact of the proposed work will be maximized through the open-access distribution of software and data products on GitHub and the Virginia Tech library. Implementation of the software will be integrated into graduate and undergraduate courses at Virginia Tech to increase student involvement and participation. Additionally, this research will be conducted by PhD and undergraduate students, several of whom are underrepresented minorities. The professional development of these students will be prioritized through close mentorship by the PIs and attendance of scientific conferences and career-building workshops. Results of this research will be provided on the Aylward Lab website (www.aylwardlab.com) and GitHub site (https://github.com/faylward)This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.