NSF Postdoctoral Fellowship in Biology FY 2022: Determining functional correlates to the emergence and maintenance of antibiotic resistance genes in host-associated microbiomes

该行动资助了美国国家科学基金会2022财年生物学博士后研究奖学金，研究基因组，环境和表型之间相互作用的生命规则的综合研究。该奖学金支持将以创新方式对生活规则领域作出贡献的研究员的研究和培训。微生物无处不在，参与大量的生物过程，如产生和降解抗生素，帮助宿主代谢。由于大多数微生物可能尚未被发现，对这些微生物的持续研究将有助于更好地了解微生物的生态重要性和技术潜力。微生物学的许多最新发现都依赖于DNA和RNA测序方法来识别新的微生物。虽然测序是一种很好的方法，但目前存在大量未使用的测序数据。这个项目将开发工具来帮助在微生物群落中恢复这些数据，提高我们对微生物在不同生态系统中所起作用的理解。该研究员还将参与多种培训工作，以提高生物学家从本科到博士后水平的计算素养。该项目的总体目标是改进基因注释，丰度估计和功能表征的宏基因组和亚转录组。该项目将首先扩展当前依赖组装图的方法，以实现元组测序数据的同源级注释和丰度估计。装配图包含元组库中的所有序列，因此不会遭受与当前分析技术相同的数据丢失。元组测序实验也经常缺乏能力，因为在给定的环境中只有少量的群落。迁移学习克服了这个问题，首先在一个大型微生物数据集上训练一个模型，然后再利用这个学习到的信息来预测一个较小数据集的模式。该项目将建立一个迁移学习模型，该模型将在公开可用的细菌分离RNA测序数据上进行训练，并将用于发现亚转录组的模式。这些方法将用于研究猪肠道微生物组的抗菌素耐药性。虽然抗菌素耐药性通常可以通过测序来检测，但抗菌素耐药性基因的发生可能被低估，因为这些序列在许多基因组中共享，这使得所得测序数据复杂。在这个项目中开发的工具将改善复杂序列的检测，如抗菌素耐药基因。该项目还将研究饲养环境和抗生素暴露等宿主性状在暴露于各种抗生素时如何影响猪肠道微生物组中抗菌素耐药性的发生和出现。本研究开发的工具将改善不同微生物生态系统的注释和功能表征，并在其他系统中使用元组测序数据产生预测价值。最后，本提案旨在利用开放科学原则来开发和传播研究产品，并通过培训努力促进进入数据密集型生物学领域，从而扩大科学参与。该奖项反映了美国国家科学基金会的法定使命，并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，被认为值得支持。

This action funds an NSF Postdoctoral Research Fellowship in Biology for FY 2022, Integrative Research Investigating the Rules of Life Governing Interactions Between Genomes, Environment and Phenotypes. The fellowship supports research and training of the fellow that will contribute to the area of Rules of Life in innovative ways. Microbes are everywhere, and participate in broad number of biological processes such as producing and degrading antibiotics, and aiding in host metabolism. As most microbes have not yet likely been discovered, continued investigation of these organisms will lead to better understanding of the ecological importance and technological potential of microbes. Many recent discoveries in microbiology have relied on DNA and RNA sequencing approaches to identify new microbes. While sequencing is a excellent approach, there exists a large amount of sequencing data that is currently being unused. This project will develop tools to help recover this data in microbial communities, improving our understanding of the role microbes play in different ecosystems. The fellow will also engage in multiple training efforts to improve computational literacy of biologists from the undergraduate to post doctoral level.The overarching goal of this project is to improve gene annotation, abundance estimation, and functional characterization of metagenomes and metatranscriptomes. This project will first extend current approaches that rely on assembly graphs to achieve ortholog-level annotation and abundance estimation in meta-omic sequencing data. Assembly graphs contain all the sequences in a meta-omic library and thus do not suffer the same data loss as current analysis techniques. Meta-omic sequencing experiments also frequently lack power given small numbers of profiled communities from a given environment. Transfer learning overcomes this problem by first training a model on a large microbial dataset, then re-using this learned information to predict patterns in a smaller dataset. This project will build a transfer learning model that will be trained on publicly available bacterial isolate RNA sequencing data and will be used to discover patterns in metatranscriptomes. These approaches will be used to study antimicrobial resistance in pig gut microbiomes. While antimicrobial resistance can often be detected through sequencing, the occurrence of antimicrobial resistance genes is likely underestimated because these sequences are shared across many genomes which makes the resultant sequencing data complex. The tools developed in this project will improve the detection of complex sequences like antimicrobial resistance genes. The project will also investigate how host traits like rearing environment and antibiotic exposure impact the occurrence and emergence of antimicrobial resistance in pig gut microbiomes when exposed to various antibiotics. The tools developed in this study will improve annotation and functional characterization of diverse microbial ecosystems, generating predictive value in other systems with meta-omic sequencing data. Lastly, this proposal aims to broaden participation in science using open science principles for the development and dissemination of research products, and through training endeavors that facilitate entry into the field of data-intensive biology.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.