NSF Postdoctoral Fellowship in Biology: Identifying and Validating Missing Links in the Global Bat-Virus Network

这一行动资助了NSF 2023财年生物学博士后研究奖学金，扩大了生物学中代表性不足的群体的参与。该研究金支持一项针对该研究员的研究和培训计划，该计划将增加生物学中代表性不足的群体的参与。迫切需要确定未观察到的宿主-病毒关联，因为人类、家畜和野生动物中许多令人担忧的传染病是由病毒引起的。然而，由于物流成本高，有关已知关联的信息稀少，对预测这些相互作用提出了挑战。开发基于计算机的模型来识别可能的关联，可以通过集中实地研究工作来测试预测的相互作用来削减成本。该项目将使用预测模型来确定蝙蝠和病毒之间的联系，然后通过实地研究进行测试。研究结果将证实(或不证实)蝙蝠病毒的相关性，这将改善预测模型。该研究员将通过导师指导、一系列研究生院准备研讨会和社区外展，扩大未被充分代表的群体在生物学领域的参与。该研究员将结合机器学习和系统验证现场和实验室研究，以扩大全球蝙蝠病毒网络，并测试蝙蝠栖息生态相对于其他蝙蝠和病毒特征的重要性。蝙蝠之所以特别令人感兴趣，是因为已知它们携带许多具有人畜共患病潜力的病毒，导致病原体监测增加，并显示出种内对人为结构的栖息地偏好的差异，这可能会增加向人类或从人类溢出的机会。该研究员将使用链接预测模型来量化蝙蝠物种和病毒家族之间相互作用的可能性，然后测试这些预测，并评估在俄克拉荷马州和德克萨斯州采样的蝙蝠不同栖息地结构类型(人为或自然)的种内差异。创建模型预测、验证和改进的迭代循环通常是精确和自适应建模的一个被忽视的步骤。为了解决这一问题，链接预测模型将使用包含已验证预测的更新数据集再次运行，并将评估模型性能的变化。为了扩大代表不足的群体在生物学领域的参与，该研究员将指导学生，并在俄克拉荷马大学举办一系列关于本科生研究、研究生经历和职业探索的研讨会。此外，该研究员还将通过教育社区志愿者了解蝙蝠栖息生态、野生动物疾病生态学和/或城市哺乳动物来参与社区外展。该项目由生物科学局生物基础设施司和已建立的激励竞争性研究计划(EPSCoR)共同资助。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势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 2023, Broadening Participation of Groups Underrepresented in Biology. The Fellowship supports a research and training plan for the Fellow that will increase the participation of groups underrepresented in biology. There is a pressing need to identify unobserved host-virus associations, as many infectious diseases of concern in humans, domestic animals, and wildlife are caused by viruses. However, predicting these interactions is challenged by a sparsity of information on known associations due to high logistical costs. Developing computer-based models to identify likely associations can cut costs by focusing field research efforts to test predicted interactions. This project will use predictive models to identify associations between bats and viruses, which will then be tested by field research. The research results will provide confirmation (or not) of bat-virus associations that will improve the predictive model. The fellow will broaden participation of underrepresented groups in biology through mentorship, a series of graduate school preparation workshops, and community outreach.The fellow will combine machine learning and systematic validation field and laboratory studies to expand the global bat–virus network and test the importance of bat roosting ecology relative to other bat and virus traits. Bats are of particular interest because they are known to host many viruses of zoonotic potential, leading to increased pathogen surveillance, and show intraspecific variation in roost preference for anthropogenic structures that could increase spillover opportunities to or from humans. The fellow will quantify the probability of interactions between bat species and virus families with link prediction models and then test these predictions as well as assess intraspecific variation in associations across roost structure type (anthropogenic or natural) in bats sampled in Oklahoma and Texas. Creating an iterative loop of model prediction, validation, and improvement is often a neglected step to accurate and adaptive modeling. To address this, the link prediction model will be run again with an updated dataset containing the validated predictions, and changes in model performance will be evaluated. To broaden participation of underrepresented groups in biology, the fellow will mentor students and present a series of workshops about undergraduate research, the graduate student experience, and career exploration at the University of Oklahoma. Additionally, the fellow will engage in community outreach through education of community volunteers on bat roosting ecology, wildlife disease ecology, and/or urban mammals. This project is jointly funded by the Division of Biological Infrastructure in the Directorate for Biological Sciences, and the Established Program to Stimulate Competitive Research (EPSCoR).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.