NSF Convergence Accelerator - Track D: Data-Driven Disease Control and Prevention in Veterinary Health

NSF 融合加速器 - 轨道 D：兽医健康中数据驱动的疾病控制和预防

• 批准号: 2040680
• 负责人: Beatriz Martinez Lopez
• 金额: $ 94.49万
• 依托单位: University of California-Davis
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-15 至 2022-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2040680&HistoricalAwards=false
• 关键词: NSF; Convergence; Accelerator; Track; Data

The NSF Convergence Accelerator supports use-inspired, team-based, multidisciplinary efforts that address challenges of national importance and will produce deliverables of value to society in the near future. The broader impact and potential societal benefit of this Convergence Accelerator Phase I project is to improve animal production and health. Specifically, this project aims to address the maintenance of good swine health as a means of achieving high productivity and efficiency in the high production global pork market. The US is the world’s second-largest pork producer and a major player in the world pork market, ranking second as importer and exporter country (the export value for pork in 2019 was $7 billion). Thus, sustainability of this industry is important both as a food source as well as from an economic standpoint. Sustainability requires informed and timely decisions using scientific-based analytical tools and prediction models, based on reliable and current data to better manage swine health. However, in reality, while a vast amount of data has been collected in all steps of primary production -from farrowing and weaning to slaughter- such data is often incomplete, inconsistent and scattered at different stakeholders -producers, veterinary diagnostic labs, and veterinary clinics. Intertwined with this lack of integrated and high-quality data, there is also a dire need for effective artificial intelligence (AI) algorithms specifically designed to address key veterinary health challenges. The potential impact of this project extends far beyond swine health and could be used as a model across animal production and health in the US and globally not only assuring high quality food supply but also providing economic advantage. The team will build on their existing Disease BioPortal platform to facilitate the integration and sharing of key datasets and develop new data-driven models specifically adapted to animal health. Outcomes will not only have a direct beneficial impact in the swine industry saving producers millions of dollars yearly but also will significantly improve animal health/welfare, food safety and, ultimately, public health.The team proposes both data-centric and model-centric approaches with the following objectives: 1) development of a pipeline for effective multi-level data connection and integration, including animal diagnostics; pathogen genomes; animal genetics, production and trade; farm management practices including biosecurity and treatment protocols and environmental information, among others; 2)elaboration, implementation and validation of advanced bioinformatic pipelines (i.e., read-based and assembly based methods) and AI algorithms (i.e., cost-aware adaptive sampling and explainable machine learning models) to solve key problems in the swine industry, in particular, antimicrobial resistance (AMR) and swine influenza infections and; 3) expansion and adoption of the Disease BioPortal platform to facilitate data sharing and AI user-friendly usage and visualization by veterinarians, producers as well as other stakeholders and the general public. This effort will enable us to address critical challenges: early detection of infectious diseases and timely outbreak investigations; better understanding of the variability and spread patterns of pathogens within and between farms; identification of the main drivers contributing to AMR; cost-effectiveness of the surveillance, treatment, vaccination, and biosecuritystrategies implemented at animal-, farm- and system- level. While the work focuses on the swine industry and its two most imminent challenges, the proposed data pipeline, data integration, bioinformatics and, AI algorithm development can be applied to other livestock. Thus, this work has the potential to improve animal health and welfare and to secure the sustainability of US agriculture and food systems by providing data-driven decision tools that push the frontier of precision epidemiology.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.

NSF融合加速器支持以使用为灵感，以团队为基础，多学科的努力，以应对国家重要性的挑战，并将在不久的将来为社会提供有价值的成果。这个融合加速器第一阶段项目的更广泛的影响和潜在的社会效益是改善动物生产和健康。 具体而言，该项目旨在解决保持良好的猪健康作为在高产量全球猪肉市场实现高生产力和效率的手段。 美国是世界第二大猪肉生产国，也是世界猪肉市场的主要参与者，作为进口国和出口国排名第二（2019年猪肉出口额为70亿美元）。因此，这一行业的可持续性无论是作为食物来源还是从经济角度来看都很重要。可持续发展需要使用基于科学的分析工具和预测模型，根据可靠的最新数据做出明智和及时的决策，以更好地管理猪的健康。然而，在现实中，虽然在初级生产的所有步骤中收集了大量数据-从分娩和断奶到屠宰-但这些数据往往不完整，不一致，分散在不同的利益相关者-生产者，兽医诊断实验室和兽医诊所。除了缺乏综合和高质量的数据外，还迫切需要专门设计用于解决关键兽医健康挑战的有效人工智能（AI）算法。该项目的潜在影响远远超出了猪的健康，可以作为美国和全球动物生产和健康的典范，不仅确保高质量的食品供应，还提供经济优势。该团队将建立在现有的疾病BioPortal平台上，以促进关键数据集的整合和共享，并开发专门适用于动物健康的新数据驱动模型。结果不仅会对养猪业产生直接的有益影响，每年为生产者节省数百万美元，而且会显著改善动物健康/福利、食品安全，最终改善公共健康。该团队提出了以数据为中心和以模型为中心的方法，目标如下：1）开发一个有效的多级数据连接和集成管道，包括动物诊断;病原体基因组;动物遗传学、生产和贸易;农场管理实践，包括生物安全和处理协议以及环境信息等; 2）高级生物信息学管道的制定、实施和验证（即，基于读取和基于汇编的方法）和AI算法（即，成本感知的自适应采样和可解释的机器学习模型），以解决养猪业的关键问题，特别是抗菌素耐药性（AMR）和猪流感感染; 3）扩大和采用疾病BioPortal平台，以促进兽医，生产商以及其他利益相关者和公众的数据共享和人工智能用户友好的使用和可视化。这一努力将使我们能够应对以下关键挑战：传染病的早期发现和及时的疫情调查;更好地了解农场内和农场之间病原体的变异性和传播模式;确定导致AMR的主要驱动因素;在动物、农场和系统层面实施的监测、治疗、疫苗接种和生物安全战略的成本效益。虽然这项工作的重点是养猪业及其两个最紧迫的挑战，但拟议的数据管道、数据集成、生物信息学和人工智能算法开发可以应用于其他牲畜。 因此，这项工作有可能改善动物健康和福利，并通过提供数据驱动的决策工具，推动精确流行病学的前沿，以确保美国农业和食品系统的可持续性。该奖项反映了NSF的法定使命，并已被认为值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估的支持。