RAPID: Exploring Causes and Cures for COVID-19 through Improved Access to Biomedical Research

解决新冠肺炎大流行和未来的危机，需要迅速和全面地获得相关研究知识。该项目通过增强国家理解相关全球研究文献和确定与抗击病毒相关的不同领域的发现的能力，促进科学进步，以促进国民健康和福利。这篇相关文献包括数以万计的生物医学文章，而且每周都在爆炸性地增长。该项目将提供获取重要发现的手段，并开发工具来发现跨学科以及新冠肺炎的病因、生物标志物、病情和治疗之间的联系。该项目将在一个开放的网站上提供每两周一次的关于新冠肺炎的生物医学文献的关键发现摘要。这些信息将按主题、国家和组织进行排列，以便研究人员和临床医生可以轻松获取。该项目还将发现跨研究领域的联系，可以确定新的治疗方法。这两种方法都努力将不同的研究结果和创新发展联系起来。第一种方法描述直接相关的研究文献，使研究结果更容易获得，并允许研究人员发现互补的知识。基于文献的发现方法，使用机器学习和相关方法，有助于弥合不同的研究结果。该项目将提供一系列研究知识，以帮助开发预防和治疗冠状病毒的方法。基于文献的发现方法用于提取目标领域的关键成分，然后探索其他不同领域的潜在原因、重要生物机制和可以重新定位的治疗方法。这一方法试图在对其他病毒的研究中发现以前未被认识到的共性，这些病毒影响易受目标病毒影响的关键生理系统(例如呼吸和免疫系统)。这种共性，例如不同病毒影响特定生物系统的方式，可能为认识共病或开发新的治疗方法打开大门。该项目将促进对文本挖掘和科学发现的理解。这项工作将使用人类评委在多个维度(如病毒类型、所述生理机制、原因、药物类型和其他治疗方式)对适度数量的摘要记录进行分类，然后使用软件对30,000篇和不断增长的新冠肺炎相关研究文章进行自动分类。从这种分类中出现的线索可能会打开进入其他生物医学研究和临床研究的窗口，以寻求新的方法。通过立即公开这些数据，该项目将服务于在该地区进行研究和实践的社区，并加速围绕冠状病毒的科学发现。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。

Resolving the COVID-19 pandemic, and future crises, demands rapid and comprehensive access to pertinent research knowledge. This project promotes the progress of science to advance national health and welfare by enhancing the nation’s capability to understand the related global research literature and to identify findings in disparate fields that are relevant for combating the virus. This body of relevant literature comprises tens of thousands of biomedical articles and is growing explosively each week. This project will provide means to access vital findings and to develop tools to discover connections across disciplines and among causes, biomarkers, conditions, and treatments for COVID-19. The project will provide bi-weekly summaries of key findings in the biomedical literature pertaining to COVID-19 on an open website. The information will be arranged by topic, country, and organization, so that it can be easily accessed by researchers and clinicians. The project will also uncover connections across research domains that could identify novel treatments. Both approaches strive to connect disparate research findings and innovative developments. The first approach profiles the directly relevant research literature to make findings more accessible and allow researchers to discover complementary knowledge. Literature Based Discovery methods, using machine learning and related methods, help bridge distinct research findings. The project will make available a spectrum of research knowledge to help develop prevention and treatment for corona viruses. Literature Based Discovery methods serve to extract key components of a target domain and then explore other distinct domains for potential causes, vital biomechanisms, and treatments that could be repurposed. This approach seeks to discover previously unrecognized commonalities in research on other viruses that affect key physiological systems susceptible to the target viruses (for example respiratory and immune systems). Such commonalities, such as in ways that different viruses affect particular biosystems, may open doors to recognition of co-morbidities or development of novel treatments. The project will advance understanding of text mining and scientific discovery. The work will employ human judges to classify a moderate number of abstract records on multiple dimensions (such as virus type, physiological mechanisms addressed, causes, drug types, and other treatment modalities), then use software to auto-classify the 30,000 and growing COVID-19-related research articles. Clues emerging from such classification could open windows into other biomedical research and clinical studies to seek novel approaches. By making these data immediately and publicly available, the project will serve the community of those researching and practicing in the area and accelerate scientific discovery around the coronavirus.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.