Linking Text Mining and Data Mining for Biomedical Knowledge Discovery

描述（由申请人提供）： 系统集成正在成为21世纪生物学的驱动力。研究人员正在系统地解决基因功能和复杂的调控过程，通过研究不同组织水平的生物体，从基因组，转录组和蛋白质组到代谢组和相互作用组。为了充分实现这种高通量数据的价值，需要先进的生物信息学进行整合，挖掘，比较分析和功能解释。此外，随着越来越多的生物医学文献，现在电子，有一个迫切的需要和一个很好的机会，充分利用文本挖掘工具的知识提取。然而，尽管最近取得了进展，文本挖掘工具并没有被生物学家广泛使用。这种差距部分是由于文本挖掘和生物用户社区之间缺乏密切的互动。该应用程序的目标是开发一个数字研究基础设施，将文本挖掘与系统生物学背景下的数据挖掘联系起来，用于生物医学知识发现，特别关注系统在真实的世界科学应用中的实用性和可用性。基于我们已经开发的生物信息学框架，以及我们与生物医学研究界的密切互动，具体目标是：（i）整合现有的文本挖掘工具，以从科学文献中识别和提取蛋白质和网络信息，（ii）将文本挖掘和数据挖掘与组学数据整合和网络接口连接，以捕获和可视化网络知识，以及（iii）进行用户研究，开发科学用例，提供培训和推广，并向广大生物医学用户社区传播该系统。本文提出的数字信息资源将作为生物医学研究人员从文献和公共数据库中提供的大量信息中破译知识的有利环境，更好地了解生物和疾病过程，作为基本了解人类健康和疾病的关键。

DESCRIPTION (provided by applicant): Systems integration is becoming the driving force for the 21st century biology. Researchers are systematically tackling gene functions and complex regulatory processes by studying organisms at different levels of organization, from genomes, transcriptomes and proteomes to metabolomes and interactomes. To fully realize the value of such high-throughput data requires advanced bioinformatics for integration, mining, comparative analysis, and functional interpretation. Furthermore, with an ever-increasing volume of biomedical literature now available electronically, there is both a pressing need and a great opportunity to fully utilize text mining tools for knowledge extraction. However, despite recent advancements, text mining tools are not being broadly used by biologists. Such a gap is partly due to the lack of close interactions between the text mining and the biological user communities. The goal of this application is to develop a digital research infrastructure that links text mining with data mining in the systems biology context for biomedical knowledge discovery, with a special focus on the utility and usability of the system for real world scientific applications. Building upon the bioinformatics framework we have already developed, as well as our close interactions with the biomedical research community, the specific aims are to: (i) integrate existing text mining tools to identify and extract protein and network information from scientific literature, (ii) connect text mining and data mining with omics data integration and web interface to capture and visualize network knowledge, and (iii) conduct user studies, develop scientific use cases, provide training and outreach, and disseminate the system to the broad biomedical user community. The digital information resource proposed herein will serve as an enabling environment for biomedical researchers to decipher knowledge from a plethora of information available in the literature and public databases, gaining a better understanding of biological and disease processes as a key to the basic understanding of human health and disease.