SBIR Phase I: Data Integration and Analysis for the Pharmaceutical Industry

这个小型企业创新研究（SBIR）第一阶段项目旨在为生物医学研究人员提供新的基于网络的机制，以便对大量生物医学文献进行整合和有效分析。分析将包括检测候选活性物质（药物）、临床适应症（疾病）之间的重要联系，甚至与这些药物和疾病的学术专家/专家之间的联系。提供这种新颖的基于网络的机制需要研究实体解析、灵活的在线分析处理和受控的众包。实体解析是必要的，因为药物研究人员从多个来源获取信息，这些来源通常使用不同的术语和标识方案来标识相同的实体（例如，相同的物质）。受控的众包是必要的，因为人类和自动代理将合作创造高质量的信息资产，重要的是要清楚地追踪每条信息的提供者及其可信度。最后，在线分析处理技术在企业商业智能（例如销售分析）中非常成功，将需要扩展到分析生物医学文献的目的。该项目更广泛的影响/商业潜力是降低药物发现过程中的成本，并使有希望的新药发现成为可能。该项目可以影响在制药公司或学术组织中从事药物发现工作的研究人员和决策主管的效率。影响将体现在以下几个方面：1)减少研究人员在大量生物医学文献中搜索和浏览有趣的新兴观点所花费的时间，从而降低研究部门的成本。提出的生物医学文献分析的新方法将使研究人员能够获得、总结和观察最新科学事实的趋势，而无需一次大量浏览数千条引文。2)通过揭示药物/疾病/分子靶点之间的联系，使新药发现成为可能，这些联系可以为药物发现管道提供信息，否则会被忽视。鉴于该国面临的医疗成本不断上升，发现新药的影响可能是巨大的。

This Small Business Innovation Research (SBIR) Phase I project aims to provide novel web-based mechanisms to biomedical researchers for the integration and efficient analysis of the vast amount of biomedical literature. The analysis will include the detection of important connections between candidate active substances (drugs), clinical indications (diseases), and even connections to academic experts/specialists on such drugs and diseases. Delivering such novel web-based mechanisms requires research on entity resolution, flexible on-line analytical processing and controlled crowdsourcing. Entity resolution is needed because the pharmaceutical researcher draws his information from multiple sources, which often use different terminologies and identification schemes for the same entity (e.g., the same substance). Controlled crowdsourcing is needed because humans and automatic agents will cooperate in creating a quality information asset, where it is important that there is clear trace to who provided each piece of information and how credible it is. Finally, on-line analytical processing technologies, which have been very successful in corporate business intelligence (e.g., sales analysis) will need to be extended for the purpose of analyzing the biomedical literature. The broader impact/commercial potential of this project is reduced costs in the drug discovery process and enablement of promising new drug discoveries. The project can impact the effectiveness of researchers and decision making executives working on drug discovery either in pharmaceutical companies or in academic organizations. Impact will come in the following ways: 1) Reduction of costs in the research department by reducing the time that the researchers spend in searching and browsing the vast amount of biomedical literature for interesting emerging ideas. The proposed novel ways of biomedical literature analysis will allow the researcher to obtain, summarize and observe trends in the latest scientific facts, without copiously browsing thousands of citations at a time. 2) Enabling of new drug discovery by revealing drug/disease/molecular target connections that can feed the drug discovery pipeline and would otherwise be overlooked. The impact of discovering novel drugs can be huge given the rising medical costs that the country faces.