Natural Language Processing-Based Comprehensive Data Analysis for Interaction Between Chemicals and Drug Metabolism Network

药物代谢与药物的有效性和安全性密切相关，多种药物联合给药后的药物相互作用涉及药物代谢的抑制或诱导。为了预测药物相互作用，不仅需要全面了解化学品与单个药物代谢酶的相互作用，而且需要全面了解调节酶表达和活性的网络系统。在本研究中，我们开发了一种基于自然语言处理的文本挖掘技术，用于系统地收集化学品和药物代谢相关网络之间相互作用的信息。此外，根据收集到的信息，我们系统地进行了结构-活性关系分析。这些结果将为旨在开发更安全药物的药物设计和发现提供线索。

Drug metabolism is closely related to efficacy and safety of drugs, of which inhibition or induction is involved in drug-drug interaction following administration of multiple drugs. In order to predict drug-drug interactions, it is required to comprehensively understand not only interaction of chemicals with individual drug-metabolizing enzymes but that with a network system regulating expression and activity of the enzymes. In the present study, we developed a natural language processing-based text-mining technique for systematically collecting information on interaction between chemicals and a drug metabolism-related network. In addition, based on the information collected, we systematically performed structure-activity relationship analyses. These results will present a clue to drug design and discovery intended for development of safer drugs.