Jaylet Thomas, Jornod Florence, Capdet Quentin, Armant Olivier, Audouze Karine
Université Paris Cité, Inserm, HealthFex, Paris F-75006, France.
PSE-ENV/SERPEN/LECO, Institut de Radioprotection et de Sûreté Nucléaire (IRSN), Saint-Paul-Lez-Durance, France.
Bioinformatics. 2025 Jul 1;41(7). doi: 10.1093/bioinformatics/btaf381.
The Adverse Outcome Pathways (AOP) framework advances alternative toxicology by prioritizing the mechanisms underlying toxic effects. It organizes existing knowledge in a structured way, tracing the progression from the initial perturbation of a molecular event, caused by various stressors, through key events across different biological levels, ultimately leading to adverse outcomes that affect human health and ecosystems. However, the increasing volume of toxicological data presents a significant challenge for integrating all available knowledge effectively.
Text mining techniques, including natural language processing and graph-based approaches, provide powerful methods to analyze and integrate large, heterogeneous data sources. Within this framework, the AOP-helpFinder TM tool, accessible as a web server, was created to identify stressor-event and event-event relationships by automatically screening scientific literature in the PubMed database, facilitating the development of AOPs. The proposed new version introduces enhanced functionality by incorporating additional data sources, automatically annotating events from the literature with toxicological database information in a systems biology context. Users can now visualize results as interactive networks directly on the web server. With these advancements, AOP-helpFinder 3.0 offers a robust solution for integrative and predictive toxicology, as demonstrated in a case study exploring toxicological mechanisms associated with radon exposure.
AOP-helpFinder is available at https://aop-helpfinder-v3.u-paris-sciences.fr.
不良结局途径（AOP）框架通过对毒性效应背后的机制进行优先排序，推动了替代毒理学的发展。它以结构化的方式组织现有知识，追踪从各种应激源引起的分子事件的初始扰动开始，经过不同生物水平的关键事件，最终导致影响人类健康和生态系统的不良结局的进展。然而，毒理学数据量的不断增加给有效整合所有可用知识带来了重大挑战。
文本挖掘技术，包括自然语言处理和基于图的方法，提供了强大的方法来分析和整合大型、异构数据源。在此框架内，创建了可作为网络服务器访问的AOP-helpFinder TM工具，通过自动筛选PubMed数据库中的科学文献来识别应激源-事件和事件-事件关系，促进AOP的开发。提议的新版本通过纳入额外的数据源引入了增强功能，在系统生物学背景下用毒理学数据库信息自动注释文献中的事件。用户现在可以直接在网络服务器上以交互式网络的形式可视化结果。通过这些进展，AOP-helpFinder 3.0为综合和预测毒理学提供了一个强大的解决方案，如在一项探索与氡暴露相关的毒理学机制的案例研究中所示。
AOP-helpFinder可在https://aop-helpfinder-v3.u-paris-sciences.fr获得。
