相似性(几何)
风险评估
计算机科学
数据挖掘
数据科学
风险分析(工程)
人工智能
医学
计算机安全
图像(数学)
作者
Julia E. Rager,Cynthia V. Rider
标识
DOI:10.1016/j.cotox.2023.100417
摘要
Human health risk assessments for complex mixtures can address real-world exposures and protect public health. While risk assessors typically prefer whole mixture approaches over component-based approaches, data from the precise exposure of interest are often unavailable and surrogate data from a sufficiently similar mixture(s) are required. This review describes recent advances in determining sufficient similarity of whole, complex mixtures spanning the comparison of chemical features, bioactivity profiles, and statistical evaluation to determine “thresholds of similarity.” Case studies, including water disinfection byproducts, botanical ingredients, and wildfire emissions, are used to highlight tools and methods. Limitations to application of sufficient similarity in risk-based decision making are reviewed and recommendations presented for developing best practice guidelines.
科研通智能强力驱动
Strongly Powered by AbleSci AI