多任务深度学习模型用于使用ToxCast生物测试的基于机制的发育和生殖毒性 (DART) 预测
Siyeol Ahn1, Hojun Jung2, Jinwon Hwang1
1School of Environmental Engineering, University of Seoul, Seoul, Republic of Korea.
Frontiers in toxicology
|February 19, 2026
概括
这项研究引入了一种新的深度学习框架,用于使用体外生物活性数据预测发育和生殖毒性 (DART). 基于机制的DGCL模型显示出卓越的性能,为DART评估提供了一种可靠的非动物替代方案.
科学领域:
- 毒理学和药理学 毒理学和药理学
- 计算生物学和生物信息学
- 在体外和替代测试方法.
背景情况:
- 对于发育和生殖性毒性 (DART) 的传统动物研究是资源密集的,在道德上具有挑战性.
- 新方法方法 (NAM) 对于推进DART评估至关重要.
- 将机械数据与计算模型集成为DART预测提供了一个有希望的途径.
研究的目的:
- 开发和评估一个基于机制的深度学习框架,用于预测DART.
- 将最先进的深度学习架构与传统机器学习算法的性能进行比较.
- 用外部参考化学数据验证开发模型的预测能力.
主要方法:
- 利用了23个与发育和生殖途径相关的ToxCast测定中的体外生物活性数据.
- 开发和微调了四个深度学习架构:DGCL,TransFoxMol,MolPath和MolFormer. 这些架构包括:
- 采用多任务学习框架来提高模型稳定性和性能,特别是在不平衡的数据集.
主要成果:
- 该DGCL深度学习模型显著超过了基线机器学习算法 (例如随机森林,XGBoost).
- 在DGCL的多任务学习扩展改善了有限的活跃数据的终点的性能和稳定性.
- 外部验证证明了平衡的预测性能 (F1 = 0.68),证实了模型的可靠性和通用性.
结论:
- 开发的基于机制的深度学习框架,特别是DGCL模型,为DART预测提供了可靠和准确的非动物方法.
- 这种方法有效地处理机理上不同的和不平衡的测试数据,解决传统方法的局限性.
- 该研究强调了将机械生物测试数据与深度学习相结合的潜力,用于DART评估中的监管应用.
更多相关视频
17:28Human Pluripotent Stem Cell Based Developmental Toxicity Assays for Chemical Safety Screening and Systems Biology Data Generation
Published on: June 17, 2015
13.2K
05:45Developmental Toxicity Assay Based on Real-Time Monitoring of Fibroblast Growth Factor Signal Disruption in Human Induced Pluripotent Stem Cells
Published on: October 10, 2025
568
相关概念视频
01:23Toxicity Testing in Animals
25
Toxicity tests in animals are grounded on two main assumptions: first, the effects observed in laboratory animals can be extrapolated to humans, especially when adjusted for body surface area; second, high-dose exposure in animals is essential to identify potential human hazards from lower doses. This is based on the quantal dose-response concept, which faces the challenge of extrapolating results from relatively few test animals to much larger human populations. For example, a 0.01% incidence...
25
01:21Toxicokinetics: Overview
35
Studies that assess how a drug is absorbed, distributed, metabolized, and excreted (ADME) at toxic doses are termed toxicokinetics. Understanding toxicokinetics helps predict adverse drug reactions (ADRs) and manage toxicity in humans.Toxicokinetics differs from pharmacokinetics mainly in the dose levels studied, with toxicokinetics focusing on higher toxic doses. The kinetics at these levels can be non-linear due to altered physiological processes. Toxicodynamics examines the relationship...
35