BRLA-DDI:一种新的药物相互作用提取框架
Zhu Yuan1, Shuailiang Zhang2, Zongjin Li3
1Department of Information Management, The National Police University for Criminal Justice, Baoding 071000, China.
本研究介绍了BRLA-DDI,这是一种用于药物相互作用 (DDI) 提取的新型模型,通过集成先进的深度学习技术来提高识别药物不良反应 (ADR) 的准确性.
科学领域:
- 生物医学信息学 生物医学信息学
- 自然语言处理自然语言处理.
- 机器学习 机器学习
背景情况:
- 药物相互作用 (DDI) 提取对于识别药物不良反应 (ADR) 至关重要.
- 现有的模型面临的挑战是复杂的句子和隐含的药物关系.
- 需要更准确和更普遍的DDI提取方法.
研究的目的:
- 介绍BRLA-DDI,一种用于增强DDI提取的新型模型.
- 提高DDI提取模型的精度和概括能力.
- 解决处理复杂和隐含药物相互作用的局限性.
主要方法:
- 该BRLA-DDI模型整合了BioBERT-LSTM用于特征提取和关系图卷积网络 (R-GCN) 的多头注意力.
- 一个创新的注意力损失机制将交叉损失与基于注意力的规范化相结合.
- 使用动态负采样策略来缓解零损失问题并提高稳定性.
主要成果:
- 在2013年DDI提取数据集上,BRLA-DDI取得了高性能,精度为87.68%,回忆率为88.06%,F1评分为87.87%.
- 该模型在外部TAC 2018数据集上表现出优越且稳健的性能,这表明其具有很强的概括性.
- 提出的方法显著优于现有的DDI提取方法.
结论:
- BRLA-DDI模型在药物相互作用提取方面取得了重大进展.
- 生物BERT-LSTM,R-GCN和注意力损失的协同集成提高了模型性能和通用性.
- 公开发布的代码和数据有助于进一步研究生物医学信息处理.
更多相关视频
07:02A Computerized Test Battery to Study Pharmacodynamic Effects on the Central Nervous System of Cholinergic Drugs in Early Phase Drug Development
Published on: February 11, 2019
07:40A Data Integration Workflow to Identify Drug Combinations Targeting Synthetic Lethal Interactions
Published on: May 27, 2021
相关概念视频
Pharmacokinetics: Drug–Drug Interactions
Pharmacokinetics: Drug–Food and Drug–Viral Interactions
Drug-Receptor Interactions
Several parameters, such as the drug's affinity for its receptor and its efficacy, which is its ability to activate the receptor, determine the drug's effect on the tissue....
Factors Affecting Protein-Drug Binding: Drug Interactions
Displacement interactions can have varying outcomes, ranging from toxicity to virtually...
Factors Affecting Renal Clearance: Drug Distribution and Drug Interactions
One important factor is the relationship between renal clearance and the apparent volume of distribution. Renal clearance tends to be inversely proportional to the apparent volume of distribution. Drugs with an extensive distribution volume or those...
Drug-Receptor Interaction: Antagonist
Antagonists can be classified as competitive or noncompetitive based on their...
