Scholarlink

Background <p>Drug–drug interactions (DDIs) frequently occur in combination therapy and may cause adverse effects or reduced efficacy. Existing computational approaches often fail to capture both the semantic information in drug sequences and the structural properties of drug molecules, limiting predictive power.</p> Results <p>We propose MDG-DDI, a deep learning framework that integrates a Frequent Consecutive Subsequence (FCS)-based Transformer encoder with a Deep Graph Network (DGN) to extract complementary semantic and structural features. These representations are fused and fed into a Graph Convolutional Network (GCN) for DDI prediction. Experiments on three benchmark datasets under transductive and inductive settings show that MDG-DDI consistently outperforms state-of-the-art methods, with particularly strong gains when predicting interactions involving unseen drugs.</p> Conclusion <p>By jointly modeling substructure-level semantics and molecular graph structure, MDG-DDI achieves robust and accurate DDI prediction. The framework demonstrates improved generalization and offers potential for enhancing drug safety assessment and discovery.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

MDG-DDI: multi-feature drug graph for drug-drug interaction prediction

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月30日

查看全文
  • Wenjun Li,
  • Yiting Zhou,
  • Wanjun Ma,
  • Weijun Liang,
  • Xiwei Tang

摘要

Background

Drug–drug interactions (DDIs) frequently occur in combination therapy and may cause adverse effects or reduced efficacy. Existing computational approaches often fail to capture both the semantic information in drug sequences and the structural properties of drug molecules, limiting predictive power.

Results

We propose MDG-DDI, a deep learning framework that integrates a Frequent Consecutive Subsequence (FCS)-based Transformer encoder with a Deep Graph Network (DGN) to extract complementary semantic and structural features. These representations are fused and fed into a Graph Convolutional Network (GCN) for DDI prediction. Experiments on three benchmark datasets under transductive and inductive settings show that MDG-DDI consistently outperforms state-of-the-art methods, with particularly strong gains when predicting interactions involving unseen drugs.

Conclusion

By jointly modeling substructure-level semantics and molecular graph structure, MDG-DDI achieves robust and accurate DDI prediction. The framework demonstrates improved generalization and offers potential for enhancing drug safety assessment and discovery.