Scholarlink

Background <p>Drug-target affinity (DTA) prediction is crucial for drug discovery. While protein evolutionary features are vital for identifying conserved regions, current methods rely on computationally expensive multiple sequence alignment (MSA), creating a significant bottleneck. To address this, we propose MAFI-DTA, a novel model that circumvents the computational burden of MSA by deriving evolutionary context directly from sequences using the advanced protein language model, ESM-3. Furthermore, we introduced a multi-scale protein graph construction strategy based on varying numbers of residues, enabling the effective extraction of structural information across different scales through a neural network integrating graph neural networks, BiLSTM, and transformer modules.</p> Results <p>Experimental validation on multiple benchmark datasets demonstrates that MAFI-DTA achieves significantly better performance compared to existing approaches. This improved accuracy is attributed to the effective incorporation of both the predicted evolutionary information and the multi-scale graph representations. The source code is available at <a href="https://github.com/aliveadult/MAFI">https://github.com/aliveadult/MAFI</a>.</p> Conclusions <p>MAFI-DTA successfully captures crucial evolutionary and multi-scale structural context efficiently, overcoming the limitations of traditional MSA methods. This approach provides a high-performing tool that facilitates the study of drug–target interactions and accelerates the drug discovery process.</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

A novel method for drug-target affinity prediction by integrating predicted evolutionary information and multi-scale protein graphs

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月30日

查看全文
  • Mingjian Jiang,
  • Huaibin Hang,
  • Zihao Cui,
  • Junxiao Feng,
  • Weina Pang,
  • Teng Ma,
  • Zhi Zhang,
  • Yaping Fan,
  • Shunpeng Pang,
  • Wei Zhou,
  • Yuanyuan Zhang

摘要

Background

Drug-target affinity (DTA) prediction is crucial for drug discovery. While protein evolutionary features are vital for identifying conserved regions, current methods rely on computationally expensive multiple sequence alignment (MSA), creating a significant bottleneck. To address this, we propose MAFI-DTA, a novel model that circumvents the computational burden of MSA by deriving evolutionary context directly from sequences using the advanced protein language model, ESM-3. Furthermore, we introduced a multi-scale protein graph construction strategy based on varying numbers of residues, enabling the effective extraction of structural information across different scales through a neural network integrating graph neural networks, BiLSTM, and transformer modules.

Results

Experimental validation on multiple benchmark datasets demonstrates that MAFI-DTA achieves significantly better performance compared to existing approaches. This improved accuracy is attributed to the effective incorporation of both the predicted evolutionary information and the multi-scale graph representations. The source code is available at https://github.com/aliveadult/MAFI.

Conclusions

MAFI-DTA successfully captures crucial evolutionary and multi-scale structural context efficiently, overcoming the limitations of traditional MSA methods. This approach provides a high-performing tool that facilitates the study of drug–target interactions and accelerates the drug discovery process.