EUVIMEDEUVIMED entdecken
Open Access

DALI-Syn: integrating chemical LLMs with multi-level attention architectures for synergistic drug combination prediction

Huanhuan Du, Yayu Tian, Weiguang Wang, Bo Wei, Ting Mei, Hongbo Zhu, Zidan Yang, Xuan He iD

Health Information Science and Systems · 05.04.2026 · 14 (1)

DOI: 10.1007/s13755-026-00441-2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Health Information Science and Systems, online erschienen am 05.04.2026. Verfasst von Huanhuan Du und 7 weiteren Autor:innen. Beteiligte Einrichtung: College of Medicine and Biological Information Engineering, Northeastern University, Shenyang, 110819 Liaoning China. Die Arbeit nennt 54 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: National Natural Science Foundation of China, Fundamental Research Funds for the Central Universities.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Predicting drug synergy remains a critical challenge in personalized cancer treatment. Existing computational methods have made significant progress, offering the potential to accelerate the discovery of novel drug combinations. However, topology-based methods often fail to fully capture the deep chemical semantics within SMILES sequences. Moreover, models employing shallow feature fusion fail to effectively capture the complex interactions between drugs and cell lines.

Methods

We propose DALI-Syn, which integrates a chemical large language model and incorporates bidirectional drug-cell line interaction capabilities. The framework leverages ChemBERTa-2 to extract deep semantic representations from SMILES, transcending local topology. Simultaneously, Dynamic Multi-scale Convolution (DMC) and multi-head self-attention mechanisms refine drug and cell line features locally and globally. A symmetric cross-attention module then executes bidirectional feature fusion.

Results

Experiments on the O'Neil dataset demonstrate that DALI-Syn outperforms state-of-the-art methods, achieving an AUC of 0.970 and AUPR of 0.908. Validation on an independent AstraZeneca test set confirms its robust generalization capability.

Conclusions

DALI-Syn advances drug synergy prediction through enhanced chemical semantic understanding and deep feature interaction, positioning it as a reliable tool to accelerate combination therapy development.

Schlagwörter

Stichwörter der Autor:innen

Attention mechanismChemBERTa-2Drug synergy predictionFeature interaction fusion

Autor:innen und Institutionen

  • College of Medicine and Biological Information Engineering, Northeastern University, Shenyang, 110819 Liaoning China.
  • School of Computer Science and Engineering, Northeastern University, Shenyang, 110819 Liaoning China.
  • School of Computing, Newcastle University, Newcastle upon Tyne, UK.
  • School of Information Science and Engineering, Shenyang Ligong University, Shenyang, China.
  • School of Computer Science, Beijing University of Posts and Telecommunications, Beijing, China.

ORCID-Profile: Xuan He

Förderung

  • National Natural Science Foundation of China
  • Fundamental Research Funds for the Central Universities

Zitieren

APA

Du, H., Tian, Y., Wang, W., Wei, B., Mei, T., Zhu, H., Yang, Z., & He, X. (2026). DALI-Syn: integrating chemical LLMs with multi-level attention architectures for synergistic drug combination prediction. Health Information Science and Systems, 14(1). https://doi.org/10.1007/s13755-026-00441-2

Vancouver

Du H, Tian Y, Wang W, Wei B, Mei T, Zhu H, et al. DALI-Syn: integrating chemical LLMs with multi-level attention architectures for synergistic drug combination prediction. Health Information Science and Systems. 2026;14(1). doi:10.1007/s13755-026-00441-2

BibTeX
@article{Du2026,
  author = {Du, Huanhuan and Tian, Yayu and Wang, Weiguang and Wei, Bo and Mei, Ting and Zhu, Hongbo and Yang, Zidan and He, Xuan},
  title = {DALI-Syn: integrating chemical LLMs with multi-level attention architectures for synergistic drug combination prediction},
  journal = {Health Information Science and Systems},
  year = {2026},
  volume = {14},
  number = {1},
  doi = {10.1007/s13755-026-00441-2}
}
Kostenlos starten →