Adaptive multi-view learning method for enhanced drug repurposing using chemical-induced transcriptional profiles, knowledge graphs, and large language models

Abstract

This study introduces adaptive multi-view learning (AMVL), a novel methodology that integrates chemical-induced transcriptional profiles (CTPs), knowledge graph (KG) embeddings, and large language model (LLM) representations to enhance the predictive power of drug repurposing. Drug repurposing offers a promising alternative to traditional drug development by identifying new therapeutic uses for existing drugs, which significantly reduces cost and time. However, current methods often rely on limited data sources and simplistic assumptions, restricting their ability to capture the multifaceted nature of biological systems.

Highlights

• AMVL improves drug repurposing efficacy by leveraging multi-source similarities within a heterogeneous network.
• This study is the first to integrate chemical-induced transcriptional profiles, knowledge graph embeddings, and large language model representations for this purpose.
• On three benchmark datasets and one validation set, AMVL outperforms current state-of-the-art methods across multiple metrics, including AUC, AUPR, and F1.

Publication Details

• DOI: 10.1016/j.jpha.2025.101275
• Journal Impact Factor (IF): 6.1
• CAS Ranking: Q1
• Authors: Yudong Yan#, Yinqi Yang#, Zhuohao Tong, Yu Wang, Fan Yang, Zupeng Pan, Chuan Liu, Mingze Bai, Yongfang Xie, Yuefei Li, Kunxian Shu, Yinghong Li*

Recommended citation: Yan, Y., Yang, Y., Tong, Z., Wang, Y., Yang, F., Pan, Z., Liu, C., Bai, M., Xie, Y., Li, Y., Shu, K., & Li, Y. (2025). "Adaptive multi-view learning method for enhanced drug repurposing using chemical-induced transcriptional profiles, knowledge graphs, and large language models." Journal of Pharmaceutical Analysis, 15(6), 101275.
Download Paper