MAGED: Multimodal attentive graph learning with gene expression dynamics on knowledge graphs for TCM target prediction
Clinical Snapshot
PICO Framework
| P — Population | Traditional Chinese Medicine (TCM) herbs and their documented therapeutic effects |
| I — Intervention | MAGED multimodal graph attention learning framework integrating knowledge graphs with gene expression dynamics |
| C — Comparator | Existing baseline methods for herb-target interaction prediction |
| O — Outcomes | Herb-Target Interaction (HTI) prediction accuracy, ranking performance, recall under cold-start scenarios, and interpretability of predictions |
Bottom Line
MAGED represents a significant computational advance in predicting herb-target interactions by integrating traditional Chinese medicine knowledge with modern molecular data. The 59.8% improvement in prediction accuracy and literature validation of 80% of top predictions in the case study suggest strong potential for drug discovery applications. However, this is a computational methodology study requiring further validation across diverse herbal compounds and clinical contexts. While promising for research applications, clinical translation would require extensive validation studies and regulatory consideration. Australian researchers working with complementary medicines could benefit from this approach, particularly given the TGA's emphasis on evidence-based evaluation of herbal products. The framework's ability to connect traditional therapeutic knowledge with molecular mechanisms addresses a key gap in herbal medicine research, though practical implementation requires significant computational expertise and resources.
Key Findings
P Value: Not reported
Effect Size: 59.8% improvement over baseline methods
Primary Outcome: Herb-Target Interaction prediction accuracy measured by HR@10
Nnt Or Sensitivity: 8/10 top predictions validated by literature in Scutellaria baicalensis case study
Confidence Interval: Not reported
Clinical Application
Requires computational resources and expertise in bioinformatics; applicable to research settings rather than direct clinical use Relevant for complementary medicine research in Australia, though TGA regulation of herbal medicines requires evidence-based approaches. Could support research into herbal products listed on ARTG. Researchers and clinicians working with traditional Chinese medicine and herbal therapeutics
Abstract
ETHNOPHARMACOLOGICAL RELEVANCE: Plants used in traditional medicine have long provided valuable clues for drug discovery, yet systematically connecting their documented therapeutic effects with modern understanding of biological mechanisms remains a key scientific question. Bridging the gap between traditional knowledge, clinical symptom profiles, and contemporary molecular-level regulation is essential for advancing the study of herbal medicine. OBJECTIVES: This study aims to develop a multimodal graph attention learning framework, named MAGED, which integrates knowledge graphs with gene expression dynamics to improve the accuracy and interpretability of Herb-Target Interaction (HTI) prediction. METHODS: MAGED incorporates multi-source heterogeneous data, including traditional Chinese medicine (TCM) properties, hierarchical clinical symptom information, and molecular-level biological data, such as gene expression dynamics under drug intervention and causal biological networks. A multimodal fusion encoder is employed to embed herb attributes into efficacy-related feature vectors, which are then combined with functional representations (e.g., normalized enrichment scores) to form a unified contextualized representation. Additionally, a hierarchical graph attention network is designed to integrate macroscopic symptom-gene associations with microscopic regulatory pathways, thereby establishing a coherent inference chain from herbal effects to biological targets and phenotypic symptoms. RESULTS: Systematic experimental evaluations demonstrated that MAGED significantly outperforms existing baseline methods across multiple evaluation metrics, achieving a 59.8 % improvement in HR@10. The model also exhibited superior ranking and recall performance under cold-start scenarios. In a case study on Scutellaria baicalensis, 8 out of the top 10 predicted targets were supported by existing literature, with functional or direct interaction evidence, and some targets were further validated through molecular docking. CONCLUSION: MAGED provides an accurate and interpretable framework for predicting herb-target interactions, effectively integrating traditional knowledge with modern molecular evidence. This approach shows strong potential to facilitate the discovery of herbal mechanisms and the identification of novel therapeutic targets.
References
- 1.Chen, F., Fang, S., Zhao, R., Han, X., Li, H., Xu, B., & Tang, Z. (2026). MAGED: Multimodal attentive graph learning with gene expression dynamics on knowledge graphs for TCM target prediction. Journal of Ethnopharmacology, 121218. https://doi.org/10.1016/j.jep.2026.121218
Related Research
Journal of chemical information and modeling
Reinforcement Learning-Driven Multiproperty Optimization in Molecular Design Using Multicontext Transcriptome Data
29 July 2026
Journal of chemical information and modeling
PegaPlus─Interactive Machine Learning by Human Observation for Efficient Clustering and Analysis of Structure-Activity Data
14 July 2026
Cell host & microbe
Mining the code of life for new antibiotics.
10 July 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service