Systematic prediction of synergistic drug combinations through network-based deep learning framework
文献类型：期刊论文
| 作者 | Zhang, Jun; Chen, Shi-Long; Wang, Yong-Cui |
| 刊名 | EXPERT SYSTEMS WITH APPLICATIONS
 |
| 出版日期 | 2025 |
| 卷号 | 270 |
| 关键词 | Anticancer treatment Deep learning Drug-target relationship Graph attention network Synergistic combination |
| 英文摘要 | Drug combination therapy was an effective strategy to overcome anticancer resistance, typically resulting from intrinsic tumor heterogeneity. Computational methods enabled conducting high throughput screening and exploring synergistic drug combinations for a specific cancer type on a large scale. Current studies mainly focused on the model architecture and the representation of molecule with structural information, while previous work has demonstrated the crucial role of drug-target information in the molecular mode of action in combination therapies. Hence, a network-based framework, TAG-CP, was proposed here to identify pre-clinical synergistic drug combinations for cancer treatment by integrating drug-target relationships into compound representation with a graph attention mechanism. Particularly, the compounds were first regarded as nodes and connected if they share common targets. Then, the molecular representations of the compounds were learned by a modified attention-based graph neural network. Further, the compound-compound pair was represented through S-kernel to overcome the systematic issue and concatenated with the features of the cancer cell line (CCL) to be input for training a deep learning model, while the output was calculated using Bliss score from the experimental data. The independent tests and the insights from the prediction mechanism demonstrated that the TAG-CP model could effectively discern similar drug combinations and preferred to unveil combinations with functionally related targets. These results indicated that, besides the model complexity and the commonly input feature, efficiently integrating biological information such as drug-target interactions offered an alternative approach to improve prediction accuracy. |
| 源URL | [http://210.75.249.4/handle/363003/62549]  |
| 专题 | 西北高原生物研究所_中国科学院西北高原生物研究所 |
| 推荐引用方式 GB/T 7714 | Zhang, Jun,Chen, Shi-Long,Wang, Yong-Cui. Systematic prediction of synergistic drug combinations through network-based deep learning framework[J]. EXPERT SYSTEMS WITH APPLICATIONS,2025,270. |
| APA | Zhang, Jun,Chen, Shi-Long,&Wang, Yong-Cui.(2025).Systematic prediction of synergistic drug combinations through network-based deep learning framework.EXPERT SYSTEMS WITH APPLICATIONS,270. |
| MLA | Zhang, Jun,et al."Systematic prediction of synergistic drug combinations through network-based deep learning framework".EXPERT SYSTEMS WITH APPLICATIONS 270(2025). |
入库方式： OAI收割
来源：西北高原生物研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。