**郑州大学团队创新AI工具，精准识别药物-靶标相互作用**
近日，郑州大学和电子科技大学的研究团队取得突破，推出全新药物-靶标相互作用识别工具MIDTI。该工具采用多视角相似性网络融合策略和深度交互式注意机制，大大提升了药物-靶标相互作用预测的准确性和效率。
在药物研发和重新定位过程中，准确识别药物与靶标之间的相互作用至关重要。过去虽然有许多计算模型被提出并应用于这一领域，但如何有效融合与药物和靶标相关的多视角相似性网络仍是研究的难点。郑州大学团队的这项研究，正为解决这一难题提供了新思路。
新方法MIDTI不仅考虑药物与靶标之间的直接相互作用，还融合了多种视角的相似性网络，确保了预测的全面性和准确性。此外，深度交互式注意机制的应用，使模型能够更好地捕捉药物和靶标之间的复杂关系。研究结果显示，MIDTI在DTI预测任务上的表现显著超越其他方法。
该研究成果以《Drug-target interaction predictions with multi-view similarity network fusion strategy and deep interactive attention mechanism》为题，于2024年6月6日正式发表在《Bioinform》杂志，标志着人工智能在药物研发领域的又一重要进展。
英语如下：
News Title: Zhengzhou University Team Develops New AI Tool: Accurate Prediction of Drug-Target Interaction
Keywords: AI Drug Recognition, Multi-view Network Integration, MIDTI Technology
News Content: **Zhengzhou University Team Innovates AI Tool for Accurate Identification of Drug-Target Interaction**
Recently, a research team from Zhengzhou University and the University of Electronic Science and Technology of China has made a breakthrough, launching a new tool for drug-target interaction identification called MIDTI. The tool adopts a multi-view similarity network fusion strategy and deep interactive attention mechanism, greatly improving the accuracy and efficiency of drug-target interaction prediction.
In the process of drug development and repositioning, accurately identifying the interaction between drugs and targets is crucial. Although many computational models have been proposed and applied in this field in the past, how to effectively integrate multi-view similarity networks related to drugs and targets remains a research challenge. The research by the Zhengzhou University team provides new ideas for solving this problem.
The new method MIDTI not only considers the direct interaction between drugs and targets but also integrates similarity networks from multiple perspectives, ensuring comprehensive and accurate predictions. In addition, the application of deep interactive attention mechanisms enables the model to better capture the complex relationships between drugs and targets. The results show that MIDTI significantly outperforms other methods in DTI prediction tasks.
The research findings were formally published in the journal Bioinform on June 6th, 2024 with the title “Drug-target interaction predictions with multi-view similarity network fusion strategy and deep interactive attention mechanism,” marking another important development in artificial intelligence in the field of drug development.
【来源】https://www.jiqizhixin.com/articles/2024-06-27-4
Views: 2