GATv2EPI: Predicting Enhancer–Promoter Interactions with a Dynamic Graph Attention Network
Author Information
Author(s): Zhang Tianjiao, Zhao Xingjie, Sun Hao, Gao Bo, Liu Xiaoqi, Lonardi Stefano, Qin Zhaohui S.
Primary Institution: Northeast Forestry University
Hypothesis
Can a novel framework using dynamic graph attention networks improve the prediction of enhancer-promoter interactions?
Conclusion
GATv2EPI enhances EPI prediction accuracy by capturing complex topological structure information from gene regulatory networks through graph neural networks.
Supporting Evidence
- GATv2EPI demonstrated superior EPI recognition accuracy compared to existing similar methods.
- The model showed a training time improvement of 95.29% over TransEPI.
- Results emphasize the importance of epigenetic features surrounding enhancers and promoters in EPI prediction.
Takeaway
This study created a new model to better understand how parts of DNA called enhancers and promoters work together to control gene activity.
Methodology
The study used a novel EPI prediction framework based on dynamic graph attention networks, leveraging epigenetic information and a connectivity-based sampling method for dataset partitioning.
Limitations
The model requires considerable computational resources when dealing with particularly large datasets.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website