MSGTrans: circRNA-Disease Association Prediction With Cluster-Based Negative Sampling and Multi-Scale Graph...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Circular RNAs (circRNAs) play crucial roles in biological processes and human diseases.
• Identifying circRNA-disease associations is vital for biomarker discovery and therapeutic development.
• Current wet-lab methods are time-consuming and costly, necessitating computational approaches.

Purpose of the Study:

• To develop an efficient computational method for predicting circRNA-disease associations.
• To address the challenge of imbalanced datasets in circRNA-disease association studies.
• To offer a reliable tool for identifying potential circRNA biomarkers for diseases.

Main Methods:

• Developed MSGTrans, a novel computational model for circRNA-disease association prediction.
• Integrated multi-source nodes and employed cluster-based negative sampling to handle data imbalance.
• Utilized multi-scale graph transformers and a cross-scale attention mechanism to capture node features.
• Employed LightGBM for the final prediction of circRNA-disease associations.

Main Results:

• MSGTrans demonstrated superior performance compared to existing state-of-the-art methods.
• The model effectively captured both local and global features of nodes in the network.
• Identified reliable circRNA candidates associated with specific human diseases.

Conclusions:

• MSGTrans provides an efficient and accurate computational approach for predicting circRNA-disease associations.
• The method offers a valuable tool for accelerating biomarker discovery and therapeutic strategy development.
• Highlights the potential of graph transformers and attention mechanisms in biological network analysis.