Related Experiment Video
Updated: Aug 5, 2026
08:25
Identification of Circular RNAs using RNA Sequencing
Published on: November 14, 2019
HSimGCDA: A novel higher-order similarity graph representation learning method for identifying CircRNA-disease
Yang Li1, Lei Wang2, Zhu-Hong You3
1School of Mathematics and Statistics, Weinan Normal University, Weinan 714099, China.
Bioorganic Chemistry
|July 29, 2026
Summary
This study introduces HSimGCDA, a novel computational model that improves the prediction of circular RNA-disease associations (CDAs) by effectively integrating higher-order similarity information for better disease diagnosis and treatment strategies.
Area of Science:
- Bioinformatics
- Genomics
- Computational Biology
Background:
- Circular RNAs (circRNAs) play crucial roles in biological processes and disease.
- Predicting circRNA-disease associations (CDAs) is vital for disease diagnosis and treatment.
- Existing computational models struggle to capture complex network information for CDA prediction.
Purpose of the Study:
- To develop an enhanced graph convolutional network (HSimGCDA) for improved CDA prediction.
- To effectively integrate higher-order similarity and diverse data features of circRNAs and diseases.
- To provide a practical tool for exploring CDAs and aiding disease diagnosis and therapeutic development.
Main Methods:
- Construction of multi-similarity graph networks integrating circRNA and disease data.
- Implementation of a higher-order similarity strategy to aggregate remote neighbor information.
- Utilization of a multilayer perceptron for accurate inference of CDAs.
Main Results:
- HSimGCDA achieved 95.93% prediction accuracy and 98.74% AUC on the CircR2Disease dataset.
- Ablation experiments confirmed the effectiveness of the HSimGCDA model.
- Case studies identified novel circRNAs associated with gastric and breast cancers.
Conclusions:
- HSimGCDA significantly enhances the prediction of circRNA-disease associations.
- The model offers valuable insights for disease diagnosis and therapeutic development.
- HSimGCDA demonstrates strong potential as a practical tool in biomedical research.
