Potential circRNA-Disease Association Prediction Using DeepWalk and Nonnegative Matrix Factorization.

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• Circular RNAs (circRNAs) are abundant noncoding RNAs in eukaryotes.
• circRNAs play a critical role in tumor progression.
• Understanding circRNA-disease associations is vital for disease research.

Purpose of the Study:

• To propose a new computational method, DWNMF, for predicting circRNA-disease associations.
• To leverage topological, functional, and semantic similarities for accurate predictions.
• To enhance the prediction of circRNA-disease correlations.

Main Methods:

• Utilized DeepWalk for node feature learning on the circRNA-disease association network.
• Integrated functional and semantic similarities with topological similarities.
• Employed improved weighted K-nearest neighbor (IWKNN) for network preprocessing.
• Applied nonnegative matrix factorization (NMF) with L2,1-norm and dual-graph regularization.

Main Results:

• The DWNMF method demonstrated high predictive performance in cross-validation.
• DWNMF outperformed existing state-of-the-art methods for circRNA-disease association prediction.
• The model successfully identified potential circRNA-disease relationships across multiple datasets.

Conclusions:

• DWNMF is an efficient and accurate tool for predicting circRNA-disease associations.
• The method provides valuable insights into the role of circRNAs in diseases.
• This approach can advance the study of circRNAs in disease mechanisms and therapeutic strategies.