A novel collaborative filtering model for LncRNA-disease association prediction based on the Naïve Bayesian...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Discovering long non-coding RNA (lncRNA)-disease associations is challenging due to limited experimental data.
• Developing efficient computational models is crucial for identifying disease-related lncRNAs.

Purpose of the Study:

• To propose a novel computational model, CFNBC, for inferring potential lncRNA-disease associations.
• To enhance the prediction of lncRNA-disease links without solely relying on known miRNA-disease associations.

Main Methods:

• Constructed an lncRNA-miRNA-disease tripartite network integrating known associations.
• Applied item-based collaborative filtering to update the tripartite network.
• Utilized a Naïve Bayesian Classifier for predicting potential lncRNA-disease associations.

Main Results:

• CFNBC achieved a reliable Area Under the Curve (AUC) of 0.8576 in Leave-One-Out Cross Validation (LOOCV).
• The model demonstrated superior performance compared to existing state-of-the-art methods.
• Case studies on glioma, colorectal cancer, and gastric cancer confirmed CFNBC's excellent prediction capabilities.

Conclusions:

• CFNBC shows satisfactory prediction performance, making it a valuable addition to biomedical research.
• The model offers an efficient computational approach for identifying potential lncRNA-disease associations.