Combining K Nearest Neighbor With Nonnegative Matrix Factorization for Predicting Circrna-Disease Associations

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Circular RNAs (circRNAs) are crucial regulators of gene expression.
• circRNAs are implicated in various complex human diseases.
• Experimental identification of circRNA-disease associations is resource-intensive.

Purpose of the Study:

• To develop a computational model for inferring circRNA-disease associations.
• To address the need for efficient and cost-effective methods in this field.

Main Methods:

• Proposed a novel method, KNN-NMF, combining K-nearest neighbors (KNN) and Nonnegative Matrix Factorization (NMF).
• Calculated Gaussian Interaction Profile (GIP) kernel similarity for circRNAs and diseases.
• Incorporated disease semantic similarity.
• Constructed new circRNA-disease interaction profiles using weighted KNN to mitigate false negatives.

Main Results:

• KNN-NMF demonstrated superior prediction performance compared to existing methods in five-fold cross-validation.
• Case studies confirmed KNN-NMF's efficacy in identifying potential circRNA-disease associations for common diseases.

Conclusions:

• KNN-NMF is an effective computational tool for predicting circRNA-disease associations.
• The method aids in understanding disease pathogenesis and facilitates diagnostic strategies.