Predicting miRNA-Disease Association Based on Modularity Preserving Heterogeneous Network Embedding.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• MicroRNAs (miRNAs) are crucial regulators of biological processes implicated in human diseases.
• Accurate prediction of miRNA-disease associations is vital for disease diagnosis, prevention, and therapeutic development.

Purpose of the Study:

• To propose a novel heterogeneous network embedding-based method for predicting miRNA-disease associations.
• To integrate diverse biological information and network structures for improved prediction accuracy.

Main Methods:

• Developed MDN-NMTF (Module-based Dynamic Neighborhood Non-negative Matrix Tri-Factorization), a heterogeneous network embedding approach.
• Constructed a network integrating disease similarity, miRNA similarity, and known miRNA-disease associations.
• Learned latent vector representations for miRNAs and diseases, considering module properties.

Main Results:

• MDN-NMTF effectively integrates diverse biological data for miRNA-disease association prediction.
• The method preserves heterogeneous network structural information and properties.
• An extended version, MDN-NMTF2, further improved prediction capabilities compared to existing methods across four databases.

Conclusions:

• MDN-NMTF offers a powerful tool for predicting miRNA-disease associations.
• The method demonstrates significant improvements over existing approaches, aiding in disease research and drug discovery.
• Integrating modular properties enhances the predictive performance for miRNA-disease associations.