Hierarchical Hypergraph Learning in Association- Weighted Heterogeneous Network for miRNA- Disease Association...

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• MicroRNAs (miRNAs) are crucial regulators in cellular processes and disease development.
• Existing computational methods for miRNA-disease association prediction often overlook crucial attribute information within associated edges.
• Accurate identification of miRNA-disease associations is vital for understanding disease mechanisms and developing targeted therapies.

Purpose of the Study:

• To propose a novel computational method, Hierarchical Hypergraph learning in Association-Weighted heterogeneous network for MiRNA-Disease association identification (HHAWMD).
• To enhance the prediction accuracy of miRNA-disease associations by fully exploring attribute information in heterogeneous networks.
• To develop a robust tool for identifying potential miRNA-disease relationships.

Main Methods:

• Adaptive fusion of multi-view similarities using channel attention.
• Construction of an association-weighted heterogeneous graph by assigning edge weights and attribute features.
• Generation of a hypergraph by extracting subgraphs and creating hyperedges between miRNA-disease node pairs.
• Application of a hierarchical hypergraph learning approach with node-aware and hyperedge-aware attention mechanisms.

Main Results:

• The HHAWMD method adaptively fuses similarity information and distinguishes relationship relevance based on expression levels and similarity data.
• The constructed association-weighted heterogeneous graph and subsequent hypergraph capture rich semantic information.
• Experimental results demonstrate superior performance of HHAWMD compared to existing methods in miRNA-disease association identification.

Conclusions:

• HHAWMD effectively leverages attribute information within associated edges in heterogeneous networks.
• The hierarchical hypergraph learning approach enhances the aggregation of semantic information for improved prediction accuracy.
• HHAWMD serves as a powerful and accurate tool for identifying novel miRNA-disease associations.