HNMDA: heterogeneous network-based miRNA-disease association prediction

Area of Science:

• Biomedical Informatics
• Computational Biology
• Genomics

Background:

• MicroRNAs (miRNAs) are implicated in human disease development.
• Experimental methods for identifying miRNA-disease associations are costly and time-consuming.
• Computational approaches are needed for efficient prediction.

Purpose of the Study:

• To develop and validate a computational model for predicting latent miRNA-disease associations.
• To improve the accuracy and efficiency of miRNA-disease association prediction.

Main Methods:

• Proposed the Heterogeneous Network-based MiRNA-Disease Association (HNMDA) prediction model.
• Integrated known miRNA-disease associations, miRNA functional similarity, disease semantic similarity, and Gaussian interaction profile kernel similarity.
• Employed a heterogeneous network-based method with random walk with restart and optimal projection.

Main Results:

• Achieved an Area Under the Curve (AUC) of 0.8394 in cross-validation, outperforming previous methods.
• Case studies showed high confirmation rates: 82% for breast neoplasms, 76% for esophageal neoplasms, and 84% for kidney neoplasms.
• Successfully predicted novel miRNA-disease associations later confirmed by experimental reports.

Conclusions:

• HNMDA is an effective computational tool for predicting miRNA-disease associations.
• The model offers a promising alternative to experimental methods for identifying disease-related miRNAs.
• The approach demonstrates potential for discovering new therapeutic targets and understanding disease mechanisms.