Maximal entropy random walk on heterogenous network for MIRNA-disease Association prediction

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• MicroRNAs (miRNAs) play crucial roles in human disease development.
• Predicting disease-miRNA associations is of increasing interest.
• Existing in silico models leverage publicly available miRNA databases.

Purpose of the Study:

• To develop a computational model for predicting potential disease-miRNA associations.
• To integrate diverse biological data into a heterogeneous network for enhanced prediction.
• To validate the model's performance through cross-validation and case studies.

Main Methods:

• Developed the Maximal Entropy Random Walk on heterogeneous network for MiRNA-disease Association prediction (MERWMDA) model.
• Constructed a heterogeneous network integrating known disease-miRNA associations, miRNA functional relations, and disease semantic relations.
• Applied Maximal Entropy Random Walk (MERW) on the network to identify potential associations.

Main Results:

• MERWMDA achieved high performance in cross-validation, with AUCs of 0.8966 (global) and 0.8491 (local).
• Case studies on Esophageal cancer and Prostate cancer showed high confirmation rates (94% and 88%) for top-ranked miRNAs.
• Validation on Lung cancer and Lymphoma demonstrated the model's robustness and predictive power for novel associations.

Conclusions:

• MERWMDA is an effective computational tool for predicting disease-miRNA associations.
• The model provides valuable candidate miRNAs for further experimental validation.
• This approach can aid in understanding disease mechanisms and identifying potential therapeutic targets.