Weighted Network-Based Inference of Human MicroRNA-Disease Associations

The identification of disease microRNAs is vital for understanding the pathogenesis of diseases at the molecular level, and is critical for designing specific molecular tools for diagnosis, treatment and prevention. However, one major issue in microRNA studies is the lack of bioinformatics methods to accurately predict microRNA-disease associations. Herein, we proposed an approach to infer microRNA-disease associations based on a weighted network. We tested our method on benchmark dataset documented in the miR2Disease, a database system we developed previously for collecting experimentally verified microRNA-disease associations, and achieved an area under the ROC up to 0.80. The method described here presents a promising approach to infer new potential microRNA-disease associations, which will provide testable hypotheses to guide future biological experiments and contribute to the identification of true disease microRNAs.