idenMD-NRF

idenMD-NRF identifies associations between microRNAs (miRNAs) and diseases to prioritize disease-linked miRNAs and elucidate pathogenic mechanisms of complex diseases.

Key Features:

• Ranking Framework: Conceptualizes miRNA-disease association identification as an information retrieval task to systematically evaluate and prioritize candidate associations.
• Learning to Rank algorithm: Applies a Learning to Rank machine learning algorithm to order diseases by likelihood of association with a query miRNA.
• Association features and predictors: Leverages high-level association features and multiple predictive inputs to inform the ranking process.
• Experimental validation: Demonstrated superior predictive performance relative to other predictors using two independent test datasets.

Scientific Applications:

• Mechanistic investigation: Predicts disease associations for novel miRNAs to support investigation of underlying molecular mechanisms.
• Biomarker identification: Prioritizes miRNAs as candidate diagnostic or prognostic biomarkers.
• Therapeutic target discovery: Identifies candidate miRNAs for exploration as therapeutic targets.

Methodology:

Formulates the problem as an information retrieval task and employs a Learning to Rank algorithm that uses high-level association features and various predictors, with performance evaluated on two independent test datasets.