A Dual Ranking Algorithm Based on the Multiplex Network for Heterogeneous Complex Disease Analysis.

Area of Science:

• Biomedical research
• Computational biology
• Network science

Background:

• Identifying biomarkers for complex diseases is crucial but challenging.
• Existing network propagation methods often rely on single networks, limiting information capture.
• Integrating multiple networks risks data conflict and loss of individual network characteristics.

Purpose of the Study:

• To develop a robust framework for biomarker discovery in heterogeneous complex diseases.
• To address limitations of single-network and aggregated-network approaches.
• To identify biomarkers with high discrimination ability and biological interpretability for clinical applications.

Main Methods:

• Developed a multiplex network-based dual ranking framework (DualRank).
• Applied DualRank to heterogeneous complex disease analysis for diagnosis, prognosis, and classification.
• Evaluated biomarker identification performance based on quantity, prediction accuracy, and interpretability.

Main Results:

• DualRank outperformed existing competing methods in complex disease analysis.
• The framework successfully identified biomarkers with small quantity and strong predictive performance.
• Achieved an average Area Under the Curve (AUC) of 0.818, demonstrating high prediction accuracy.

Conclusions:

• DualRank provides an effective approach for biomarker discovery in heterogeneous complex diseases.
• The identified biomarkers possess biological interpretability and clinical relevance.
• This method enhances capabilities for disease diagnosis, prognosis, and classification.