PrGeFNE: Predicting disease-related genes by fast network embedding.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Identifying disease-related genes is crucial for understanding disease mechanisms, diagnosis, and treatment.
• Integrating multi-source biological data remains a challenge for improving disease-gene prediction.

Purpose of the Study:

• To propose a novel method, PrGeFNE (fast network embedding), for predicting disease-related genes.
• To effectively integrate multiple association types (phenotype-disease, disease-gene, protein-protein, gene-GO) for enhanced prediction.
• To provide a web tool for researchers to identify candidate disease genes and perform enrichment analysis.

Main Methods:

• Constructed a heterogeneous network incorporating phenotype-disease, disease-gene, protein-protein, and gene-GO associations.
• Extracted low-dimensional node representations using a fast network embedding algorithm.
• Reconstructed a dual-layer heterogeneous network and applied network propagation for disease-gene prediction.

Main Results:

• Demonstrated the significant role of various association data types in enhancing disease-gene prediction.
• Confirmed the superior performance of PrGeFNE compared to state-of-the-art algorithms through cross-validation and newly added-association validation.
• Developed a user-friendly web tool for accessing predicted candidate genes and performing enrichment analyses.

Conclusions:

• PrGeFNE effectively integrates multi-source biological data for accurate disease-gene prediction.
• The method and associated web tool can accelerate research into disease molecular mechanisms and experimental validation.
• This work highlights the importance of network embedding and heterogeneous data integration in bioinformatics.