NetCore: a network propagation approach using node coreness.

Area of Science:

• Computational Biology
• Bioinformatics
• Network Science

Background:

• Identifying phenotype-genotype associations is crucial for understanding diseases.
• Existing network propagation methods often suffer from node degree bias.
• Prioritizing candidate genes and identifying disease modules remain challenging.

Purpose of the Study:

• To introduce NetCore, a novel network propagation approach addressing node degree bias.
• To enhance the identification of phenotype-genotype associations and network modules.
• To provide a robust tool for genomics data analysis.

Main Methods:

• Developed NetCore, a network propagation method utilizing node coreness.
• Implemented a random walk with restart procedure incorporating node coreness.
• Employed a semi-supervised approach for phenotype-associated network module identification.
• Utilized ConsensusPathDB for a high-confidence protein-protein interaction (PPI) network.

Main Results:

• NetCore demonstrated improved gene re-ranking compared to standard degree-based methods.
• Evaluated NetCore on 11 GWAS traits, showing superior performance via cross-validation.
• Successfully identified disease genes and modules for Schizophrenia GWAS and pan-cancer mutation data.
• Outperformed existing network propagation approaches in comparative analyses.

Conclusions:

• NetCore effectively mitigates node degree bias in PPI networks.
• The approach enhances the accuracy of gene prioritization and module identification.
• NetCore offers a valuable and versatile tool for diverse genomics applications.