Gene relevance based on multiple evidences in complex networks.

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Multi-omics data analysis presents challenges in reconstructing comprehensive molecular disease mechanisms.
• Network-based methods are valuable for analyzing complex molecular interactions but often limited to specific omics combinations and layers.

Purpose of the Study:

• To investigate network diffusion for quantifying gene relevance using multiple molecular evidence layers.
• To introduce a novel gene scoring method, mND, for enhanced multi-omics data analysis.

Main Methods:

• Applied network diffusion to quantify gene relevance based on network proximity to altered genes across multiple data layers.
• Developed a gene scoring metric (mND) that integrates network information and multi-omics evidence.

Main Results:

• The mND score demonstrates superior performance compared to existing methods in identifying genes within network proximity across one or more layers.
• The method shows good performance in recovering known cancer-associated genes.
• The mND pipeline is versatile, accommodating multi-omics datasets, stratified data (e.g., single-cell clusters), or combined scenarios.

Conclusions:

• Network diffusion offers a powerful framework for integrated multi-omics analysis.
• The mND method provides a robust and broadly applicable tool for gene relevance quantification in complex biological systems and disease research.