A statistical physics approach to integrating multi-omics data for disease-module detection.

Area of Science:

• Computational biology
• Systems biology
• Genomics

Background:

• Genes involved in the same disease often interact within molecular networks, forming disease modules.
• Network-based methods are advancing the understanding of human disease molecular underpinnings.
• Current computational methods for disease module extraction often lack multi-omics integration capabilities.

Purpose of the Study:

• To develop a novel computational approach for disease module detection using multi-omics data.
• To address the gap in methods that leverage multiple omics profiles for identifying context-dependent disease modules.

Main Methods:

• Developed a statistical physics approach based on the random-field O(n) model (RFOnM).
• Applied RFOnM to integrate gene-expression data with genome-wide association studies (GWAS) or mRNA with DNA methylation data.
• Utilized the human interactome network in conjunction with multi-omics profiles for several complex diseases.

Main Results:

• The RFOnM approach successfully integrated diverse omics data (gene expression, GWAS, DNA methylation) with the human interactome.
• Demonstrated that RFOnM outperforms existing single-omics methods in identifying disease modules across multiple complex diseases.
• Identified context-dependent disease modules by leveraging the power of multi-omics integration.

Conclusions:

• The developed RFOnM approach offers a robust framework for multi-omics data integration in disease module detection.
• This method enhances the elucidation of complex disease molecular mechanisms by capturing inter-omic relationships.
• The findings highlight the potential of statistical physics models in advancing network-based disease research.