Using empirical biological knowledge to infer regulatory networks from multi-omics data

Area of Science:

• Systems Biology
• Genomics
• Computational Biology

Background:

• Multi-omics data integration is crucial for understanding complex diseases and developing personalized therapeutics.
• Existing tools often lack comprehensive frameworks for genome-level regulatory network inference incorporating prior knowledge.

Purpose of the Study:

• To present IntOMICS, an efficient integrative framework for multi-omics data analysis and regulatory network inference.
• To address the need for a framework that incorporates prior biological knowledge and handles diverse omics data types.

Main Methods:

• IntOMICS utilizes Bayesian networks to systematically analyze gene expression, DNA methylation, and copy number variation data.
• It incorporates both explicit prior biological knowledge and empirically derived knowledge from experimental data.
• The framework infers regulatory networks, providing insights into genetic information flow.

Main Results:

• IntOMICS demonstrates increased accuracy in regulatory network inference compared to single-modality algorithms.
• The framework successfully captures crosstalks between multi-omics data, validated in colon cancer samples.
• Performance comparison showed IntOMICS competitive with other multi-omics inference algorithms incorporating prior knowledge.
• Potential predictive biomarkers were identified in microsatellite stable stage III colon cancer.

Conclusions:

• IntOMICS offers a novel approach to biological knowledge discovery through multi-omics data integration.
• It serves as a powerful resource for systems biology research.
• The framework provides valuable insights into biological mechanisms relevant to personalized medicine.