Executable Network Models of Integrated Multiomics Data.

Area of Science:

• Computational biology
• Systems biology
• Bioinformatics

Background:

• Multiomics profiling offers a comprehensive view of biological conditions by capturing complex signaling events across molecular layers.
• Pathway enrichment analysis is crucial for understanding signaling mechanisms but faces challenges due to data variability in multiomics datasets.
• Integrating diverse omics data remains a computational challenge for a holistic understanding of biological systems.

Purpose of the Study:

• To introduce a novel Boolean network-based method, multiomics Boolean Omics Network Invariant-Time Analysis (mBONITA), for integrating multiomics data.
• To develop a computational approach that overcomes technical and biological variability in layered omics data.
• To identify consistently modulated genes and pathways across multiple molecular layers.

Main Methods:

• mBONITA utilizes prior knowledge networks for topology-based pathway analysis.
• The method integrates fold-changes, variance, node influence, and cross-dataset evidence to identify modulated genes.
• Boolean network modeling is employed to handle the complexity of multiomics data integration.

Main Results:

• mBONITA successfully integrated multiomics data from RAMOS B cells treated with cyclosporine A under varying oxygen tensions.
• The analysis identified pathways involved in hypoxia-mediated chemotaxis.
• Comparative analysis showed mBONITA outperforms other methods in identifying pathways modulated across all omics layers.

Conclusions:

• mBONITA provides a robust framework for integrating multiomics data, enabling a more accurate characterization of biological signaling.
• The method facilitates the discovery of key pathways in complex biological processes like hypoxia-mediated chemotaxis.
• mBONITA is a valuable tool for systems biology research and is publicly available.