DEXOM: Diversity-based enumeration of optimal context-specific metabolic networks

Area of Science:

• Systems biology
• Metabolic network reconstruction
• Multi-omics data integration

Background:

• Metabolic activity identification is complex due to post-transcriptional modifications and protein degradation.
• Gene expression alone is insufficient for accurate metabolic predictions.
• Genome-scale metabolic networks (GSMN) can be refined using multi-omics data for context-specific models.

Purpose of the Study:

• To address the limitation of multiple optimal sub-networks in context-specific metabolic modeling.
• To formalize and develop an approach for enumerating diverse optimal metabolic networks.
• To improve the predictive power and biological interpretability of metabolic models.

Main Methods:

• Developed DEXOM, a unified approach for diversity-based enumeration of context-specific metabolic networks.
• Implemented various enumeration strategies within DEXOM.
• Validated methods using simulated and real biological data.

Main Results:

• DEXOM successfully enumerates alternative optimal metabolic networks.
• Ensembles of metabolic networks improved in silico predictions of essential genes in Saccharomyces cerevisiae.
• Alternative solutions aided in detecting enriched pathways in human cancer cell lines.

Conclusions:

• Enumerating diverse metabolic network solutions enhances biological interpretation.
• DEXOM provides a robust framework for exploring metabolic states.
• The open-source DEXOM library facilitates broader application in systems biology research.