Computational methods to identify metabolic sub-networks based on metabolomic profiles

Area of Science:

• Bioinformatics
• Systems Biology
• Metabolomics

Background:

• Untargeted metabolomics identifies concentration changes in biological samples under different conditions.
• Metabolomic profiles indicate perturbations but don't fully elucidate biochemical mechanisms.
• Interpreting complex metabolomic data requires integrating it with comprehensive metabolic knowledge.

Purpose of the Study:

• To review and present the primary graph-based computational approaches for interpreting metabolomic data using metabolic networks.
• To discuss the advantages, disadvantages, and parameter impacts of various graph algorithms applied to metabolic network analysis.
• To provide guidelines for effective sub-network extraction and suggest applications for these methods.

Main Methods:

• Utilizes graph theory algorithms to mine metabolic networks.
• Focuses on algorithms designed to extract sub-networks relevant to identified compounds from metabolomic profiles.
• Reviews existing adaptations of graph algorithms for improved biological interpretability.

Main Results:

• Highlights the current lack of consensus on the analysis of metabolic networks using graph approaches.
• Details the strengths and weaknesses of different graph-based methods for metabolomic data interpretation.
• Emphasizes the critical influence of parameter choices on the outcomes of sub-network extraction.

Conclusions:

• Graph theory offers powerful tools for uncovering biochemical mechanisms from metabolomic data.
• Standardized guidelines and further research are needed to establish consensus in metabolic network analysis.
• The reviewed methods provide a foundation for advancing the interpretation of complex biological systems.