Integrating Prior Knowledge From Genome-Scale Metabolic Model With Metabolomics for Diet Assessment.

Area of Science:

• Metabolomics
• Systems Biology
• Nutritional Science

Background:

• Dietary biomarker metabolite detection is crucial but lacks biomechanistic understanding.
• Previous work introduced features integrating genome-scale metabolic models with metabolomes.
• Existing methods face limitations with small feeding trial cohorts.

Purpose of the Study:

• To investigate the combined impact of reaction and subsystem features for diet assessment.
• To evaluate the effect of prior knowledge volume on predictive modeling accuracy.
• To assess the robustness of proposed features for multi-diet classification.

Main Methods:

• Integration of novel features derived from genome-scale metabolic models and metabolomic data.
• Comparative analysis of feature combinations (reaction, subsystem, and novel feature).
• Evaluation of model performance across varying volumes of prior knowledge and multiple dietary conditions.

Main Results:

• Combined reaction and subsystem features demonstrate improved performance in several experimental settings.
• Diet assessment accuracy increases with higher volumes of prior knowledge, with diminishing returns.
• The proposed features exhibit good generalization capabilities for multi-diet assessment.

Conclusions:

• Integrating prior metabolic knowledge with metabolomics offers a robust approach to diet assessment.
• Novel features enhance the understanding of diet-related metabolic pathways.
• The developed methodology shows promise for improving predictive modeling in nutritional science.