Systems Biology Approaches to Metabolic Pathway Modeling

Abstract

Systems biology integrates experimental data with computational models to understand how metabolic pathways function as interconnected networks rather than isolated reactions, yet the adoption of these modelling frameworks into routine biomedical and biotechnological workflows varies widely across methodology types. We evaluated 214 systems biology metabolic modelling programmes active across centres in Spain and Switzerland between 2015 and 2021, spanning five methodology categories: constraint-based flux balance analysis, kinetic ordinary differential equation models, stochastic simulation approaches, hybrid multiscale frameworks, and machine-learning-augmented metabolic models. A Metabolic Modelling Adoption Index (MMAI) was constructed from five sub-scores -- predictive accuracy for metabolic phenotypes, model scope and pathway coverage, parameter identifiability, experimental validation depth, and computational accessibility -- with weights from regression against sustained adoption into active research or industrial pipelines. MMAI correlated with adoption at r = +0.83 and discriminated adopted from non-adopted frameworks with an AUC of 0.878. Constraint-based flux balance analysis scored highest (mean MMAI 0.822), while stochastic simulation trailed at 0.594. Only 34.6 percent of programmes exceeded the 0.75 threshold. Predictive accuracy carried the largest regression weight (beta = +0.278), followed by computational accessibility (beta = +0.228).