Systematic sensitivity analyses of growth modelling to evaluate the robustness of Genome Scale Metabolic Network models — case study with the filamentous fungus Penicillium rubens

The functionality—and thereby the quality—of a Genome-Scale Metabolic Network (GSMN) model is commonly evaluated by using a biomass reaction as the objective function to determine flux distributions under optimal network states through popular analyses such as Flux Balance Analysis (FBA) or Flux Variance Analysis (FVA). However, assessing model functionality solely by its ability to produce biomass is reductive; the predicted growth must also be biologically consistent. The definition and integration of this artificial reaction should theoretically reflect the organism-specific macromolecular composition and stoichiometry under defined environmental conditions. In practice, however, reconstruction decisions — whether documented, reasoned, or arbitrary — often lead to model refinements aimed at ensuring functionality, which can potentially introduce biases. Such accommodations, along with limited transparency in model development, may compromise the reliability of simulation outcomes. To mitigate this issue, we advocate for the systematic implementation of rapid and simple sensitivity analyses during the final stages of GSMN reconstruction. This approach enables the evaluation of model robustness and the predictive reliability of flux distributions with respect to the biomass reaction formulation. To illustrate this framework, we apply two sensitivity tests to the i Prub22 GSMN of the filamentous fungus Penicillium rubens , an organism known for its capacity to produce a broad range of high-value natural products. The first test evaluates sensitivity to the quantitative definition of the biomass reaction, and the second examines model behaviour under alternative nutrient conditions. Adopting this approach enables a quantitative assessment of the degree of independence exhibited by a model’s flux productions with respect to the composition of biomass reactions.