Possibilistic validation of a constraint-based model under data scarcity: application to Pichia pastoris cultures

Constraint-based modelling allows building structured models of cells without accounting for intracellular kinetics. These models can be combined with experimental data to estimate the (pseudosteady) state or phenotype exhibited by cells at given conditions, standing out as a useful analytical tool. In this work, a simplified, constraint-based model of Pichia pastoris, a widely recognized platform for recombinant protein expression, is derived from its metabolic network. Then, the model is validated against experimental data provided by different research groups: possibility theory is used to analyse the consistency between model and measurements. Afterwards, the biomass growth rate is estimated to illustrate the ability of the model to predict non-measured fluxes. The approach followed in this contribution is particularly useful in scenarios lacking data; it makes it possible to link the extracellular behaviour of Pichia pastoris during cultivation with its internal state, being a promising tool for optimization and monitoring industrial processes.