Waluga, ThomasThomasWalugaNazarpoor Najafabadi, HosseinHosseinNazarpoor NajafabadiZavrel, MichaelMichaelZavrelSkiborowski, MirkoMirkoSkiborowski2026-10-272026-10-272026-09-16Digital Chemical Engineering 21: 100344 (2026)https://hdl.handle.net/11420/65476Lignocellulosic biomass has a high potential as resource for the chemical industry, but its conversion to fermentable sugars requires efficient pre-treatment and a subsequent enzymatic saccharification. Here, we present a model-based workflow for the design of enzyme cocktails for lignocellulosic biomass saccharification. Saccharification performance depends on several factors, such as substrate structure and pre-treatment severity, enzyme synergy, and inhibition effects. Because of this complexity, optimization by experimental trial-and-error is expensive and time-consuming, and enzyme mixtures optimized for one substrate often lack transferability to other feedstocks. To address these challenges, this work couples a stochastic simulator of lignocellulose hydrolysis with multi-objective optimization, explicitly resolving the trade-off between conversion speed and enzyme usage as a proxy for costs. The results demonstrate the feasibility and potential of model-based optimisation for rational enzyme cocktail design.en2772-50812026Elsevierhttps://creativecommons.org/licenses/by/4.0/Bayesian optimizationLignocelluloseSaccharificationTechnology::660: Chemistry; Chemical Engineering::660.6: BiotechnologyJournal Article10.1016/j.dche.2026.100344