Abstract In this study, a constrained mixture design was applied to evaluate the potential of agro-industrial by-products, namely cocoa fruit shell (CFS), cocoa bean shell (CBS), and palm oil (PO) as substrates for enzyme production. A desirability function approach was employed to optimize the simultaneous production of endoglucanase (EGL), exoglucanase (EXG), and β-glucosidase (BGL) by Penicillium roqueforti ATCC 10110 under solid-state fermentation (SSF). The desirability function reached a value of 0.84, indicating successful fulfillment of the optimization criteria. The optimized substrate proportions CFS (7.5 g), CBS (1.0 g), and PO (1.5 g) resulted in enzyme activities of (5.38 IUg −1 ) for EGL, (1.01 IUg −1 ) for EXG, and (2358.75 IUg −1 for BGL). All enzymes exhibited high stability at 50 °C and at pH 4.5 and 6.0. The presence of CoSO 4 and EDTA enhanced the activities of EGL, EXG, and BGL by more than 100%. The effects of solvent addition were further analyzed using artificial neural networks based on a Kohonen Self-Organizing Map (KSOM) to identify correlations among the observed responses. Overall, the application of desirability based optimization combined with constrained mixture design proved effective in enhancing cellulase production by P. roqueforti ATCC 10110. The characterized enzymes demonstrated have shown significant potential for industrial applications. Graphical Abstract
Araújo et al. (Tue,) studied this question.