Application of MOGA-ANN tool for the production of cellulase and xylanase using de-oiled rice bran (DORB) for bioethanol production

India is the second-largest producer of paddy in the world, with an estimated annual production of 165 metric tons (MT) and about 10.8 MT of rice bran. Partially utilized agro-industrial residues, de-oiled rice bran (DORB) is an unexplored substrate for the production of industrial enzymes. The present study focused on producing and optimizing fungal cellulases and xylanase enzymes by Aspergillus niger VSRK09 under SSF conditions. The effect of physiological parameters was confirmed by one-factor-at-a-time (OFAT), followed by the RSM-based FCCCD method using statistical design. The enzyme-producing abilities of A. niger VSRK09 were improved using a hybrid statistical tool viz. artificial neural network incorporated with a multi-objective genetic algorithm (MOGA-ANN). This study revealed that the MOGA-based model resulted in an optimized enzyme activity, i.e., 24.8 FPU gds⁻¹ and 520 IU gds⁻¹ of cellulase and xylanase, respectively, at incubation time of 3.5 days, a substrate to moisture (SM) ratio of 1:1, and inoculum size of 3.0 × 10⁶ spores mL⁻¹. The validation performance of ANN indicates that the designed model is trained and tested with a minimum mean square error and inserted in MOGA, which produced the best-optimized results. Saccharification of DORB released 320.5 ± 8.75 (mg gds⁻¹) of reducing sugar which was subsequently converted to bioethanol using Saccharomyces cerevisiae yielded 13.48 ± 1.21 g L⁻¹ ethanol.