Hybrid in silico and ultrasound-assisted enzymatic extraction as a functionalizing strategy for cassava starch:...

Cassava (Manihot esculenta Crantz) is a key source of starch, yet conventional extraction methods face significant drawbacks, such as excessive water consumption and low efficiency. This study combined in silico and experimental approaches to optimize cassava starch extraction through ultrasound-assisted enzymatic biocatalysis, while assessing its effects on physicochemical, morphological, structural, and pasting properties. Molecular docking and dynamics simulations revealed strong binding affinities of pectin lyase (Aspergillus aculeatus) and 1,4-β-cellobiohydrolase (Trichoderma reesei) toward pectin and cellulose (-6.2 and - 7.2 kcal/mol), validating their potential to degrade the plant matrix. In the experimental phase, cellulase and pectinase were employed, individually and in combination with ultrasound (37 kHz, 100 W/cm²), to extract starch from cassava cv. Venezolana MCOL 2215. Ultrasound-assisted pectinase treatment (PEC + UTS) resulted in high starch purity (93.86%), within technologically acceptable ranges and not statistically different from the conventional process, while enhancing crystallinity, water absorption capacity, and thermal stability, and reducing apparent amylose content and gelatinization enthalpy. Granule morphology was preserved, with only slight surface erosions under ultrasonic treatments. Pasting profiles showed reduced retrogradation, a desirable feature for food systems requiring paste stability. This integrative in silico-experimental framework demonstrates that ultrasound-assisted enzymatic biocatalysis can substantially enhance cassava starch extraction while preserving its structural integrity. The confirmed enzyme-substrate interactions provide molecular insight into the observed synergy. These findings establish a scientific basis for future scale-up and optimization of extraction processes to improve starch recovery and functional performance.