Engineering of β-Agarase with Enhanced Thermostability via Multitool Consensus Prediction and Structure-Guided Screening

β-Agarase plays a crucial role in preparing bioactive agar oligosaccharides, but its insufficient thermostability limits industrial application. In this study, we developed a stepwise design strategy integrating multitool consensus prediction, structure-based rational screening, and greedy algorithm-based optimization to enhance AgaDcat’s thermostability. This yielded mutant M3 (N120S-D243N-Q246A-S287E-A335D), which exhibits an 11 °C higher melting temperature ( T m ) and 14-fold longer half-life ( t 1/2 ) at 50 °C than the wild-type. Molecular dynamics simulations showed that mutations strengthened hydrophobic interactions, salt bridges, and hydrogen bond networks, while optimizing surface charge. The M3 variant performed well in high-temperature agarose hydrolysis, mainly producing neoagarotetraose (NA4) and neoagarohexaose (NA6), showing great industrial potential. This framework improves the efficiency of enzyme thermostability engineering and provides a general approach for industrial enzymes modification.