In summary the CSFp approach yielded a total of 45 mutants through evolutionary processes with a notable 69 exhibiting D-Lac activity Among these 16 mutants displayed activity levels exceeding an orde

In summary, the CSFp approach yielded a total of 45 mutants through evolutionary processes, with a notable 69% exhibiting D-Lac activity. Among these, 16 mutants displayed activity levels exceeding an order of magnitude, with the top-performing variant, N96S/A271E/F274Y/F308G, demonstrating a remarkable 56-fold increase in activity (Figure 6). The results evince that the combination of In Silico and experimental methodologies can expedite the enhancement of D-Lac activity. This approach enables us to evolve enzymes with unknown crystal structures more rapidly and robustly. While employing a solitary in silico software package may yield marginal enhancements in enzyme activity, the synergistic impact of multiple in silico modules working in tandem is more likely to produce the optimal outcome we seek. Therefore, the CSFp strategy proposed in this study can facilitate the development of computational enzyme evolution by significantly reducing screening efforts and rapidly evolving enzymes without experimental structural information and with uncertain enzyme transition state models. This approach further expands the application of the structure predictor in the field of enzyme design