ABSTRACT Artificial intelligence (AI) has emerged as a paradigm‐shifting force in enzyme engineering, enabling data‐driven prediction of catalytic activity, stability, and substrate specificity. By integrating large‐scale datasets from structured databases (e.g., PDB, BRENDA) and high‐throughput experimentation (e.g., deep mutational scanning, microfluidics), machine learning (ML) approaches—including random forests, support vector machines, and deep neural networks—have accelerated enzyme optimization across key domains: mutational profiling, catalytic condition refinement, and mechanistic elucidation. Notably, AlphaFold has revolutionized structure prediction, while AI‐directed evolution enhanced enantioselectivity in nonnatural reactions (e.g., C─Si bond formation). Nevertheless, persistent challenges include data heterogeneity, model overfitting with sparse datasets, and limited interpretability of deep learning frameworks. Future advancements necessitate hybrid strategies merging AI with physics‐based simulations (e.g., molecular dynamics), rigorous standardization of databases (aligned with FAIR principles), and synergistic integration of rational design with data‐driven optimization. This review critically evaluates AI's transformative potential and methodological gaps in enzyme engineering, highlighting implications for sustainable biomanufacturing and industrial biocatalysis.
Du et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: