Aug 2026· Journal of Agricultural and Food Chemistry· Vol 74 32, pp.
24761-24780
· 0 citations· 252 references
Medicine
TL;DR
This review examines enzyme engineering from classical methods to AI-assisted biocatalyst development, highlighting key advances, challenges, and emerging trends in autonomous laboratories, sustainable biocatalysis, and computational protein design.
Abstract
Nowadays, enzyme engineering has moved from traditional structure-based mutagenesis and directed evolution to data-intensive, AI-assisted design paradigms that involve the rapid discovery and optimization of biocatalysts. Whereas classical approaches relied on rational design and experimental screening, advances in high-throughput sequencing, modeling, and machine learning have enabled predictive exploration of sequence-structure-function relationships in enzymes. Importantly, the latest protein language models and deep learning approaches enable accurate prediction of mutational outcomes, stability engineering, and functional annotation at an unprecedented scale. Generative AI models also enable the design of novel enzymes by predicting protein sequences with tailored catalytic functions and broadened substrate specificity. AI combined with design-build-test-learn (DBTL) automation and synthetic biology has enabled the creation of closed-loop engineering workflows for rapid, iterative optimization. This review examines enzyme engineering from classical methods to AI-assisted biocatalyst development, highlighting key advances, challenges, and emerging trends in autonomous laboratories, sustainable biocatalysis, and computational protein design.
Natural enzymes often fail to meet industrial demands for catalytic efficiency, stability, and substrate specificity, creating a critical bottleneck in biomanufacturing. This review examines how artificial intelligence (AI) and automation are reshaping enzyme engineering from empirical trial‑and‑error toward data-driven, closed-loop design. We trace AI development from feature-engineered machine learning to supervised deep learning and self-supervised protein language models, and automation from standalone task execution to cascade integration and biofoundry-enabled build-test workflows. Their convergence is analyzed through a stage-based autonomy framework, highlighting the transition from semi-automated workflows to conditional and high-autonomy DBTL systems. Recent studies demonstrate that AI-guided prediction, automated experimentation, and active learning can accelerate enzyme optimization; however, key barriers remain, including biased datasets, limited out-of-distribution generalization, weak mechanistic interpretability, automation interoperability constraints, and unresolved multi-objective trade-offs. We discuss future directions involving FAIR-compliant data infrastructure, hybrid sequence-structure-physics models, modular automation platforms, and autonomous closed-loop systems. By integrating historical evolution, representative case studies, success and failure analysis, and practical bottlenecks, this review provides a roadmap for advancing AI-guided and autonomous enzyme engineering.
Kexin Hao, Jianguang Liu, Hui Tang et al.· Bioresources and Bioprocessi...· 0 citations
The combination of computational and experimental methods has become indispensable for optimization and rational enzyme design. Recently, the development of artificial intelligence (AI)-based tools has further streamlined enzyme engineering pipelines, enabling more accurate designs, while reducing the number of variants required for experimental validation. However, due to the intricate complexity of enzymatic systems, significant challenges must be addressed before we take the next step to fully optimize the use of these AI-guided enzyme design methodologies. These challenges include un-curated datasets, the need to consider both the static and dynamic structure of enzymes, and the requirement for effective interdisciplinary collaborations to ensure the integration of computational and experimental approaches. Here, we present recent advances in AI-based computational enzyme design, discussing the main challenges in the field and how a combination with classical physics-based methods could help overcome them. We further explore novel trends that could completely modulate the future of protein design and provide our outlook on the key concepts and future opportunities that will shape the next steps of enzyme design.
Rosa Teijeiro-Juiz, Thomas B Brück, Bernhard Loll· Molecules· 0 citations
A comprehensive introduction and overview of several current artificial intelligence (AI)‐driven methods available for enzyme design, with a focus on reaction‐to‐sequence design, structure prediction, substrate scope prediction, engineering of stable variants, design of enzymes with non‐canonical amino acids, and de novo design is offered.
Rosa Teijeiro-Juiz, Nina Egeler, Grzegorz Jamróg et al.· Protein Science· 2 citations
Results indicate that supervised machine learning can help guide the construction of high-value enzyme libraries with expanded catalytic scope, and suggest that supervised machine learning can help guide the construction of high-value enzyme libraries with expanded catalytic scope.
Ravi G. Lal, Jason Yang, Ziyan Zhang et al.· bioRxiv· 0 citations
Protein engineering is a critical technology that modifies protein sequences and structures to optimize specific functions or introduce new ones, and it is widely applied in biomanufacturing, pharmaceutical research and development, synthetic biology and other fields. Traditional protein engineering mainly relies on the modification of natural proteins, and its strategies usually include simulating natural evolution processes or designing based on protein structures, with the aim of enhancing protein stability, activity and other properties. However, the high complexity of protein sequences and the limited understanding of the correspondence between amino acid sequences, protein structures and biological functions have restricted the further development of traditional methods. In recent years, the rapid development of artificial intelligence has greatly driven the paradigm shift in protein engineering research, enabling researchers to efficiently explore proteins with specific structures and functions in a broader sequence space. This paper reviews the basic principles and development history of protein engineering, and introduces traditional strategies such as directed evolution and rational design. It focuses on protein structure prediction tools represented by AlphaFold, generative models such as RFdiffusion and ProteinMPNN, and the de novo protein design methods driven by these tools. Furthermore, combined with classic applications of protein engineering including metabolic engineering, enzyme engineering and antibody engineering, this paper analyzes the progress of artificial intelligence in protein design and optimization. By sorting out relevant studies, this paper aims to provide a reference for understanding the development context of protein engineering and the application of artificial intelligence in this field.
Zonghao Cheng· Theoretical and Natural Scie...· 0 citations
Enzyme enantioselectivity remains a central challenge in the biocatalytic synthesis of chiral pharmaceuticals and fine chemicals. Conventional evolution methods, constrained by prolonged experimental cycles and computationally intensive processes, struggle to comprehensively explore complex protein sequences. Although machine learning (ML) has been successfully applied to optimize enzymatic properties, such as catalytic efficiency and stability, its potential in enantioselectivity engineering remains underexploited. This review systematically evaluates how machine learning integrates multimodal datasets, including sequence, structural, and functional performance data, to advance the discovery and engineering of naturally enantioselective enzymes, and enable the de novo design of artificial enzymes for reaction-relevant applications. Furthermore, key algorithmic frameworks are reviewed. Persistent challenges such as data heterogeneity and limited model generalization capabilities are also critically examined. Moreover, machine learning is increasingly expected to bridge molecular-level enantioselective design with pathway optimization and reactor-scale process engineering, enabling a more integrated and efficient chiral biomanufacturing pipeline. Overall, these advancements open new avenues for efficient chiral biosynthesis and provide theoretical foundations and technical blueprints for a paradigm shift toward intelligence-driven biocatalysis engineering.
Jie Gu, Yan Xu, Xiaoyan Sun et al.· ACS Synthetic Biology· 0 citations