News, Recent papers

Machine learning in predictive biocatalysis: A comparative review of methods and applications, Biotechnology Advances

Harnessing Artificial Intelligence for Predictive Biocatalysis


We are advancing the field of predictive biocatalysis through cutting-edge machine learning approaches. Our team (BRS team at MICALIS, INRAE, Jouy-en-Josas and LISSB/Genoscope, CEA–Évry) recently published a review article, “Machine learning in predictive biocatalysis: A comparative review of methods and applications”, in Biotechnology Advances (https://www.sciencedirect.com/science/article/pii/S0734975025001843). This work is supported by PEPR B BEST.

Biocatalysis, the use of enzymes to accelerate chemical reactions, is central to sustainable innovation in green chemistry, biotechnology, and pharmacy. Designing efficient new biocatalysts remains a major challenge given the complexity of biological systems. By leveraging artificial intelligence, we are developing novel and powerful tools to anticipate enzyme function, predict reaction efficiency, and design new catalysts.

In our review, we compare a range of machine learning methods, from deep neural networks to graph and transformer models, and highlight their respective strengths and limitations. The article provides a comprehensive overview of methods for various annotation tasks in biocatalysis.  We also emphasize the importance of large datasets and molecular representation techniques in accelerating enzyme discovery and enabling sustainable processes. At the same time, we examine the key challenges that remain, including data quality and the inherent biochemical complexities of enzyme systems. Addressing these challenges will be critical for moving the field forward.

This review is aimed to be a valuable resource for researchers interested in advancing enzyme engineering and exploring new opportunities in synthetic biology, artificial metabolism, and green biocatalysis.

Ref : Tripathi N, Hérisson J, Faulon JL. Machine learning in predictive biocatalysis: A comparative review of methods and applications. Biotechnology Advances (2025). 10.1016/j.biotechadv.2025.108698