Avoiding Replicates in Biocatalysis Experiments: Machine Learning for Enzyme Cascade Optimization

ChemCatChem, vol. 17

Abstract

Abstract The optimization of enzyme cascades is a complex and resource‐demanding task due to the multitude of parameters and synergistic effects involved. Machine learning can support the identification of optimal reaction conditions, for example, in the case of Bayesian optimization (BO), by proposing new experiments based on Gaussian process regression (GPR) and expected improvement (EI). Here, in this research BO is used to optimize the concentrations of the reaction components of an enzyme cascade. The productivity‐cost‐ratio is chosen as the optimization objective in order to achieve the highest possible productivity, which was normalized to the costs of the materials used to prevent convergence to ever‐increasing enzyme concentrations. To reduce the experimental effort, contrary to common practice in biological experiments, replicates were not used; instead, the algorithm's proposed experiments and inherent uncertainty quantification were relied upon. This approach balances parameter space exploration and exploitation, which is critical for the efficient and effective identification of optimal reaction conditions. At the optimized reaction conditions identified in this study, the productivity‐cost ratio is doubled to 38.6 mmol L−1 h−1 €−1 compared to a reference experiment. The parameter optimization required only 52 experiments while being robust to outlying experimental results.

Authors 6

1. Regine Siedentop

   TU Dortmund University

   Affiliation as printed

   Department of Biochemical and Chemical Engineering TU Dortmund University Emil‐Figge‐Straße 66 44227 Dortmund Germany

   TU Dortmund University Biochemical and Chemical Engineering Emil-Figge-Str. 66 44227 Dortmund GERMANY

2. Maximilian Siska

   Forschungszentrum Jülich

   Affiliation as printed

   Institute of Bio‐ and Geosciences Forschungszentrum Jülich GmbH Wilhelm‐Johnen‐Straße 52428 Jülich Germany

   Forschungszentrum Julich GmbH Institute of Bio- and Geosciences (IBG) Wilhelm-Johnen-Straße 52428 Jülich GERMANY

3. Johanna Hermes

   TU Dortmund University

   Affiliation as printed

   Department of Biochemical and Chemical Engineering TU Dortmund University Emil‐Figge‐Straße 66 44227 Dortmund Germany

   TU Dortmund University Biochemical and Chemical Engineering GERMANY

4. Stephan Lütz

   TU Dortmund University

   Affiliation as printed

   Department of Biochemical and Chemical Engineering TU Dortmund University Emil‐Figge‐Straße 66 44227 Dortmund Germany

   TU Dortmund University Biochemical and Chemical Engineering Emil-Figge-Str. 66 44227 Dortmund GERMANY

5. Eric von Lieres corresponding Aachen Computational Systems Biotechnology

   Forschungszentrum Jülich · RWTH Aachen University

   Affiliation as printed

   Computational Systems Biotechnology RWTH Aachen University Forckenbeckstraße 51 52074 Aachen Germany

   Institute of Bio‐ and Geosciences Forschungszentrum Jülich GmbH Wilhelm‐Johnen‐Straße 52428 Jülich Germany

   Forschungszentrum Julich GmbH Institute of Bio- and Geosciences (IBG) Wilhelm-Johnen-Straße 52428 Jülich GERMANY

6. Katrin Rosenthal corresponding

   Constructor University

   Affiliation as printed

   School of Science Constructor University Campus Ring 6 28759 Bremen Germany

   Constructor University School of Science Campus Ring 1 28759 Bremen GERMANY