Accelerated enzyme engineering by machine-learning guided cell-free expression
Problem
Engineering enzymes is slow because generating large sequence-function datasets for predictive design is difficult.
Why it matters
Faster enzyme engineering enables cheaper biological production of pharmaceuticals and chemicals.
What the AI did
An ML-guided platform integrating cell-free DNA assembly, cell-free expression, and functional assays mapped fitness landscapes; augmented ridge regression models predicted amide synthetase variants for pharmaceutical synthesis.
Results
1,217 enzyme variants evaluated in 10,953 unique reactions; models predicted variants for 9 small molecule pharmaceuticals
Impact
Demonstrated ML-guided engineering of amide synthetases across multiple reactions in a peer-reviewed venue.
Record
| Venue | Nature Communications |
|---|---|
| Published | 2025 |
| Work type | article |
| Authors | Grant M. Landwehr, Jonathan W. Bogart, Carol Magalhaes, Eric Hammarlund, Ashty S. Karim, Michael C. Jewett |
| Citations | 134 (OpenAlex, at import time) |
| References | 75 |
| Model or method | augmented ridge regression |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41467-024-55399-0 |
| Open access | gold |
| Keywords | machine learning, enzyme engineering, cell-free expression, amide synthetase, biocatalysis |
Importance rating
real but narrower advance
ML-guided enzyme engineering accelerates biocatalyst design for pharmaceutical synthesis; real advance but demonstrated on specific amide synthetase reactions.