Advances in Machine Learning for Directed Evolution: A Five-Year Retrospective
Abstract
The last five-plus years have seen many protein engineering disciplines transformed by advances in machine learning (ML), but the same cannot be said for directed evolution. Reflecting on a previously co-authored perspective, I discuss why I believe this to be the case, arguing that a disconnect between the goals of machine-learning-assisted directed evolution (MLDE) researchers--"identify an optimal protein"--and the goals of directed evolution more broadly--"identify a sufficient protein given time and resource constraints"--is a principal culprit. As an example, I highlight how nearly all current MLDE methods neglect to account for the cost of DNA synthesis, resulting in strategies that have limited practical applicability regardless of the underlying models' capabilities. I close by discussing recent works that are exceptions to this overarching trend, and emphasize that the last five years of efforts in ML-assisted protein engineering and the prescribed reframe of MLDE objectives need not be mutually exclusive.
Related
- Machine learning–assisted directed evolution of plant Rubisco
- Therapeutic Advances in Mycobacterium avium Complex Pulmonary Disease: Current Treatments and Emerging Host-Directed Approaches
- Directed Evolution of Threonine Aldolases for the Photobiocatalytic Synthesis of β-Branched-α-Tetrasubstituted and Other Noncanonical Amino Acids
- PepSpace: An Automated, Physics-Driven Directed Evolution Platform for the De Novo Design of Specific Peptide Binders and Condensate Modulators
- Directed Evolution of a Non-Heme Iron Enzyme for Enantioselective Dearomative Nitrene Transfer: Synthesis of 3a-Aminopyrroloindoline Alkaloids
BibTeX
@misc{wittmann2026advances,
title = {{Advances in Machine Learning for Directed Evolution: A Five-Year Retrospective}},
author = {Bruce J. Wittmann},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2609.03046},
eprint = {2609.03046},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.03046},
note = {CC BY-SA 4.0},
}
RIS
TY - UNPB
TI - Advances in Machine Learning for Directed Evolution: A Five-Year Retrospective
AU - Bruce J. Wittmann
PY - 2026
DA - 2026/10/21
PB - arXiv
DO - 10.48550/arxiv.2609.03046
UR - https://arxiv.org/abs/2609.03046
AB - The last five-plus years have seen many protein engineering disciplines transformed by advances in machine learning (ML), but the same cannot be said for directed evolution. Reflecting on a previously co-authored perspective, I discuss why I believe this to be the case, arguing that a disconnect between the goals of machine-learning-assisted directed evolution (MLDE) researchers--"identify an optimal protein"--and the goals of directed evolution more broadly--"identify a sufficient protein given time and resource constraints"--is a principal culprit. As an example, I highlight how nearly all current MLDE methods neglect to account for the cost of DNA synthesis, resulting in strategies that have limited practical applicability regardless of the underlying models' capabilities. I close by discussing recent works that are exceptions to this overarching trend, and emphasize that the last five years of efforts in ML-assisted protein engineering and the prescribed reframe of MLDE objectives need not be mutually exclusive.
ER -