Machine learning-assisted directed evolution yields dramatic improvement on novel AAV engineering task
- 1Department of Engineering Science, University of Oxford, Parks Road, Oxford, OX1 3PJ.
Abstract
Directed evolution enables the discovery of protein mutants with improved fitness through iterative rounds of selection and has been widely applied to adeno-associated virus (AAV) capsid engineering. Machine learning (ML) can augment this process by prioritising mutants for experimental validation, but whether its benefits outweigh the added cost of ML-designed library construction remains unclear. Here, we address this question by considering improvements in manufacturing efficiency of AAV capsids in the context of exosomal encapsulation, which is a technique for improving AAV immunogenicity. We generated a directed evolution dataset comprising 53,974 mutants across three rounds of selection and an independent assessment dataset of 472 ML-designed mutants with detailed profiling. Directed evolution alone yielded a 3-fold improvement, while ML-assisted directed evolution yielded a 41-fold improvement in manufacturing efficiency, demonstrating that ML-assisted directed evolution outperforms directed evolution alone. We show that data quality - specifically the number of selection rounds and the statistical power of mutant counts - has a greater impact on model performance than model architecture. Finally, we provide practical guidance on experimental design and implementation of ML tuning strategies that can augment model performance in protein engineering applications.
Related
- Heterologous engineering of receptors using OrthoRep (HERO) for directed evolution of GPCRs in yeast
- Directed evolution of a thermostable family-A DNA polymerase into a reverse transcriptase for 2'-modified RNA
- Engineering Mosaic T7 Bacteriophages for Dual Peptide Display to Enable Cellular Targeting and Ligand-Directed Protein Capture.
- ECG-RSI: Recursive Self-Improvement through Targeted Electrocardiogram Waveform Evolution
- Directed Evolution of Multimodal Plastic Degrading Enzymes
BibTeX
@misc{luchner2026machine,
title = {{Machine learning-assisted directed evolution yields dramatic improvement on novel AAV engineering task}},
author = {M. R. Luchner and I. Vorobieva and R. Makkar and M. Jozinovic and S. Chester and M. J. A. Wood and S. J. Sanders and D. Gupta and H. Steel},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.06.757046},
url = {https://www.biorxiv.org/content/10.64898/2026.10.06.757046v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Machine learning-assisted directed evolution yields dramatic improvement on novel AAV engineering task
AU - M. R. Luchner
AU - I. Vorobieva
AU - R. Makkar
AU - M. Jozinovic
AU - S. Chester
AU - M. J. A. Wood
AU - S. J. Sanders
AU - D. Gupta
AU - H. Steel
PY - 2026
DA - 2026/10/06
PB - bioRxiv
DO - 10.64898/2026.10.06.757046
UR - https://www.biorxiv.org/content/10.64898/2026.10.06.757046v1
AB - Directed evolution enables the discovery of protein mutants with improved fitness through iterative rounds of selection and has been widely applied to adeno-associated virus (AAV) capsid engineering. Machine learning (ML) can augment this process by prioritising mutants for experimental validation, but whether its benefits outweigh the added cost of ML-designed library construction remains unclear. Here, we address this question by considering improvements in manufacturing efficiency of AAV capsids in the context of exosomal encapsulation, which is a technique for improving AAV immunogenicity. We generated a directed evolution dataset comprising 53,974 mutants across three rounds of selection and an independent assessment dataset of 472 ML-designed mutants with detailed profiling. Directed evolution alone yielded a 3-fold improvement, while ML-assisted directed evolution yielded a 41-fold improvement in manufacturing efficiency, demonstrating that ML-assisted directed evolution outperforms directed evolution alone. We show that data quality - specifically the number of selection rounds and the statistical power of mutant counts - has a greater impact on model performance than model architecture. Finally, we provide practical guidance on experimental design and implementation of ML tuning strategies that can augment model performance in protein engineering applications.
ER -