Machine learning-assisted directed evolution yields dramatic improvement on novel AAV engineering task

Machine learning-assisted directed evolution has proven to significantly enhance the engineering of adeno-associated virus (AAV) capsids, according to a new study. Traditionally, directed evolution involves iterative rounds of selection to discover protein mutants with enhanced fitness, which is commonly employed in AAV capid engineering.

However, the integration of machine learning (ML) into this process has sparked debate over its efficacy compared to traditional methods, primarily due to the added cost of designing ML libraries.

Researchers from the study aimed to clarify the benefits of ML-assisted directed evolution by focusing on manufacturing efficiency, particularly in the context of exosomal encapsulation—a technique that improves AAV immunogenicity. The researchers generated a comprehensive dataset comprising 53,974 mutants generated through three rounds of directed evolution, followed by an independent assessment of 472 mutants designed using ML.

The results were striking: while directed evolution alone resulted in a threefold improvement in manufacturing efficiency, ML-assisted directed evolution achieved an impressive 41-fold enhancement, showcasing the superiority of ML in this application.

The study also highlighted the critical role of data quality in the success of ML models. Specifically, the number of selection rounds and the statistical power derived from mutant counts proved to be more influential than the choice of model architecture. Furthermore, the findings offer practical guidance for researchers on designing and implementing ML-tuned strategies to optimize protein engineering applications.