Maximizing mRNA vaccine production with bayesian optimization

Sara Sousa Rosa,Davide Nunes,Luis Antunes,Duarte M. F. Prazeres,Marco P. C. Marques,Ana M. Azevedo

DOI: https://doi.org/10.1002/bit.28216

IF: 3.8

2022-08-28

Biotechnology and Bioengineering

Abstract:mRNA vaccines are a new alternative to conventional vaccines with a prominent role in infectious disease control. These vaccines are produced in in vitro transcription (IVT) reactions, catalyzed by RNA polymerase in cascade reactions. To ensure an efficient and cost‐effective manufacturing process, essential for a large‐scale production and effective vaccine supply chain, the IVT reaction needs to be optimised. IVT is a complex reaction that contains a large number of variables that can affect its outcome. Traditional optimisation methods rely on classic Design of Experiments (DoE) methods, which are time‐consuming and can present human bias or based on simplified assumptions. In this contribution, we propose the use of Machine Learning approaches to perform a data‐driven optimization of an mRNA IVT reaction. A Bayesian optimization method and model interpretability techniques were used to automate experiment design, providing a feedback loop. IVT reaction conditions were found under 60 optimization runs that produced 12 g.L−1 in solely two hours. The results obtained outperform published industry standards and data reported in literature in terms of both achievable reaction yield and reduction of production time. Furthermore, this shows the potential of Bayesian optimization as a cost‐effective optimisation tool within (bio)chemical applications.This article is protected by copyright. All rights reserved.

biotechnology & applied microbiology