High-Throughput Algorithmic Optimization of In Vitro Transcription for SARS-CoV-2 mRNA Vaccine Production

Area of Science:

• Biotechnology
• Molecular Biology
• Machine Learning Applications

Background:

• Optimizing in vitro transcription (IVT) is crucial for producing high-quality messenger ribonucleic acid (mRNA).
• Efficient mRNA production is essential for developing vaccines and therapeutics, particularly for rapidly emerging viral variants like SARS-CoV-2 Delta.
• Current IVT methods often face challenges with yield, purity, reaction time, and reagent costs.

Purpose of the Study:

• To optimize in vitro transcription (IVT) conditions for messenger ribonucleic acid (mRNA) production from the SARS-CoV-2 Delta variant DNA template.
• To enhance total mRNA yield and purity (percent intact mRNA) using machine learning and automated liquid handling.
• To evaluate commercially available T7 RNA polymerases for improved mRNA synthesis.

Main Methods:

• Utilized an iterative Bayesian optimization approach to fine-tune 11 critical process parameters.
• Employed automated, high-throughput liquid handling technology for efficient reaction execution.
• Conducted automated, high-throughput screening of T7 RNA polymerases under optimized conditions.

Main Results:

• Achieved a 12% improvement in total mRNA yield.
• Reduced reaction time by 50%.
• Decreased the use of expensive reagents by up to 44% while improving mRNA purity.

Conclusions:

• Developed a novel, efficient platform for optimizing IVT reactions for mRNA production.
• Demonstrated the successful application of machine learning and automation in enhancing mRNA synthesis.
• The optimized process offers significant improvements in yield, speed, cost-effectiveness, and purity for mRNA production.