McMinn Spencer E, Miller Danielle V, Yur Daniel, Stone Kevin, Xu Yuting, Vikram Ajit, Murali Shashank, Raffaele Jessica, Holland David, Wang Sheng-Ching, Smith Joseph P
Process Research and Development, Merck & Co., Inc., West Point, Pennsylvania 19486, United States.
Biometrics Research, Merck & Co., Inc., Rahway, New Jersey 07065, United States.
Biochemistry. 2024 Nov 5;63(21):2793-2802. doi: 10.1021/acs.biochem.4c00188. Epub 2024 Oct 20.
The transcription (IVT) of messenger ribonucleic acid (mRNA) from the linearized deoxyribonucleic acid (DNA) template of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Delta variant (B.1.617.2) was optimized for total mRNA yield and purity (by percent intact mRNA) utilizing machine learning in conjunction with automated, high-throughput liquid handling technology. An iterative Bayesian optimization approach successfully optimized 11 critical process parameters in 42 reactions across 5 experimental rounds. Once the optimized conditions were achieved, an automated, high-throughput screen was conducted to evaluate commercially available T7 RNA polymerases for rate and quality of mRNA production. Final conditions showed a 12% yield improvement and a 50% reduction in reaction time, while simultaneously significantly decreasing (up to 44% reduction) the use of expensive reagents. This novel platform offers a powerful new approach for optimizing IVT reactions for mRNA production.
利用机器学习结合自动化高通量液体处理技术，对严重急性呼吸综合征冠状病毒2（SARS-CoV-2）Delta变异株（B.1.617.2）线性化脱氧核糖核酸（DNA）模板的信使核糖核酸（mRNA）转录（IVT）进行优化，以提高总mRNA产量和纯度（以完整mRNA百分比计）。一种迭代贝叶斯优化方法在5轮实验中的42个反应中成功优化了11个关键工艺参数。一旦达到优化条件，便进行了自动化高通量筛选，以评估市售T7 RNA聚合酶的mRNA生产速率和质量。最终条件显示产量提高了12%，反应时间减少了50%，同时显著减少（最多减少44%）了昂贵试剂的使用。这个新平台为优化用于mRNA生产的IVT反应提供了一种强大的新方法。
