Application of novel Raman spectroscopy device and advanced data analytics to optimise viral vector production and design

有两种方法来优化病毒载体生产，第一种是改善生物反应器内的微环境，第二种是改善病毒载体构建体的设计。这两种方法都有可能提高病毒载体滴度、产品质量并使规模扩大更容易。本项目的第一部分将重点关注优化生物反应器微环境的控制，因为这对于确保产品异质性保持在商业良好生产规范（GMP）操作规定的预定义质量标准范围内至关重要。然而，在病毒载体生产中，连续在线监测仅适用于少数工艺参数（即pH、温度和DO 2）。目前，这些生物反应器工艺的控制实施了基本和次优的控制策略，这些策略严重依赖于离线数据分析和不频繁的操作员决策。机器学习技术与先进的过程分析技术（PAT）（如拉曼光谱）相结合，可以实现关键CPP和CQA的真实的预测。这些实时预测可以与先进的反馈控制器相结合，用于优化工艺性能，例如最大化病毒滴度和最小化生产杂质，如空衣壳。一旦开发出控制算法，该项目的第二部分将是修改cap或rep基因的基因序列，并评价这些设计如何改善明确定义和受控的生物反应器系统中的临床生产工艺。

There are two methods to optimise viral vector production, the first is to improve the micro-environment within the bioreactor and the second is to improve the design of the viral vector constructs. Both of these approaches have the potential to improve viral vector titres, product quality and enable easier scale-up. The first part of this project will focus on optimising the control of the bioreactor micro-environment as this is of paramount importance to ensure product heterogeneity remains within a predefined specification defined by commercial good manufacturing practice (GMP) operations. However, continuous online monitoring is available only for a small number of process parameters (i.e pH, temperature and DO2) in viral vector manufacturing.Currently, the control of these bioreactor processes implements rudimentary and suboptimal control strategies that rely heavily on off-line data analytics and infrequent operator decisions. Machine learning techniques combined with advanced Process Analytical Technologies (PATs) such as Raman spectroscopy can enable real time predictions of key CPPs and CQAs. These real-time predictions can be combined with advanced feedback controllers that can be used to optimise process performance such as maximising viral titres and minimising manufacturing impurities such as empty capsids. Once the control algorithms have been developed, the second part of the project will be to modify the gene sequence of the cap or rep gens and evaluate how the design of these will improve clinical manufacturing processes within a well-defined and controlled bioreactor system.