Big data analytics streamlining therapeutic drug development from micro-scale through to industrial-scale (cell culture)

该EngD的目的是生成一个数据驱动的决策工具，以加快从细胞系开发到全面生产的新型疗法的交付。该项目将专注于简化数据管理，可视化和分析所有可用的在线，离线和产品质量数据，这些数据是由多个设施和规模的细胞培养操作记录产生的。该项目将开发先进的机器学习算法，并将其应用于与工业相关的细胞培养数据集相关的新应用。这些算法的应用将使探索性分析能够测试假设，识别模式和相关性，并揭示趋势，从而解锁新的见解，这些见解将直接转化为更快地递送新型治疗性蛋白质。该项目最初将侧重于细胞系开发，以评估定义稳健和稳定细胞系的关键性能驱动因素，这些细胞系在从微量（AMBR（15 ml））到实验室规模（3-7 L）并最终到商业生产（15，000 L）的规模扩大过程中表现符合预期。该项目还旨在量化可变性并评估每个细胞系在扩大规模期间的性能，以更好地预测更大规模下的任何潜在故障或预期工艺问题。该项目旨在提供一个机会，通过使用Python和R等软件来分析高维数据，研究不同的大数据和机器学习算法。其他要获得的技能也集中在哺乳动物细胞培养和实验室自动化以及熟练掌握数据科学和机械建模。

The aim of this EngD is to generate a data-driven decisional tool to speed up the delivery of novel therapeutics from cell line development to full-scale manufacturing. The project will focus on streamlining the data management, visualisation and analytics of all available on-line, off-line and product quality data generated by cell culture operations recorded across multiple facilities and scales. The project will develop advanced machine learning algorithms and apply them to novel applications related to industrially-relevant cell culture datasets. The application of these algorithms will enable exploratory analysis to test hypotheses, identify patterns and correlations and uncover trends unlocking novel insights that will directly translate into the faster delivery of novel therapeutic proteins. The project will initially focus on cell line development to evaluate the key performance drivers defining robust and stable cell lines that perform as expected as they scale-up from micro-scale (AMBR (15 ml) to lab-scale (3-7 L) and ultimately to commercial manufacturing (15,000 L). The project also aims to quantify the variability and assess the performance of each cell line during scale-up to better forecast any potential failures or expected process issues at the larger scales. The project aims to provide an opportunity to investigate different big data and machine learning algorithms through the use of software's such as Python and R to analyse data of high dimensions. Other skills to be gained also focus on mammalian cell culture and laboratory automation as well as proficiency in data science and mechanistic modelling.