The process of lentiviral vector production is complex and presents unique challenges. Optimal production is heavily dependent on the concentration and ratios of the plasmid DNA components contributing to the transfection process. Improving titres can determine whether a new gene therapy product is commercially viable. A considerable amount of research is aimed at establishing the best titre and one of the ways to do this is by ensuring optimal transfection ratios that can improve the productivity or recovery of upstream/downstream platform processes, respectively. 

This means that a multifactorial design approach is often required to address the biological complexity of the system and achieve optimised vector particle production. We have developed and optimised a cutting-edge digital and physical platform coupled with automated liquid handlers enabling:  

1) Automated Design of Experiments (DoE) and multivariate systems optimisation;  

2) High-throughput screening, including small molecules, siRNA, shRNA;  

3) Machine Learning and Deep Learning Data Analysis; 

4) Robustness and Ruggedness Testing to Support Analytical Quality by Design (AQbD) Methods.  

Our platform routinely screens and models 5-10 factors. We have proven capacity for modelling 100+ factors with execution of thousands of experimental runs in a single iteration. The new platform, High-throughput Optimisation Services (HtOS), has significantly reduced manual operator time, increased throughput while meeting high demand and with improvements in precision and robustness. We will describe the capabilities and present a Case Study where the new platform was used to successfully optimise the transient transfection plasmid ratios in upstream production of lentiviral vector.