Advanced Data Analytics Application in Biomanufacturing Processes

Biomanufacturing drug substance processes include cell culture (thaw, seed train, inoculum, and protein production), harvest, and purification steps. Due to the living nature, the bioprocess inherently has higher variability when compared to chemical reactions. Ensuring consistent process performance and continuous process improvement are ongoing challenges for bioprocess. Thorough characterization of process and product attributes using statistical design of experiments (DoE) during the process design (PD) stage is a proven means to establish the process and product knowledge. Recent accelerated development timelines are putting pressure and limitations on the extent of characterization during the PD stage, and, therefore, increasing the importance of a lifecycle approach to accumulate process and product knowledge during the post-approval stage of process validation (i.e., continued process verification). Following telecoms, advertising, and insurance, the biopharma industry has started to embrace methods like Bayesian statistics and advanced data analytics (ADA) to gain additional process and product understanding, improved process control, and process performance. Using ADA may elucidate previously undetected relationships between process inputs and outputs, which hold promise as an additional tool to augment traditional DoE as a means to gain actionable insights, process and product knowledge. An internal multidisciplinary team (e.g., ADA) comprising data engineers, data scientists, bioprocess experts, and a translator /project manager is formed. IT infrastructure to support ADA projects is established, the standard approaches to run the case studies are developed, and a library of models for bioprocess is built. An upstream process case study to improve upstream productivity and process robustness is discussed. The team is able to extract, cleanse, ingest, and process 800 GB of data from 11 disparate data sources into a cloud environment to evaluate more than 100 hypotheses. Implementing the insights leads to improved process performance. Through the case study, the strategy to sustain the internal capability is developed. The importance of building ADA capabilities to enable more efficient and reliable bioprocesses is discussed.