OR WAIT 15 SECS
Advanced data analytics, including statistical modeling and machine learning technique, can enable more efficient and reliable bioprocesses.
Continued process verification is one of the required stages of process validation, which necessitates a documented process and product monitoring program. During routine process and product monitoring, one commercial process (process A) demonstrated higher process variation compared to other commercial products manufactured at the same site. Advanced data analytics, including statistical modeling and machine learning technique, was applied to further understand the process variation in cooperation with a consultant company (McKinsey/Quantum Black). A cross-functional team composed of data engineers, data scientists, bioprocess experts, and translator/project manager was formed and worked together for approximately six months. The analysis consolidated disparate data sources, including information from raw material, equipment, process information, environment, human, etc. Data engineering technique, modeling methodology, and model insights are shared. The importance of building advanced data analytics capabilities to enable more efficient and reliable bioprocesses is discussed.
Submitted: Jan. 24, 2020
Accepted: April 8, 2020
Jun Luo*, email@example.com, is senior manager, and Sid Kundu is engineer II; both are at Manufacturing Sciences, Genentech, Vacaville, CA. Yiming Peng is principal statistical scientist at Non Clinical Biostatistics, Genentech, South San Francisco, CA. Jesse Bergevin is associate director at Process Engineering, Genentech, Vacaville, CA.
*To whom all correspondence should be addressed
Vol. 33, No. 6
When referring to this article, please cite it as J. Luo, et al., “Understanding Commercial Cell Culture Process Performance Variation Through Advanced Data Analytics,” BioPharm International 33 (6) 2020.