Predicting raw material impact on cell culture parameters in commercial biotherapeutic manufacturing

• Published in: Biochemistry and biophysics reports Vol. 43; p. 102192
• Main Authors: Pugh, Phyllis C., Khanal, Bijay R., Lemons, Jenna L., Murillo, Mario A., Patel, Jacika N., Boppana, Preeya K., Padmanabhan, Veera
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.09.2025; Elsevier
• Subjects: Chemometrics; Fourier-transform infrared spectroscopy (FTIR); Multivariate analysis; Partial least squares; Process analytical technology (PAT); Silico modeling; Spectroscopy; Fourier-transform infrared spectroscopy (FTIR); Process analytical technology (PAT); Spectroscopy; Silico modeling; Multivariate analysis; Partial least squares; Chemometrics
• ISSN: 2405-5808; 2405-5808
• DOI: 10.1016/j.bbrep.2025.102192

Complex, chemically-undefined media components are often used as nutrients in the production of biological products via mammalian cell culture. Variability in the compositions of these complex raw materials can significantly impact product yields. This paper investigates the influence of raw material quality on the cell culture process by developing data-based models to estimate final productivity in an industrial antibody production operation at AstraZeneca. Fourier Transform Infrared (FTIR) spectroscopy measurements of selected raw material components were obtained. These measurements were processed, derivatized, and used to create Partial Least Squares (PLS) regression chemometric models. The resulting models were then employed to predict the influence of such raw variability on the yields of biotherapeutic molecules. [Display omitted] •Chemometric models assess raw material impact on antibody production.•PLS regression models can be created from FTIR spectroscopy.•Models effectively predict productivity at commercial scale.•FTIR screening cuts raw material evaluation time from a month to two days.