Enabling Raman Spectroscopy-Based Cell Culture Process Control via Real-Time Chemometric Model Calibration

Raman spectroscopy, combined with chemometric models such as Partial Least Squares (PLS), is a powerful enabling technology for real-time monitoring and control of industrial cell culture processes. Specifically, the non-destructiveness and near-real time measurement nature of the technology offers great potential in how processes are designed, monitored, and controlled, yet its implementation in industrial settings has been limited. This is mainly due to extensive dataset requirements for chemometric model development, as dozens to hundreds of cell‑culture constituents are known to contribute to Raman scattering effects. Experimental artifacts and anomalies arising from different sites, scales, or equipment further exacerbate this issue. To address these challenges, we introduce a novel one‑point calibration methodology that reframes the model training and validation paradigm from the "best chemometric model for the historical data" to the "best model for the most recent and representative data". Implemented within an adaptive‑learning framework, this approach drastically reduces dependence on extensive historical datasets while better accommodating site-, process-, and run-to-run- variabilities for the 9-14 day fed‑batch Chinese Hamster Ovary (CHO) cell culture processes investigated.