Generic Chemometric Models for Metabolite Concentration Prediction Based on Raman Spectra

Abstract

Chemometric models for on-line process monitoring have become well established in pharmaceutical bioprocesses. The main drawback is the required calibration effort and the inflexibility regarding system or process changes. So, a recalibration is necessary whenever the process or the setup changes even slightly. With a large and diverse Raman dataset, however, it was possible to generate generic partial least squares regression models to reliably predict the concentrations of important metabolic compounds, such as glucose-, lactate-, and glutamine-indifferent CHO cell cultivations. The data for calibration were collected from various cell cultures from different sites in different companies using different Raman spectrophotometers. In testing, the developed “generic” models were capable of predicting the concentrations of said compounds from a dilution series in FMX-8 mod medium, as well as from an independent CHO cell culture. These spectra were taken with a completely different setup and with different Raman spectrometers, demonstrating the model flexibility. The prediction errors for the tests were mostly in an acceptable range (<10% relative error). This demonstrates that, under the right circumstances and by choosing the calibration data carefully, it is possible to create generic and reliable chemometric models that are transferrable from one process to another without recalibration.

Assessing the capabilities of 2D fluorescence monitoring in microtiter plates with data-driven modeling for secondary substrate limitation experiments of Hansenula polymorpha

“…For example, a common regression model for the pH value of all three experiments of this study could be established. However, comparable to the recent study on Raman spectroscopy by Yousefi-Darani et al [ 51 ], this demands additional workflows for spectral alignment.…”

“…In conclusion, despite the changes in the scattered light and fluorescence dynamics, the resulting PLS models provide a reasonable overall accuracy. Although for individual cultivation conditions the predictive performance showed shortcomings of the models, with a relative RMSE below 10%, the acceptance criterium for a transferable PLS model described by Yousefi-Darani et al [ 51 ] is met.…”

“…These conditions were not covered by the calibration dataset and, thus, represented an extrapolation of the PLS model, for which the performance of the regression method is known to be low [ 58 ]. Consequently, these conditions increased the RMSE above the threshold of 10% and thus rendered the models not acceptable, according to Yousefi-Darani et al [ 51 ]. In a subsequent data evaluation study, it should therefore be reevaluated, if a changed composition of the calibration dataset can improve the PLS modeling performance.…”

Building blocks needed for mechanistic modeling of bioprocesses: A critical review based on protein production by CHO cells

“…It stands out for its spectra's sharpness and compatibility with aqueous systems ( Li et al, 2018 ). Several studies have reported the use of Raman spectroscopy for online monitoring in CHO systems in the last few years ( Li et al, 2018 ; Santos et al, 2018 ; Feidl et al, 2019 ; Yilmaz et al, 2020 ; W Eyster et al, 2021 ; Chen et al, 2021 ; A Gibbons et al, 2022 ; Domján et al, 2022 ; Schwarz et al, 2022 ; Romann et al, 2022 ; Yousefi-Darani et al, 2022 ; Yang et al, 2024 ). The online monitoring of key process variables facilitates the development of mathematical feedback models where operating conditions can be adjusted to obtain the maximum yield in protein production.…”

Comprehensive modeling of cell culture profile using Raman spectroscopy and machine learning

“…By detecting this light, the concentration of specific molecules can be measured. In previous studies, Raman measurement systems were developed for primary medium components and metabolite concentrations, including glucose, lactate, and amino acids 12 , 13 . Additionally, feedback control systems have been established to achieve optimal concentrations 14 – 17 .…”