Boosted structured additive regression for Escherichia coli fed‐batch fermentation modeling

Abstract

The quality of biopharmaceuticals and patients' safety are of highest priority and there are tremendous efforts to replace empirical production process designs by knowledge-based approaches. Main challenge in this context is that real-time access to process variables related to product quality and quantity is severely limited. To date comprehensive on- and offline monitoring platforms are used to generate process data sets that allow for development of mechanistic and/or data driven models for real-time prediction of these important quantities. Ultimate goal is to implement model based feed-back control loops that facilitate online control of product quality. In this contribution, we explore structured additive regression (STAR) models in combination with boosting as a variable selection tool for modeling the cell dry mass, product concentration, and optical density on the basis of online available process variables and two-dimensional fluorescence spectroscopic data. STAR models are powerful extensions of linear models allowing for inclusion of smooth effects or interactions between predictors. Boosting constructs the final model in a stepwise manner and provides a variable importance measure via predictor selection frequencies. Our results show that the cell dry mass can be modeled with a relative error of about ±3%, the optical density with ±6%, the soluble protein with ±16%, and the insoluble product with an accuracy of ±12%. Biotechnol. Bioeng. 2017;114: 321-334. © 2016 Wiley Periodicals, Inc.

Prediction of the Quantity and Purity of an Antibody Capture Process in Real Time

“…The final concentration may deviate slightly from the actual value, but the product meets specified criteria and an optimized yield can be achieved. Other modeling techniques such as random forest, neural networks, or boosted structured additive regression [26,32] may improve the prediction, but such an assessment was beyond the scope of this work. Variations in the training data set will reflect process variations, but additional artificially introduced errors will affect the accuracy of prediction.…”

Integrated process development: The key to improve Fab production in E. coli

“…[ 29 ] This is in line with a quality‐by‐design (QbD) approach in bioprocessing, which demands a thorough understanding of the whole process. [ 1,30 ]…”

Real‐time monitoring and model‐based prediction of purity and quantity during a chromatographic capture of fibroblast growth factor 2

“…So far, QbD and PAT have limited applications in biomanufacturing. Several studies for real‐time monitoring of upstream processes were performed where the most important criteria to be controlled are the product formation and feed strategies and corresponding offline methods (e.g., cell‐based assays) can last up to days (Dabros, Amrhein, Bonvin, Marison, & von Stockar, ; Luchner et al, ; Melcher et al, ; Melcher, Scharl, Luchner, Striedner, & Leisch, ; Pais, Carrondo, Alves, & Teixeira, ; von Stosch, Hamelink, & Oliveira, ). Chromatography is the main purification method for biopharmaceutical proteins.…”

“…As HCP and dsDNA are critical host cell impurities they were addressed in this study. Data analysis was performed with boosted STAR, which gives promising results in settings, where the number of variables (greatly) exceeds the number of observations, as it is common, when spectroscopic sensor systems are involved (Melcher et al, ).…”