Hybrid-EKF: Hybrid Model coupled with Extended Kalman Filter for real-time monitoring and control of mammalian cell culture.

In a decade when industry 4.0 and Quality by Design are major technology drivers of biopharma, automated and adaptive process monitoring and control are inevitable requirements and model-based solutions are key enablers in fulfilling these goals. Despite strong advancement in process digitalization, in most cases, the generated data sets are not sufficient for relying on purely data-driven methods, while the underlying complex bioprocesses are still not completely understood. In this regard, hybrid models are emerging as a timely pragmatic solution to synergistically combine available process data and mechanistic understanding. In this work we show a novel application of Hybrid-EKF framework, that is hybrid models coupled with extended Kalman filter for real-time monitoring, control and automated decision making in mammalian cell culture processing. We show that, in the considered application, the predictive monitoring accuracy of such framework improves by at least 35% when developed with hybrid models with respect to industrial benchmark tools based on PLS models. Additionally, we also highlight the advantages of this approach in industrial applications related to conditional process feeding and process monitoring. With regards to the latter, for an industrial use case we demonstrate that the application of Hybrid-EKF as a soft sensor for titer shows a 50% improvement in prediction accuracy compared to state-of-the-art soft sensor tools. This article is protected by copyright. All rights reserved.

[1]  R. E. Kalman,et al.  A New Approach to Linear Filtering and Prediction Problems , 2002 .

[2]  Manuel Remelhe,et al.  Between the Poles of Data‐Driven and Mechanistic Modeling for Process Operation , 2017 .

[3]  Rimvydas Simutis,et al.  Fuzzy-aided neural network for real-time state estimation and process prediction in the alcohol formation step of production-scale beer brewing , 1993 .

[4]  Denis Dochain,et al.  Review and classification of recent observers applied in chemical process systems , 2015, Comput. Chem. Eng..

[5]  S. D. Jong.  SIMPLS: an alternative approach to partial least squares regression , 1993 .

[6]  H. Pataki,et al.  On‐line prediction of the glucose concentration of CHO cell cultivations by NIR and Raman spectroscopy: Comparative scalability test with a shake flask model system , 2017, Journal of pharmaceutical and biomedical analysis.

[7]  W Y Svrcek,et al.  The extended Kalman filter applied to a continuous culture model , 1974, Biotechnology and bioengineering.

[8]  Robert Gustavsson,et al.  Mini‐review: soft sensors as means for PAT in the manufacture of bio‐therapeutics , 2015 .

[9]  Ana P. Teixeira,et al.  Advances in on-line monitoring and control of mammalian cell cultures: Supporting the PAT initiative. , 2009, Biotechnology advances.

[10]  Gerald Striedner,et al.  Quality by control: Towards model predictive control of mammalian cell culture bioprocesses. , 2017, Biotechnology journal.

[11]  Cleo Kontoravdi,et al.  A multi‐pronged investigation into the effect of glucose starvation and culture duration on fed‐batch CHO cell culture , 2015, Biotechnology and bioengineering.

[12]  Brian Glennon,et al.  Process model comparison and transferability across bioreactor scales and modes of operation for a mammalian cell bioprocess , 2013, Biotechnology progress.

[13]  Martin F. Luna,et al.  Combining Mechanistic Modeling and Raman Spectroscopy for Monitoring Antibody Chromatographic Purification , 2019, Processes.

[14]  Sandro Macchietto,et al.  Model-based design of experiments for parameter precision: State of the art , 2008 .

[15]  Hector M Budman,et al.  Development of a soft‐sensor based on multi‐wavelength fluorescence spectroscopy and a dynamic metabolic model for monitoring mammalian cell cultures , 2014, Biotechnology and bioengineering.

[16]  J. A. Wilson,et al.  Monitoring bioprocesses using hybrid models and an extended Kalman filter , 1996 .

[17]  Yao-ming Huang,et al.  Identifying and eliminating cell culture process variability , 2014 .

[18]  Tao Yu,et al.  On-line Estimation in Fed-batch Fermentation Process Using State Space Model and Unscented Kalman Filter , 2010 .

[19]  S. Feyo de Azevedo,et al.  Hybrid modelling of biochemical processes: A comparison with the conventional approach , 1997 .

[20]  Rimvydas Simutis,et al.  A fuzzy-supported Extended Kalman Filter: a new approach to state estimation and prediction exemplified by alcohol formation in beer brewing , 1992 .

[21]  Rui Oliveira,et al.  Hybrid modeling for quality by design and PAT-benefits and challenges of applications in biopharmaceutical industry. , 2014, Biotechnology journal.

[22]  Zizhuo Xing,et al.  Modeling kinetics of a large‐scale fed‐batch CHO cell culture by Markov chain Monte Carlo method , 2010, Biotechnology progress.

[23]  Hamidreza Mehdizadeh,et al.  Generic Raman‐based calibration models enabling real‐time monitoring of cell culture bioreactors , 2015, Biotechnology progress.

[24]  Rimvydas Simutis,et al.  Using measurement data in bioprocess modelling and control , 1994 .

[25]  Rimvydas Simutis,et al.  Control of cultivation processes for recombinant protein production: a review , 2008, Bioprocess and biosystems engineering.

[26]  Christos Georgakis,et al.  How To NOT Make the Extended Kalman Filter Fail , 2013 .

[27]  Laurent Dewasme,et al.  Experimental validation of an Extended Kalman Filter estimating acetate concentration in E. coli cultures , 2013 .

[28]  Kannan M. Moudgalya,et al.  Multi-criterion control of a bioprocess in fed-batch reactor using EKF based economic model predictive control , 2018 .

[29]  Alan G. Ryder,et al.  Performance monitoring of a mammalian cell based bioprocess using Raman spectroscopy. , 2013, Analytica chimica acta.

[30]  Massimo Morbidelli,et al.  Decision Tree‐PLS (DT‐PLS) algorithm for the development of process: Specific local prediction models , 2019, Biotechnology progress.

[31]  M. Morbidelli,et al.  Enhanced process understanding and multivariate prediction of the relationship between cell culture process and monoclonal antibody quality , 2017, Biotechnology progress.

[32]  Reiner Luttmann,et al.  Advanced monitoring and control of pharmaceutical production processes with Pichia pastoris by using Raman spectroscopy and multivariate calibration methods , 2017, Engineering in life sciences.

[33]  Rimvydas Simutis,et al.  Hybrid Approach to State Estimation for Bioprocess Control , 2017, Bioengineering.

[34]  Massimo Morbidelli,et al.  Bioprocessing in the Digital Age: The Role of Process Models , 2019, Biotechnology journal.

[35]  Henry C. Lim,et al.  Fed-Batch Cultures: Phenomena That Favor Fed-Batch Operations , 2013 .

[36]  Sebastião Feyo de Azevedo,et al.  A general hybrid semi-parametric process control framework , 2012 .

[37]  Rimvydas Simutis,et al.  Advanced Supervision of Mammalian Cell Cultures Using Hybrid Process Models , 1995 .

[38]  Alex Eon-Duval,et al.  Application of Quality by Design to the characterization of the cell culture process of an Fc-Fusion protein. , 2012, European journal of pharmaceutics and biopharmaceutics : official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik e.V.

[39]  Brandon Berry,et al.  Cross‐scale predictive modeling of CHO cell culture growth and metabolites using Raman spectroscopy and multivariate analysis , 2015, Biotechnology progress.

[40]  Massimo Morbidelli,et al.  A new generation of predictive models: The added value of hybrid models for manufacturing processes of therapeutic proteins , 2019, Biotechnology and bioengineering.