Search

Bioprocess modeling has become a useful tool for prediction of the process future with the aim to deduce operating decisions (e.g. transfer or feeds). Due to variabilities, which often occur between and within batches, updating (re-estimation) of model parameters is required at certain time intervals (dynamic parameter estimation) to obtain reliable predictions. This can be challenging in the presence of low sampling frequencies (e.g. every 24 h), different consecutive scales and large measurement errors, as in the case of cell culture seed trains. This contribution presents an iterative learning workflow which generates and incorporates knowledge concerning cell growth during the process by using a moving horizon estimation (MHE) approach for updating of model parameters. This estimation technique is compared to a classical weighted least squares estimation (WLSE) approach in the context of model updating over three consecutive cultivation scales (40–2160 L) of an industrial cell culture seed train. Both techniques were investigated regarding robustness concerning the aforementioned challenges and the required amount of experimental data (estimation horizon). It is shown how the proposed MHE can deal with the aforementioned difficulties by the integration of prior knowledge, even if only data at two sampling points are available, outperforming the classical WLSE approach. This workflow allows to adequately integrate current process behavior into the model and can therefore be a suitable component of a digital twin. Bioprocess modeling has become a useful tool for prediction of the process future with the aim to deduce operating decisions (e.g. transfer or feeds). Due to variabilities, which often occur between and within batches, updating (re-estimation) of model parameters is required at certain time intervals (dynamic parameter estimation) to obtain reliable predictions. This can be challenging in the presence of low sampling frequencies (e.g. every 24 h), different consecutive scales and large measurement errors, as in the case of cell culture seed trains. This contribution presents an iterative learning workflow which generates and incorporates knowledge concerning cell growth during the process by using a moving horizon estimation (MHE) approach for updating of model parameters. This estimation technique is compared to a classical weighted least squares estimation (WLSE) approach in the context of model updating over three consecutive cultivation scales (40–2160 L) of an industrial cell culture seed train. Both techniques were investigated regarding robustness concerning the aforementioned challenges and the required amount of experimental data (estimation horizon). It is shown how the proposed MHE can deal with the aforementioned difficulties by the integration of prior knowledge, even if only data at two sampling points are available, outperforming the classical WLSE approach. This workflow allows to adequately integrate current process behavior into the model and can therefore be a suitable component of a digital twin.

Abstract Kinetic growth models are a useful tool for a better understanding of microalgal cultivation and for optimizing cultivation conditions. The evaluation of such models requires experimental data that is laborious to generate in bioreactor settings. The experimental shake flask setting used in this study allows to run 12 experiments at the same time, with 6 individual light intensities and light durations. This way, 54 biomass data sets were generated for the cultivation of the microalgae Chlorella vulgaris. To identify the model parameters, a stepwise parameter estimation procedure was applied. First, light-associated model parameters were estimated using additional measurements of local light intensities at differ heights within medium at different biomass concentrations. Next, substrate related model parameters were estimated, using experiments for which biomass and nitrate data were provided. Afterwards, growth-related model parameters were estimated by application of an extensive cross validation procedure. Graphic abstract

Bioprocess development and optimization are still cost- and time-intensive due to the enormous number of experiments involved. In this study, the recently introduced model-assisted Design of Experiments (mDoE) concept (Möller et al. in Bioproc Biosyst Eng 42(5):867, 10.1007/s00449-019-02089-7, 2019) was extended and implemented into a software (“mDoE-toolbox”) to significantly reduce the number of required cultivations. The application of the toolbox is exemplary shown in two case studies with Saccharomyces cerevisiae. In the first case study, a fed-batch process was optimized with respect to the pH value and linearly rising feeding rates of glucose and nitrogen source. Using the mDoE-toolbox, the biomass concentration was increased by 30% compared to previously performed experiments. The second case study was the whole-cell biocatalysis of ethyl acetoacetate (EAA) to (S)-ethyl-3-hydroxybutyrate (E3HB), for which the feeding rates of glucose, nitrogen source, and EAA were optimized. An increase of 80% compared to a previously performed experiment with similar initial conditions was achieved for the E3HB concentration.

Model-based concepts and simulation techniques in combination with digital tools emerge as a key to explore the full potential of biopharmaceutical production processes, which contain several challenging development and process steps. One of these steps is the time- and cost-intensive cell proliferation process (also called seed train) to increase cell number from cell thawing up to production scale. Challenges like complex cell metabolism, batch-to-batch variation, variabilities in cell behavior, and influences of changes in cultivation conditions necessitate adequate digital solutions to provide information about the current and near future process state to derive correct process decisions.For this purpose digital seed train twins have proved to be efficient, which digitally display the time-dependent behavior of important process variables based on mathematical models, strategies, and adaption procedures.This chapter will outline the needs for digitalization of seed trains, the construction of a digital seed train twin, the role of parameter estimation, and different statistical methods within this context, which are applicable to several problems in the field of bioprocessing. The results of a case study are presented to illustrate a Bayesian approach for parameter estimation and prediction of an industrial cell culture seed train for seed train digitalization. This chapter outlines the needs for digitalization of cell proliferation processes (seed trains), the construction of a digital seed train twin as well as the role of parameter estimation and different statistical methods within this context, which are applicable to several problems in the field of bioprocessing. The results of a case study are presented to illustrate a Bayesian approach for parameter estimation and prediction of an industrial cell culture seed, as an example for seed train digitalization. It has been shown in which way prior knowledge and input uncertainty can be considered and be propagated to predictive uncertainty.

Model-based concepts and simulation techniques in combination with digital tools emerge as a key to explore the full potential of biopharmaceutical production processes, which contain several challenging development and process steps. One of these steps is the time- and cost-intensive cell proliferation process (also called seed train) to increase cell number from cell thawing up to production scale. Challenges like complex cell metabolism, batch-to-batch variation, variabilities in cell behavior, and influences of changes in cultivation conditions necessitate adequate digital solutions to provide information about the current and near future process state to derive correct process decisions.For this purpose digital seed train twins have proved to be efficient, which digitally display the time-dependent behavior of important process variables based on mathematical models, strategies, and adaption procedures.This chapter will outline the needs for digitalization of seed trains, the construction of a digital seed train twin, the role of parameter estimation, and different statistical methods within this context, which are applicable to several problems in the field of bioprocessing. The results of a case study are presented to illustrate a Bayesian approach for parameter estimation and prediction of an industrial cell culture seed train for seed train digitalization. This chapter outlines the needs for digitalization of cell proliferation processes (seed trains), the construction of a digital seed train twin as well as the role of parameter estimation and different statistical methods within this context, which are applicable to several problems in the field of bioprocessing. The results of a case study are presented to illustrate a Bayesian approach for parameter estimation and prediction of an industrial cell culture seed, as an example for seed train digitalization. It has been shown in which way prior knowledge and input uncertainty can be considered and be propagated to predictive uncertainty.

In dieser Studie werden belastbare Prognoseintervalle der wochentlichen Gesamtarbeitszeit von aktiven vollzeitbeschaftigten Universitatsprofessorinnen und -professoren in Deutschland aus Daten einer Umfrage aus dem Jahre 2016 und a‑priori Informationen aus fruheren Studien bestimmt. Neben der Gesamtarbeitszeit werden auch Teilarbeitszeiten zum Beispiel fur Lehre und Forschung ermittelt. Das Ziel, ein moglichst detailreiches Bild der Arbeitszeit fur verschiedene Tatigkeiten und Fachergruppen zu erhalten, wird mit der methodologischen Fragestellung verbunden, inwieweit frequentistische und Bayesianische Analyse in diesem Beispiel zu ahnlichen Ergebnissen fuhren. Aus den gultigen Fragebogen von aktiven Vollzeit arbeitenden Universitatsprofessorinnen und -professoren ergeben sich bei der direkten Schatzung 56 h fur die durchschnittliche wochentliche Gesamtarbeitszeit und 95 %-Prognoseintervalle innerhalb $$[35,\,80]$$ h. Frequentistische und Bayesianische Analyse fuhren zu ahnlichen Ergebnissen, Fachergruppen und Geschlechter unterscheiden sich wenig. Wird die Gesamtarbeitszeit als Summe der Arbeitszeiten fur Teilaufgaben geschatzt, fuhrt dies zu einem wesentlich groseren Mittelwert von 63 h und deutlich unterschiedlichen 95 %-Prognoseintervallen im Bayesianischen Fall mit $$[43,\,85]$$ h und im frequentistischen Fall mit ungefahr $$[27,\,114]$$ h. Messungen fur die Gesamtarbeitszeit aus unabhangig voneinander ermittelten Teilarbeitszeiten erscheinen deshalb nur verlasslich, wenn eine Bayesianische Analyse mit Vorinformationen uber die Gesamtarbeitszeit durchgefuhrt wird, denn offenbar sind Summen von Teilarbeitszeiten tendenziell groser als eine Gesamtarbeitszeitschatzung, sowohl im Mittel als auch in der Variation. Mogliche Grunde sind unvollstandige Abgrenzungen zwischen den Teilaufgaben, fehlende Ubersicht uber die insgesamt angegebene Arbeitszeit, weil kein Summenzahler wahrend des Ausfullens des Fragebogens mitgefuhrt wird, oder eine Unterschatzung der Gesamtarbeitszeit, wenn diese nur summarisch abgefragt wird. Der Anteil forschungsnaher Tatigkeiten an der Arbeitszeit erscheint mit etwa 60 % deutlich hoher als der Anteil von Lehre, Betreuung und Prufung von Studierenden mit 23 % und der Anteil administrativer Tatigkeiten mit 17 %. Die grosten signifikanten Differenzen in den Erwartungswerten der Fachergruppen treten immer zwischen den Geistes‑/Sozialwissenschaften und einer der anderen Fachergruppen auf, sowohl bei der Gesamtarbeitszeit als auch bei Teilarbeitszeiten. Der Unterschied zwischen dem erwarteten Gesamtarbeitsaufwand von Professorinnen und Professoren ist eher klein.

For the production of biopharmaceuticals, a procedure called seed train or inoculum train is required to generate an adequate number of cells for the inoculation of the production bioreactor. This seed train is time- and cost-intensive but offers potential for optimization. A method and a protocol are described for seed train mapping, directed modeling, and simulation as well as its optimization regarding selected optimization criteria such as optimal points in time for cell passaging. Furthermore, the method can also be applied for the transfer of a seed train to a different production plant or the design of a new seed train, for example, for a new cell line. Another application is to support the selection of the optimal clone for a new process. Seed train prediction can be performed for different clones, and so it can be analyzed how the seed train protocol would look like and for which clones a suitable seed train protocol could be found.Although the chapter is directed toward suspension cell lines, the method is also generally applicable, e.g., for adherent cell lines.

Cell culture technology has become a substantial domain of modern biotechnology, particularly in the pharmaceutical market. Today, products manufactured from cells itself dominate the biopharmaceutical industry. In addition, a limited number of products made of in vitro cultivated cells for regenerative medicine were launched to the market. Modeling of such processes is an important task since these systems are usually nonlinear and complex. In this chapter, a framework for the estimation of process model parameters and its implementation is shown. It is aimed to support the parameter estimation task, which increases the potential of implementation and improvement of mathematical process models into the novel and existing bioprocesses. Apart from the parameter estimation, evaluation of the estimated parameters plays an essential role in order to verify these parameters and subsequently the selected model. The workflow is outlined and shown specifically on the basis of a mathematical process model describing a mammalian cell culture batch process.

Cell culture technology has become a substantial domain of modern biotechnology, particularly in the pharmaceutical market. Today, products manufactured from cells itself dominate the biopharmaceutical industry. In addition, a limited number of products made of in vitro cultivated cells for regenerative medicine were launched to the market. Modeling of such processes is an important task since these systems are usually nonlinear and complex. In this chapter, a framework for the estimation of process model parameters and its implementation is shown. It is aimed to support the parameter estimation task, which increases the potential of implementation and improvement of mathematical process models into the novel and existing bioprocesses. Apart from the parameter estimation, evaluation of the estimated parameters plays an essential role in order to verify these parameters and subsequently the selected model. The workflow is outlined and shown specifically on the basis of a mathematical process model describing a mammalian cell culture batch process.

For the production of biopharmaceuticals, a procedure called seed train or inoculum train is required to generate an adequate number of cells for the inoculation of the production bioreactor. This seed train is time- and cost-intensive but offers potential for optimization. A method and a protocol are described for seed train mapping, directed modeling, and simulation as well as its optimization regarding selected optimization criteria such as optimal points in time for cell passaging. Furthermore, the method can also be applied for the transfer of a seed train to a different production plant or the design of a new seed train, for example, for a new cell line. Another application is to support the selection of the optimal clone for a new process. Seed train prediction can be performed for different clones, and so it can be analyzed how the seed train protocol would look like and for which clones a suitable seed train protocol could be found.Although the chapter is directed toward suspension cell lines, the method is also generally applicable, e.g., for adherent cell lines.

Reliable scale-up of biopharmaceutical production processes is key in Quality by Design. In this study, a model-based workflow is described to evaluate the bioprocess dynamics during process transfer and scale-up computationally. First, a mathematical model describes the bioprocess dynamics of different state variables (e.g., cell density, titer). Second, the model parameter probability distributions are de- termined at different scales due to measurement uncertainty. Third, the quantified parameter distribu- tions are statistically compared to evaluate if the process dynamics have been changed. This workflow was tested for the scale-up of an antibody-producing CHO fed-batch process. Significant differences were identified between the process development (30 ml) and implementation (250 ml) scale, and the feeding strategy was validated using model-assisted Design of Experiments. Then, the validated process strategy was successfully scaled up to 2 l laboratory and 50 l pilot scale. In summary, the proposed workflow enables a knowledge-driven evaluation tool for bioprocess development. ©2020 Elsevier Ltd. All rights reserved.

For the manufacturing of complex biopharmaceuticals using bioreactors with cultivated mammalian cells, high product concentration is an important objective. The phenotype of the cells in a reactor plays an important role. Are clonal cell populations showing high cell-specific growth rates more favorable than cell lines with higher cell-specific productivities or vice versa? Five clonal Chinese hamster ovary cell populations were analyzed based on the data of a 3-month-stability study. We adapted a mechanistic cell culture model to the experimental data of one such clonally derived cell population. Uncertainties and prior knowledge concerning model parameters were considered using Bayesian parameter estimations. This model was used then to define an inoculum train protocol. Based on this, we subsequently simulated the impacts of differences in growth rates (±10%) and production rates (±10% and ±50%) on the overall cultivation time, including making the inoculum train cultures; the final production phase, the volumetric titer in that bioreactor and the ratio of both, defined as overall process productivity. We showed thus unequivocally that growth rates have a higher impact (up to three times) on overall process productivity and for product output per year, whereas cells with higher productivity can potentially generate higher product concentrations in the production vessel.

The purpose of a seed train is the generation of an adequate number of cells for the inoculation of a production bioreactor. This is time- and cost-intensive: From volumes used for cell thawing or cell line maintenance the cell number has to be increased. The cells are usually run through many cultivation systems which become larger with each passage (e.g. T-flasks, roller bottles or shake flasks, small scale bioreactor systems and subsequently larger bioreactors. Single-use systems may be applied and systems which are inoculated at a partly filled state and culture volume is increased afterwards by medium addition). The production bioreactor is inoculated out of the largest seed train scale.