Bioprocess systems analysis, modeling, estimation, and control

Antibody glycan quality predicted from CHO cell culture media markers and machine learning

N-glycosylation can have a profound effect on the quality of mAb therapeutics. In biomanufacturing, one of the ways to influence N-glycosylation patterns is by altering the media used to grow mAb cell expression systems. Here, we explore the potential of machine learning (ML) to forecast the abundances of N-glycan types based on variables related to the growth media. The ML models exploit a dataset consisting of detailed glycomic characterisation of Anti-HER fed-batch bioreactor cell cultures measured daily under 12 different culture conditions, such as changes in levels of dissolved oxygen, pH, temperature, and the use of two different commercially available media. By performing spent media quantitation and subsequent calculation of pseudo cell consumption rates (termed media markers) as inputs to the ML model, we were able to demonstrate a small subset of media markers (18 selected out of 167 mass spectrometry peaks) in a Chinese Hamster Ovary (CHO) cell cultures are important to model N-glycan relative abundances (Regression - correlations between 0.80-0.92; Classification - AUC between 75.0-97.2). The performances suggest the ML models can infer N-glycan critical quality attributes from extracellular media as a proxy. Given its accuracy, we envisage its potential applications in biomaufactucuring, especially in areas of process development, downstream and upstream bioprocessing.

Different modeling approaches for inline biochemical monitoring over the VLP-making upstream stages using Raman spectroscopy

This work aimed to set inline Raman spectroscopy models to monitor biochemically (viable cell density, cell viability, glucose, lactate, glutamine, glutamate, and ammonium) all upstream stages of a virus-like particle-making process. Linear (Partial least squares, PLS; Principal components regression, PCR) and nonlinear (Artificial neural networks, ANN; supported vector machine, SVM) modeling approaches were assessed. The nonlinear models, ANN and SVM, were the more suitable models with the lowest absolute errors. The mean absolute error of the best models within the assessed parameter ranges for viable cell density (0.01-8.83 × 106 cells/mL), cell viability (1.3-100.0 %), glucose (5.22-10.93 g/L), lactate (18.6-152.7 mg/L), glutamine (158-1761 mg/L), glutamate (807.6-2159.7 mg/L), and ammonium (62.8-117.8 mg/L) were 1.55 ± 1.37 × 106 cells/mL (ANN), 5.01 ± 4.93 % (ANN), 0.27 ± 0.22 g/L (SVM), 4.7 ± 2.6 mg/L (SVM), 51 ± 49 mg/L (ANN), 57 ± 39 mg/L (SVM) and 2.0 ± 1.8 mg/L (ANN), respectively. The errors achieved, and best-fitted models were like those for the same bioprocess using offline data and others, which utilized inline spectra for mammalian cell lines as a host.

Dynamic Energy Budget model for E. coli growth in carbon and nitrogen limitation conditions

The simulations and predictions obtained from mathematical models of bioprocesses conducted by microorganisms are not overvalued. Mechanistic models are bringing a better process understanding and the possibility of simulating unmeasurable variables. The Dynamic Energy Budget (DEB) model is an energy balance that can be formulated for any living organism and can be classified as a structured model. In this study, the DEB model was used to describe E. coli growth in a batch reactor in carbon and nitrogen substrate limitation conditions. The DEB model provides a possibility to follow the changes in the microbes' cells including their elemental composition and content of some important cell ingredients in different growth phases in substrate limitation conditions which makes it more informative compared to Monod's model. The model can be used as an optimal choice between Monod-like models and flux-based approaches. KEY POINTS: • The DEB model can be used to catch changes in elemental composition of E. coli • Bacteria batch culture growth phases can be explained by the DEB model • The DEB model is more informative compared to Monod's based models.

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

This paper reviews the key building blocks needed to develop a mechanistic model for use as an operational production tool. The Chinese Hamster Ovary (CHO) cell, one of the most widely used hosts for antibody production in the pharmaceutical industry, is considered as a case study. CHO cell metabolism is characterized by two main phases, exponential growth followed by a stationary phase with strong protein production. This process presents an appropriate degree of complexity to outline the modeling strategy. The paper is organized into four main steps: (1) CHO systems and data collection; (2) metabolic analysis; (3) formulation of the mathematical model; and finally, (4) numerical solution, calibration, and validation. The overall approach can build a predictive model of target variables. According to the literature, one of the main current modeling challenges lies in understanding and predicting the spontaneous metabolic shift. Possible candidates for the trigger of the metabolic shift include the concentration of lactate and carbon dioxide. In our opinion, ammonium, which is also an inhibiting product, should be further investigated. Finally, the expected progress in the emerging field of hybrid modeling, which combines the best of mechanistic modeling and machine learning, is presented as a fascinating breakthrough. Note that the modeling strategy discussed here is a general framework that can be applied to any bioprocess.

Machine learning modeling and additive explanation techniques for glutathione production from multiple experimental growth conditions of Saccharomyces cerevisiae

Glutathione (GSH) production is of great industrial interest due to its essential properties. This study aimed to use machine learning (ML) methods to model GSHproduction under different growth conditions of Saccharomyces cerevisiae, namely cultivation time, culture volume, pressure, and magnetic field application. Different ML and regression models were evaluated for their statistics to select the most robust model. Results showed that eXtreme Gradient Boosting (XGB) was the best predictive performance model. From the best model, additive explanation techniques were used to identify the feature importance of process. According to variable analysis, the best conditions to obtain the highest GSH concentrations would be cultivation times of 72-96 h, low magnetic field intensity (3.02 mT), low pressure (0.5 kgf.cm-2), and high culture volume (3.5 L). XGB use and additive explanation techniques proved promising for determining process optimization conditions and selecting the essential process variables.

Construction of a novel kinetic model for the production process of a CVA6 VLP vaccine in CHO cells

Bioprocess development benefits from kinetic models in many aspects, including scale-up, optimization, and process understanding. However, current models are unable to simulate the production process of a coxsackievirus A6 (CVA6) virus-like particle (VLP) vaccine using Chinese hamster ovary cell culture. In this study, a novel kinetic model was constructed, correlating (1) cell growth, death, and lysis kinetics, (2) metabolism of major metabolites, and (3) CVA6 VLP production. To construct the model, two batches of a laboratory-scale 2 L bioreactor cell culture were prepared and various pH shift strategies were applied to examine the effect of pH shift. The proposed model described the experimental data under various conditions with high accuracy and quantified the effect of pH shift. Next, cell culture performance with various pH shift timings was predicted by the calibrated model. A trade-off relationship was found between product yield and quality. Consequently, multiple objective optimization was performed by integrating desirability methodology with model simulation. Finally, the optimal operating conditions that balanced product yield and quality were predicted. In general, the proposed model improved the process understanding and enabled in silico process development of a CVA6 VLP vaccine.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10616-023-00598-8.

How Not to Make the Joint Extended Kalman Filter Fail with Unstructured Mechanistic Models

The unstructured mechanistic model (UMM) allows for modeling the macro-scale of a phenomenon without known mechanisms. This is extremely useful in biomanufacturing because using the UMM for the joint estimation of states and parameters with an extended Kalman filter (JEKF) can enable the real-time monitoring of bioprocesses with unknown mechanisms. However, the UMM commonly used in biomanufacturing contains ordinary differential equations (ODEs) with unshared parameters, weak variables, and weak terms. When such a UMM is coupled with an initial state error covariance matrix P(t=0) and a process error covariance matrix Q with uncorrelated elements, along with just one measured state variable, the joint extended Kalman filter (JEKF) fails to estimate the unshared parameters and state simultaneously. This is because the Kalman gain corresponding to the unshared parameter remains constant and equal to zero. In this work, we formally describe this failure case, present the proof of JEKF failure, and propose an approach called SANTO to side-step this failure case. The SANTO approach consists of adding a quantity to the state error covariance between the measured state variable and unshared parameter in the initial P(t = 0) of the matrix Ricatti differential equation to compute the predicted error covariance matrix of the state and prevent the Kalman gain from being zero. Our empirical evaluations using synthetic and real datasets reveal significant improvements: SANTO achieved a reduction in root-mean-square percentage error (RMSPE) of up to approximately 17% compared to the classical JEKF, indicating a substantial enhancement in estimation accuracy.

Efficient optimization of time-varying inputs in a fed-batch cell culture process using design of dynamic experiments

In cell culture process development, we rely largely on an iterative, one-factor-at-a-time procedure based on experiments that explore a limited process space. Design of experiments (DoE) addresses this issue by allowing us to analyze the effects of process inputs on process responses systematically and efficiently. However, DoE cannot be applied directly to study time-varying process inputs unless an impractically large number of bioreactors is used. Here, we adopt the methodology of design of dynamic experiments (DoDE) and incorporate dynamic feeding profiles efficiently in late-stage process development of the manufacture of therapeutic monoclonal antibodies. We found that, for the specific cell line used in this article, (1) not only can we estimate the effect of nutrient feed amount on various product attributes, but we can also estimate the effect, develop a statistical model, and use the model to optimize the slope of time-trended feed rates; (2) in addition to the slope, higher-order dynamic characteristics of time-trended feed rates can be incorporated in the design but do not have any significant effect on the responses we measured. Based on the DoDE data, we developed a statistical model and used the model to optimize several process conditions. Our effort resulted in a tangible improvement in productivity-compared with the baseline process without dynamic feeding, this optimized process in a 200-L batch achieved a 27% increase in titer and > 92% viability. We anticipate our application of DoDE to be a starting point for more efficient workflows to optimize dynamic process conditions in process development.

Steps toward a digital twin for functional food production with increased health benefits

Lactobacillus rhamnosus (L. rhamnosus) is a commensal bacterium with health-promoting properties and with a wide range of applications within the food industry. To improve and optimize the control of L. rhamnosus biomass production in batch and fed-batch bioprocesses, this study proposes the application of artificial neural network (ANN) modelling to improve process control and monitoring, with potential future implementation as a basis for a digital twin. Three ANNs were developed using historical data from ten bioprocesses. These ANNs were designed to predict the biomass in batch bioprocesses with different media compositions, predict biomass in fed-batch bioprocesses, and predict the growth rate in fed-batch bioprocesses. The immunomodulatory effect of the L. rhamnosus samples was examined and found to elicit an anti-inflammatory response as evidenced by the inhibition of IL-6 and TNF-α secretion. Overall, the findings of this study reinforce the potential of ANN modelling for bioprocess optimization aimed at improved control for maximising the volumetric productivity of L. rhamnosus as an immunomodulatory agent with applications in the functional food industry.

Hybrid modeling in bioprocess dynamics: Structural variabilities, implementation strategies, and practical challenges

Hybrid modeling, with an appropriate blend of the mechanistic and data-driven framework, is increasingly being adopted in bioprocess modeling, model-based experimental design (digital-twin), identification of critical process parameters, and optimization. However, the development of a hybrid model from experimental data is an inherently complex workflow, involving designed experiments, selection of the data-driven process, identification of model parameters, assessment fitness, and generalization capability. Depending on the complexity of the process system and purpose, each piece of these modules can flexibly be incorporated into the puzzle. However, this extra flexibility can be a cause of concern to trace an "optimal" model structure. In this paper, the development of hybrid models in a common bioprocess system, selection of data-driven components and their mapping to states, choice of parameter identification techniques, and model quality assurance are revisited. The challenges associated with hybrid-model development, and corrective actions have also been reviewed. The review also suggests the lack of data, and code sharing in communal repositories can be a hurdle in the exploration, and expansion of those tools in a bioprocess system.

Advances in bioreactor control for production of biotherapeutic products

Advanced control strategies are well established in chemical, pharmaceutical, and food processing industries. Over the past decade, the application of these strategies is being explored for control of bioreactors for manufacturing of biotherapeutics. Most of the industrial bioreactor control strategies apply classical control techniques, with the control system designed for the facility at hand. However, with the recent progress in sensors, machinery, and industrial internet of things, and advancements in deeper understanding of the biological processes, coupled with the requirement of flexible production, the need to develop a robust and advanced process control system that can ease process intensification has emerged. This has further fuelled the development of advanced monitoring approaches, modeling techniques, process analytical technologies, and soft sensors. It is seen that proper application of these concepts can significantly improve bioreactor process performance, productivity, and reproducibility. This review is on the recent advancements in bioreactor control and its related aspects along with the associated challenges. This study also offers an insight into the future prospects for development of control strategies that can be designed for industrial-scale production of biotherapeutic products.

rAAV Manufacturing: The Challenges of Soft Sensing during Upstream Processing

Recombinant adeno-associated virus (rAAV) is the most effective viral vector technology for directly translating the genomic revolution into medicinal therapies. However, the manufacturing of rAAV viral vectors remains challenging in the upstream processing with low rAAV yield in large-scale production and high cost, limiting the generalization of rAAV-based treatments. This situation can be improved by real-time monitoring of critical process parameters (CPP) that affect critical quality attributes (CQA). To achieve this aim, soft sensing combined with predictive modeling is an important strategy that can be used for optimizing the upstream process of rAAV production by monitoring critical process variables in real time. However, the development of soft sensors for rAAV production as a fast and low-cost monitoring approach is not an easy task. This review article describes four challenges and critically discusses the possible solutions that can enable the application of soft sensors for rAAV production monitoring. The challenges from a data scientist's perspective are (i) a predictor variable (soft-sensor inputs) set without AAV viral titer, (ii) multi-step forecasting, (iii) multiple process phases, and (iv) soft-sensor development composed of the mechanistic model.

Bioprocess Economic Modeling: Decision Support Tools for the Development of Stem Cell Therapy Products

Over recent years, the field of cell and gene therapy has witnessed rapid growth due to the demonstrated benefits of using living cells as therapeutic agents in a broad range of clinical studies and trials. Bioprocess economic models (BEMs) are fundamental tools for guiding decision-making in bioprocess design, being capable of supporting process optimization and helping to reduce production costs. These tools are particularly important when it comes to guiding manufacturing decisions and increasing the likelihood of market acceptance of cell-based therapies, which are often cost-prohibitive because of high resource and quality control costs. Not only this, but the inherent biological variability of their underlying bioprocesses makes them particularly susceptible to unforeseen costs arising from failed or delayed production batches. The present work reviews important concepts concerning the development of bioprocesses for stem cell therapy products and highlights the valuable role which BEMs can play in this endeavor. Additionally, some theoretical concepts relevant to the building and structuring of BEMs are explored. Finally, a comprehensive review of the existent BEMs so far reported in the scientific literature for stem cell-related bioprocesses is provided to showcase their potential usefulness.

Optimization and Scale-Up of Fermentation Processes Driven by Models

In the era of sustainable development, the use of cell factories to produce various compounds by fermentation has attracted extensive attention; however, industrial fermentation requires not only efficient production strains, but also suitable extracellular conditions and medium components, as well as scaling-up. In this regard, the use of biological models has received much attention, and this review will provide guidance for the rapid selection of biological models. This paper first introduces two mechanistic modeling methods, kinetic modeling and constraint-based modeling (CBM), and generalizes their applications in practice. Next, we review data-driven modeling based on machine learning (ML), and highlight the application scope of different learning algorithms. The combined use of ML and CBM for constructing hybrid models is further discussed. At the end, we also discuss the recent strategies for predicting bioreactor scale-up and culture behavior through a combination of biological models and computational fluid dynamics (CFD) models.

Model-based intensification of CHO cell cultures: One-step strategy from fed-batch to perfusion

There is a growing interest in continuous processing of the biopharmaceutical industry. However, the technology transfer from traditional batch-based processes is considered a challenge as protocol and tools still remain to be established for their usage at the manufacturing scale. Here, we present a model-based approach to design optimized perfusion cultures of Chinese Hamster Ovary cells using only the knowledge captured during small-scale fed-batch experiments. The novelty of the proposed model lies in the simplicity of its structure. Thanks to the introduction of a new catch-all variable representing a bulk of by-products secreted by the cells during their cultivation, the model was able to successfully predict cellular behavior under different operating modes without changes in its formalism. To our knowledge, this is the first experimentally validated model capable, with a single set of parameters, to capture culture dynamic under different operating modes and at different scales.

Modern Sensor Tools and Techniques for Monitoring, Controlling, and Improving Cell Culture Processes

The growing biopharmaceutical industry has reached a level of maturity that allows for the monitoring of numerous key variables for both process characterization and outcome predictions. Sensors were historically used in order to maintain an optimal environment within the reactor to optimize process performance. However, technological innovation has pushed towards on-line in situ continuous monitoring of quality attributes that could previously only be estimated off-line. These new sensing technologies when coupled with software models have shown promise for unique fingerprinting, smart process control, outcome improvement, and prediction. All this can be done without requiring invasive sampling or intervention on the system. In this paper, the state-of-the-art sensing technologies and their applications in the context of cell culture monitoring are reviewed with emphasis on the coming push towards industry 4.0 and smart manufacturing within the biopharmaceutical sector. Additionally, perspectives as to how this can be leveraged to improve both understanding and outcomes of cell culture processes are discussed.

High Cell Density Culture of Dairy Propionibacterium sp. and Acidipropionibacterium sp.: A Review for Food Industry Applications

The dairy bacteria Propionibacterium sp. and Acidipropionibacterium sp. are versatile and potentially probiotic microorganisms showing outstanding functionalities for the food industry, such as the production of propionic acid and vitamin B12 biosynthesis. They are the only food grade microorganisms able to produce vitamin B12. However, the fermentation batch process using these bacteria present some bioprocess limitations due to strong end-product inhibition, cells slow-growing rates, low product titer, yields and productivities, which reduces the bioprocess prospects for industrial applications. The high cell density culture (HCDC) bioprocess system is known as an efficient approach to overcome most of those problems. The main techniques applied to achieve HCDC of dairy Propionibacterium are the fed-batch cultivation, cell recycling, perfusion, extractive fermentation, and immobilization. In this review, the techniques available and reported to achieve HCDC of Propionibacterium sp. and Acidipropionibacterium sp. are discussed, and the advantages and drawbacks of this system of cultivation in relation to biomass formation, vitamin B12 biosynthesis, and propionic acid production are evaluated.