A bootstrap-aggregated hybrid semi-parametric modeling framework for bioprocess development

Bringing a scalable adaptive hybrid modeling framework closer to industrial use: Application on a multiscale fungal fermentation

Digitalization has paved the way for new paradigms such as digital shadows and digital twins for fermentation processes, opening the door for real-time process monitoring, control, and optimization. With a digital shadow, real-time model adaptation to accommodate complex metabolic phenomena such as metabolic shifts of a process can be monitored. Despite the many benefits of digitalization, the potential has not been fully reached in the industry. This study investigates the development of a digital shadow for a very complex fungal fermentation process in terms of microbial physiology and fermentation operation on pilot-scale at Novonesis and the challenges thereof. The process has historically been difficult to optimize and control due to a lack of offline measurements and an absence of biomass measurements. Pilot-scale and lab-scale fermentations were conducted for model development and validation. With all available pilot-scale data, a data-driven soft sensor was developed to estimate the main substrate concentration (glucose) with a normalized root mean squared error (N-RMSE) of 2%. This robust data-driven soft sensor was able to estimate accurately in lab-scale (volume < 20× pilot) with a N-RMSE of 7.8%. A hybrid soft sensor was developed by combining the data-driven soft sensor with a mass balance to estimate the glycerol and biomass concentrations on pilot-scale data with N-RMSEs of 11% and 21%, respectively. A digital shadow modeling framework was developed by coupling a mechanistic model (MM) with the hybrid soft sensor. The digital shadow modeling framework significantly improved the predictability compared with the MM. The contribution of this study brings the application of digital shadows closer to industrial implementation. It demonstrates the high potential of using this type of modeling framework for scale-up and leads the way to a new generation of in silico-based process development.

Hybrid deep modeling of a CHO-K1 fed-batch process: combining first-principles with deep neural networks

Introduction: Hybrid modeling combining First-Principles with machine learning is becoming a pivotal methodology for Biopharma 4.0 enactment. Chinese Hamster Ovary (CHO) cells, being the workhorse for industrial glycoproteins production, have been the object of several hybrid modeling studies. Most previous studies pursued a shallow hybrid modeling approach based on three-layered Feedforward Neural Networks (FFNNs) combined with macroscopic material balance equations. Only recently, the hybrid modeling field is incorporating deep learning into its framework with significant gains in descriptive and predictive power. Methods: This study compares, for the first time, deep and shallow hybrid modeling in a CHO process development context. Data of 24 fed-batch cultivations of a CHO-K1 cell line expressing a target glycoprotein, comprising 30 measured state variables over time, were used to compare both methodologies. Hybrid models with varying FFNN depths (3-5 layers) were systematically compared using two training methodologies. The classical training is based on the Levenberg-Marquardt algorithm, indirect sensitivity equations and cross-validation. The deep learning is based on the Adaptive Moment Estimation Method (ADAM), stochastic regularization and semidirect sensitivity equations. Results and conclusion: The results point to a systematic generalization improvement of deep hybrid models over shallow hybrid models. Overall, the training and testing errors decreased by 14.0% and 23.6% respectively when applying the deep methodology. The Central Processing Unit (CPU) time for training the deep hybrid model increased by 31.6% mainly due to the higher FFNN complexity. The final deep hybrid model is shown to predict the dynamics of the 30 state variables within the error bounds in every test experiment. Notably, the deep hybrid model could predict the metabolic shifts in key metabolites (e.g., lactate, ammonium, glutamine and glutamate) in the test experiments. We expect deep hybrid modeling to accelerate the deployment of high-fidelity digital twins in the biopharma sector in the near future.

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.

From Spatial-Temporal Multiscale Modeling to Application: Bridging the Valley of Death in Industrial Biotechnology

The Valley of Death confronts industrial biotechnology with a significant challenge to the commercialization of products. Fortunately, with the integration of computation, automation and artificial intelligence (AI) technology, the industrial biotechnology accelerates to cross the Valley of Death. The Fourth Industrial Revolution (Industry 4.0) has spurred advanced development of intelligent biomanufacturing, which has evolved the industrial structures in line with the worldwide trend. To achieve this, intelligent biomanufacturing can be structured into three main parts that comprise digitalization, modeling and intellectualization, with modeling forming a crucial link between the other two components. This paper provides an overview of mechanistic models, data-driven models and their applications in bioprocess development. We provide a detailed elaboration of the hybrid model and its applications in bioprocess engineering, including strain design, process control and optimization, as well as bioreactor scale-up. Finally, the challenges and opportunities of biomanufacturing towards Industry 4.0 are also discussed.

Machine learning in bioprocess development: from promise to practice

Fostered by novel analytical techniques, digitalization, and automation, modern bioprocess development provides large amounts of heterogeneous experimental data, containing valuable process information. In this context, data-driven methods like machine learning (ML) approaches have great potential to rationally explore large design spaces while exploiting experimental facilities most efficiently. Herein we demonstrate how ML methods have been applied so far in bioprocess development, especially in strain engineering and selection, bioprocess optimization, scale-up, monitoring, and control of bioprocesses. For each topic, we will highlight successful application cases, current challenges, and point out domains that can potentially benefit from technology transfer and further progress in the field of ML.

A General Hybrid Modeling Framework for Systems Biology Applications: Combining Mechanistic Knowledge with Deep Neural Networks under the SBML Standard

In this paper, a computational framework is proposed that merges mechanistic modeling with deep neural networks obeying the Systems Biology Markup Language (SBML) standard. Over the last 20 years, the systems biology community has developed a large number of mechanistic models that are currently stored in public databases in SBML. With the proposed framework, existing SBML models may be redesigned into hybrid systems through the incorporation of deep neural networks into the model core, using a freely available python tool. The so-formed hybrid mechanistic/neural network models are trained with a deep learning algorithm based on the adaptive moment estimation method (ADAM), stochastic regularization and semidirect sensitivity equations. The trained hybrid models are encoded in SBML and uploaded in model databases, where they may be further analyzed as regular SBML models. This approach is illustrated with three well-known case studies: the Escherichia coli threonine synthesis model, the P58IPK signal transduction model, and the Yeast glycolytic oscillations model. The proposed framework is expected to greatly facilitate the widespread use of hybrid modeling techniques for systems biology applications.

SBML2HYB: a Python interface for SBML compatible hybrid modeling

SUMMARY

Here, we present sbml2hyb, an easy-to-use standalone Python tool that facilitates the conversion of existing mechanistic models of biological systems in Systems Biology Markup Language (SBML) into hybrid semiparametric models that combine mechanistic functions with machine learning (ML). The so-formed hybrid models can be trained and stored back in databases in SBML format. The tool supports a user-friendly export interface with an internal format validator. Two case studies illustrate the use of the sbml2hyb tool. Additionally, we describe HMOD, a new model format designed to support and facilitate hybrid models building. It aggregates the mechanistic model information with the ML information and follows as close as possible the SBML rules. We expect the sbml2hyb tool and HMOD to greatly facilitate the widespread usage of hybrid modeling techniques for biological systems analysis.

AVAILABILITY AND IMPLEMENTATION

The Python interface, source code and the example models used for the case studies are accessible at: https://github.com/r-costa/sbml2hyb.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A reinforcement learning-based hybrid modeling framework for bioprocess kinetics identification

Constructing predictive models to simulate complex bioprocess dynamics, particularly time-varying (i.e., parameters varying over time) and history-dependent (i.e., current kinetics dependent on historical culture conditions) behavior, has been a longstanding research challenge. Current advances in hybrid modeling offer a solution to this by integrating kinetic models with data-driven techniques. This article proposes a novel two-step framework: first (i) speculate and combine several possible kinetic model structures sourced from process and phenomenological knowledge, then (ii) identify the most likely kinetic model structure and its parameter values using model-free Reinforcement Learning (RL). Specifically, Step 1 collates feasible history-dependent model structures, then Step 2 uses RL to simultaneously identify the correct model structure and the time-varying parameter trajectories. To demonstrate the performance of this framework, a range of in-silico case studies were carried out. The results show that the proposed framework can efficiently construct high-fidelity models to quantify both time-varying and history-dependent kinetic behaviors while minimizing the risks of over-parametrization and over-fitting. Finally, the primary advantages of the proposed framework and its limitation were thoroughly discussed in comparison to other existing hybrid modeling and model structure identification techniques, highlighting the potential of this framework for general bioprocess modeling.

Industrial data science – a review of machine learning applications for chemical and process industries

Understand and optimize industrial processes via machine learning and chemical engineering principles.

A transfer learning approach for predictive modeling of bioprocesses using small data

Predictive modeling of new biochemical systems with small data is a great challenge. To fill this gap, transfer learning, a subdomain of machine learning that serves to transfer knowledge from a generalized model to a more domain-specific model, provides a promising solution. While transfer learning has been used in natural language processing, image analysis, and chemical engineering fault detection, its application within biochemical engineering has not been systematically explored. In this study, we demonstrated the benefits of transfer learning when applied to predict dynamic behaviors of new biochemical processes. Two different case studies were presented to investigate the accuracy, reliability, and advantage of this innovative modeling approach. We thoroughly discussed the different transfer learning strategies and the effects of topology on transfer learning, comparing the performance of the transfer learning models against benchmark kinetic and data-driven models. Furthermore, strong connections between the underlying process mechanism and the transfer learning model's optimal structure were highlighted, suggesting the interpretability of transfer learning to enable more accurate prediction than a naive data-driven modeling approach. Therefore, this study shows a novel approach to effectively combining data from different resources for bioprocess simulation.

Kinetic and hybrid modeling for yeast astaxanthin production under uncertainty

Astaxanthin is a high-value compound commercially synthesized through Xanthophyllomyces dendrorhous fermentation. Using mixed sugars decomposed from biowastes for yeast fermentation provides a promising option to improve process sustainability. However, little effort has been made to investigate the effects of multiple sugars on X. dendrorhous biomass growth and astaxanthin production. Furthermore, the construction of a high-fidelity model is challenging due to the system's variability, also known as batch-to-batch variation. Two innovations are proposed in this study to address these challenges. First, a kinetic model was developed to compare process kinetics between the single sugar (glucose) based and the mixed sugar (glucose and sucrose) based fermentation methods. Then, the kinetic model parameters were modeled themselves as Gaussian processes, a probabilistic machine learning technique, to improve the accuracy and robustness of model predictions. We conclude that although the presence of sucrose does not affect the biomass growth kinetics, it introduces a competitive inhibitory mechanism that enhances astaxanthin accumulation by inducing adverse environmental conditions such as osmotic gradients. Moreover, the hybrid model was able to greatly reduce model simulation error and was particularly robust to uncertainty propagation. This study suggests the advantage of mixed sugar-based fermentation and provides a novel approach for bioprocess dynamic modeling.

Generic and specific recurrent neural network models: Applications for large and small scale biopharmaceutical upstream processes

The calculation of temporally varying upstream process outcomes is a challenging task. Over the last years, several parametric, semi-parametric as well as non-parametric approaches were developed to provide reliable estimates for key process parameters. We present generic and product-specific recurrent neural network (RNN) models for the computation and study of growth and metabolite-related upstream process parameters as well as their temporal evolution. Our approach can be used for the control and study of single product-specific large-scale manufacturing runs as well as generic small-scale evaluations for combined processes and products at development stage. The computational results for the product titer as well as various major upstream outcomes in addition to relevant process parameters show a high degree of accuracy when compared to experimental data and, accordingly, a reasonable predictive capability of the RNN models. The calculated values for the root-mean squared errors of prediction are significantly smaller than the experimental standard deviation for the considered process run ensembles, which highlights the broad applicability of our approach. As a specific benefit for platform processes, the generic RNN model is also used to simulate process outcomes for different temperatures in good agreement with experimental results. The high level of accuracy and the straightforward usage of the approach without sophisticated parameterization and recalibration procedures highlight the benefits of the RNN models, which can be regarded as promising alternatives to existing parametric and semi-parametric methods.

About Model Validation in Bioprocessing

In bioprocess engineering the Qualtiy by Design (QbD) initiative encourages the use of models to define design spaces. However, clear guidelines on how models for QbD are validated are still missing. In this review we provide a comprehensive overview of the validation methods, mathematical approaches, and metrics currently applied in bioprocess modeling. The methods cover analytics for data used for modeling, model training and selection, measures for predictiveness, and model uncertainties. We point out the general issues in model validation and calibration for different types of models and put this into the context of existing health authority recommendations. This review provides a starting point for developing a guide for model validation approaches. There is no one-fits-all approach, but this review should help to identify the best fitting validation method, or combination of methods, for the specific task and the type of bioprocess model that is being developed.

Towards in silico Process Modeling for Vaccines

Chemical, manufacturing, and control development timelines occupy a significant part of vaccine end-to-end development. In the on-going race for accelerating timelines, in silico process development constitutes a viable strategy that can be achieved through an artificial intelligence (AI)-driven or a mechanistically oriented approach. In this opinion, we focus on the mechanistic option and report on the modeling competencies required to achieve it. By inspecting the most frequent vaccine process units, we identify fluid mechanics, thermodynamics and transport phenomena, intracellular modeling, hybrid modeling and data science, and model-based design of experiments as the pillars for vaccine development. In addition, we craft a generic pathway for accommodating the modeling competencies into an in silico process development strategy.

Monitoring and control of E. coli cell integrity

Soluble expression of recombinant proteins in E. coli is often done by translocation of the product across the inner membrane (IM) into the periplasm, where it is retained by the outer membrane (OM). While the integrity of the IM is strongly coupled to viability and impurity release, a decrease in OM integrity (corresponding to increased "leakiness") leads to accumulation of product in the extracellular space, strongly impacting the downstream process. Whether leakiness is desired or not, differential monitoring and control of IM and OM integrity are necessary for an efficient E. coli bioprocess in compliance with the guidelines of Quality by Design and Process Analytical Technology. In this review, we give an overview of relevant monitoring tools, summarize the research on factors affecting E. coli membrane integrity and provide a brief discussion on how the available monitoring technology can be implemented in real-time control of E. coli cultivations.

Development of a hybrid model for a partially known intracellular signaling pathway through correction term estimation and neural network modeling

Developing an accurate first-principle model is an important step in employing systems biology approaches to analyze an intracellular signaling pathway. However, an accurate first-principle model is difficult to be developed since it requires in-depth mechanistic understandings of the signaling pathway. Since underlying mechanisms such as the reaction network structure are not fully understood, significant discrepancy exists between predicted and actual signaling dynamics. Motivated by these considerations, this work proposes a hybrid modeling approach that combines a first-principle model and an artificial neural network (ANN) model so that predictions of the hybrid model surpass those of the original model. First, the proposed approach determines an optimal subset of model states whose dynamics should be corrected by the ANN by examining the correlation between each state and outputs through relative order. Second, an L2-regularized least-squares problem is solved to infer values of the correction terms that are necessary to minimize the discrepancy between the model predictions and available measurements. Third, an ANN is developed to generalize relationships between the values of the correction terms and the system dynamics. Lastly, the original first-principle model is coupled with the developed ANN to finalize the hybrid model development so that the model will possess generalized prediction capabilities while retaining the model interpretability. We have successfully validated the proposed methodology with two case studies, simplified apoptosis and lipopolysaccharide-induced NFκB signaling pathways, to develop hybrid models with in silico and in vitro measurements, respectively.

An intracellular signaling pathway is often represented by a set of nonlinear ordinary differential equations, which translate our current knowledge about the signaling pathway into a testable mathematical model. However, predictions from such models are often subject to high uncertainty since many signaling pathways are only partially known beforehand. In this study, we propose a systematic approach to develop a hybrid model to improve model accuracy by combining machine learning and the first-principle modeling. Specifically, model correction terms are learned from discrepancy between model predictions and measurements, and these terms are added to the first-principle model to enhance the prediction accuracy. Once these correction terms are learned from the data, an artificial neural network (ANN) model is developed to find an empirical relation between the model and the correction terms so that the developed ANN can be used to posses improved predictive capabilities even in new operating conditions (i.e., generalizability). The final hybrid model is then constructed by coupling the first-principle model with the developed ANN.

Recent Developments in Bioprocessing of Recombinant Proteins: Expression Hosts and Process Development

Infectious diseases, along with cancers, are among the main causes of death among humans worldwide. The production of therapeutic proteins for treating diseases at large scale for millions of individuals is one of the essential needs of mankind. Recent progress in the area of recombinant DNA technologies has paved the way to producing recombinant proteins that can be used as therapeutics, vaccines, and diagnostic reagents. Recombinant proteins for these applications are mainly produced using prokaryotic and eukaryotic expression host systems such as mammalian cells, bacteria, yeast, insect cells, and transgenic plants at laboratory scale as well as in large-scale settings. The development of efficient bioprocessing strategies is crucial for industrial production of recombinant proteins of therapeutic and prophylactic importance. Recently, advances have been made in the various areas of bioprocessing and are being utilized to develop effective processes for producing recombinant proteins. These include the use of high-throughput devices for effective bioprocess optimization and of disposable systems, continuous upstream processing, continuous chromatography, integrated continuous bioprocessing, Quality by Design, and process analytical technologies to achieve quality product with higher yield. This review summarizes recent developments in the bioprocessing of recombinant proteins, including in various expression systems, bioprocess development, and the upstream and downstream processing of recombinant proteins.