Simulation and optimization of dynamic flux balance analysis models using an interior point method reformulation

Microbial interactions from a new perspective: reinforcement learning reveals new insights into microbiome evolution

MOTIVATION

Microbes are essential part of all ecosystems, influencing material flow and shaping their surroundings. Metabolic modeling has been a useful tool and provided tremendous insights into microbial community metabolism. However, current methods based on flux balance analysis (FBA) usually fail to predict metabolic and regulatory strategies that lead to long-term survival and stability especially in heterogenous communities.

RESULTS

Here, we introduce a novel reinforcement learning algorithm, Self-Playing Microbes in Dynamic FBA, which treats microbial metabolism as a decision-making process, allowing individual microbial agents to evolve by learning and adapting metabolic strategies for enhanced long-term fitness. This algorithm predicts what microbial flux regulation policies will stabilize in the dynamic ecosystem of interest in the presence of other microbes with minimal reliance on predefined strategies. Throughout this article, we present several scenarios wherein our algorithm outperforms existing methods in reproducing outcomes, and we explore the biological significance of these predictions.

AVAILABILITY AND IMPLEMENTATION

The source code for this article is available at: https://github.com/chan-csu/SPAM-DFBA.

Towards advanced bioprocess optimization: A multiscale modelling approach

Mammalian cells produce up to 80 % of the commercially available therapeutic proteins, with Chinese Hamster Ovary (CHO) cells being the primary production host. Manufacturing involves a train of reactors, the last of which is typically run in fed-batch mode, where cells grow and produce the required protein. The feeding strategy is decided a priori, from either past operations or the design of experiments and rarely considers the current state of the process. This work proposes a Model Predictive Control (MPC) formulation based on a hybrid kinetic-stoichiometric reactor model to provide optimal feeding policies in real-time, which is agnostic to the culture, hence transferable across CHO cell culture systems. The benefits of the proposed controller formulation are demonstrated through a comparison between an open-loop simulation and closed-loop optimization, using a digital twin as an emulator of the process.

A neuro swarm procedure to solve the novel second order perturbed delay Lane-Emden model arising in astrophysics

The current work provides a mathematical second order perturbed singular delay differential model (SO-PSDDM) by using the standard form of the Lane-Emden model. The inclusive structures based on the delay terms, singular-point and perturbation factor and shape forms of the SO-PSDDM are provided. The novel form of the SO-PSDDM is numerically solved by using the procedures of artificial neural networks (ANNs) along with the optimization measures based on the swarming procedures (PSO) and interior-point algorithm (IPA). An error function is optimized through the swarming PSO procedure along with the IPA to solve the SO-PSDDM. The precision, substantiation and validation are observed for three problems of the SO-PSDDM. The exactness of the novel SO-PSDDM is observed by comparing the obtained and exact solutions. The reliability, stability and convergence of the proposed stochastic algorithms are observed for 30 independent trials to solve the novel SO-PSDDM.

Flux balance analysis of metabolic networks for efficient engineering of microbial cell factories

Metabolic engineering principles have long been applied to explore the metabolic networks of complex microbial cell factories under a variety of environmental constraints for effective deployment of the microorganisms in the optimal production of biochemicals like biofuels, polymers, amino acids, recombinant proteins. One of the methodologies used for analyzing microbial metabolic networks is the Flux Balance Analysis (FBA), which employs applications of optimization techniques for forecasting biomass growth and metabolic flux distribution of industrially important products under specified environmental conditions. The in silico flux simulations are instrumental for designing the production-specific microbial cell factories. However, FBA has some inherent limitations. The present review emphasizes how the incorporation of additional kinetic, thermodynamic, expression and regulatory constraints and integration of omics data into the classical FBA platform improve the prediction accuracy of FBA. A programmed comparison of the simulated data with the experimental observations is presented for supporting the claim. The review further accounts for the successful implementation of classical FBA in biotechnological applications and identifies areas in which classical FBA fails to make correct predictions. The analysis of the predictive capabilities of the different FBA strategies presented here is expected to help researchers in finding new avenues in engineering highly efficient microbial metabolic pathways and identify the key metabolic bottlenecks during the process. Based on the appropriate metabolic network design, fermentation engineers will be able to effectively design the bioreactors and optimize large-scale biochemical production through suitable pathway modifications.

An Advance Computing Numerical Heuristic of Nonlinear SIR Dengue Fever System Using the Morlet Wavelet Kernel

This study is associated to solve the nonlinear SIR dengue fever system using a computational methodology by operating the neural networks based on the designed Morlet wavelet (MWNNs), global scheme as genetic algorithm (GA), and rapid local search scheme as interior-point algorithm (IPA), i.e., GA-IPA. The optimization of fitness function based on MWNNs is performed for solving the nonlinear SIR dengue fever system. This MWNNs-based fitness function is accessible using the differential system and initial conditions of the nonlinear SIR dengue fever system. The designed procedures based on the MWNN-GA-IPA are applied to solve the nonlinear SIR dengue fever system to check the exactness, precision, constancy, and efficiency. The achieved numerical form of the nonlinear SIR dengue fever system via MWNN-GA-IPA was compared with the Runge–Kutta numerical results that verify the significance of MWNN-GA-IPA. Moreover, statistical reflections through different measures for the nonlinear SIR dengue fever system endorse the precision and convergence of the computational MWNN-GA-IPA.

Bacterial metabolism and pathogenesis intimate intertwining: time for metabolic modelling to come into action

We take a snapshot of the recent understanding of bacterial metabolism and the bacterial-host metabolic interplay during infection, and highlight key outcomes and challenges for the practical implementation of bacterial metabolic modelling computational tools in the pathogenesis field.

A Systematic Strategy to Find Potential Therapeutic Targets for Pseudomonas aeruginosa Using Integrated Computational Models

Pseudomonas aeruginosa is an opportunistic human pathogen that has been a constant global health problem due to its ability to cause infection at different body sites and its resistance to a broad spectrum of clinically available antibiotics. The World Health Organization classified multidrug-resistant Pseudomonas aeruginosa among the top-ranked organisms that require urgent research and development of effective therapeutic options. Several approaches have been taken to achieve these goals, but they all depend on discovering potential drug targets. The large amount of data obtained from sequencing technologies has been used to create computational models of organisms, which provide a powerful tool for better understanding their biological behavior. In the present work, we applied a method to integrate transcriptome data with genome-scale metabolic networks of Pseudomonas aeruginosa. We submitted both metabolic and integrated models to dynamic simulations and compared their performance with published in vitro growth curves. In addition, we used these models to identify potential therapeutic targets and compared the results to analyze the assumption that computational models enriched with biological measurements can provide more selective and (or) specific predictions. Our results demonstrate that dynamic simulations from integrated models result in more accurate growth curves and flux distribution more coherent with biological observations. Moreover, identifying drug targets from integrated models is more selective as the predicted genes were a subset of those found in the metabolic models. Our analysis resulted in the identification of 26 non-host homologous targets. Among them, we highlighted five top-ranked genes based on lesser conservation with the human microbiome. Overall, some of the genes identified in this work have already been proposed by different approaches and (or) are already investigated as targets to antimicrobial compounds, reinforcing the benefit of using integrated models as a starting point to selecting biologically relevant therapeutic targets.

Combining Kinetic and Constraint-Based Modelling to Better Understand Metabolism Dynamics

To understand the phenotypic capabilities of organisms, it is useful to characterise cellular metabolism through the analysis of its pathways. Dynamic mathematical modelling of metabolic networks is of high interest as it provides the time evolution of the metabolic components. However, it also has limitations, such as the necessary mechanistic details and kinetic parameters are not always available. On the other hand, large metabolic networks exhibit a complex topological structure which can be studied rather efficiently in their stationary regime by constraint-based methods. These methods produce useful predictions on pathway operations. In this review, we present both modelling techniques and we show how they bring complementary views of metabolism. In particular, we show on a simple example how both approaches can be used in conjunction to shed some light on the dynamics of metabolic networks.

Multi-omics Study of Planobispora rosea, Producer of the Thiopeptide Antibiotic GE2270A

Planobispora rosea is the natural producer of the potent thiopeptide antibiotic GE2270A. Here, we present the results of a metabolomics and transcriptomics analysis of P. rosea during production of GE2270A. The data generated provides useful insights into the biology of this genetically intractable bacterium. We characterize the details of the shutdown of protein biosynthesis and the respiratory chain associated with the end of the exponential growth phase. We also provide the first description of the phosphate regulon in P. rosea. Based on the transcriptomics data, we show that both phosphate and iron are limiting P. rosea growth in our experimental conditions. Additionally, we identified and validated a new biosynthetic gene cluster associated with the production of the siderophores benarthin and dibenarthin in P. rosea. Together, the metabolomics and transcriptomics data are used to inform and refine the very first genome-scale metabolic model for P. rosea, which will be a valuable framework for the interpretation of future studies of the biology of this interesting but poorly characterized species.

IMPORTANCEPlanobispora rosea is a genetically intractable bacterium used for the production of GE2270A on an industrial scale. GE2270A is a potent thiopeptide antibiotic currently used as a precursor for the synthesis of two compounds under clinical studies for the treatment of Clostridium difficile infection and acne. Here, we present the very first systematic multi-omics investigation of this important bacterium, which provides a much-needed detailed picture of the dynamics of metabolism of P. rosea while producing GE2270A.

Author Video: An author video summary of this article is available.

An insight to flux-balance analysis for biochemical networks

Systems biology is one of the integrated ways to study biological systems and is more favourable than the earlier used approaches. It includes metabolic pathway analysis, modelling, and regulatory as well as signal transduction for getting insights into cellular behaviour. Among the various techniques of modelling, simulation, analysis of networks and pathways, flux-based analysis (FBA) has been recognised because of its extensibility as well as simplicity. It is widely accepted because it is not like a mechanistic simulation which depends on accurate kinetic data. The study of fluxes through the network is informative and can give insights even in the absence of kinetic data. FBA is one of the widely used tools to study biochemical networks and needs information of reaction stoichiometry, growth requirements, specific measurement parameters of the biological system, in particular the reconstruction of the metabolic network for the genome-scale, many of which have already been built previously. It defines the boundaries of flux distributions which are possible and achievable with a defined set of genes. This review article gives an insight into FBA, from the extension of flux balancing to mathematical representation followed by a discussion about the formulation of flux-balance analysis problems, defining constraints for the stoichiometry of the pathways and the tools that can be used in FBA such as FASIMA, COBRA toolbox, and OptFlux. It also includes broader areas in terms of applications which can be covered by FBA as well as the queries which can be addressed through FBA.

Fast uncertainty quantification for dynamic flux balance analysis using non-smooth polynomial chaos expansions

We present a novel surrogate modeling method that can be used to accelerate the solution of uncertainty quantification (UQ) problems arising in nonlinear and non-smooth models of biological systems. In particular, we focus on dynamic flux balance analysis (DFBA) models that couple intracellular fluxes, found from the solution of a constrained metabolic network model of the cellular metabolism, to the time-varying nature of the extracellular substrate and product concentrations. DFBA models are generally computationally expensive and present unique challenges to UQ, as they entail dynamic simulations with discrete events that correspond to switches in the active set of the solution of the constrained intracellular model. The proposed non-smooth polynomial chaos expansion (nsPCE) method is an extension of traditional PCE that can effectively capture singularities in the DFBA model response due to the occurrence of these discrete events. The key idea in nsPCE is to use a model of the singularity time to partition the parameter space into two elements on which the model response behaves smoothly. Separate PCE models are then fit in both elements using a basis-adaptive sparse regression approach that is known to scale well with respect to the number of uncertain parameters. We demonstrate the effectiveness of nsPCE on a DFBA model of an E. coli monoculture that consists of 1075 reactions and 761 metabolites. We first illustrate how traditional PCE is unable to handle problems of this level of complexity. We demonstrate that over 800-fold savings in computational cost of uncertainty propagation and Bayesian estimation of parameters in the substrate uptake kinetics can be achieved by using the nsPCE surrogates in place of the full DFBA model simulations. We then investigate the scalability of the nsPCE method by utilizing it for global sensitivity analysis and maximum a posteriori estimation in a synthetic metabolic network problem with a larger number of parameters related to both intracellular and extracellular quantities.