Metabolica: a statistical research tool for analyzing metabolic networks

Effect of Suboptimal Neuromuscular Control on the Risk of Massive Wear in Total Knee Replacement

The optimal neuromuscular control (muscle activation strategy that minimises the consumption of metabolic energy) during level walking is very close to that which minimises the force transmitted through the joints of the lower limbs. Thus, any suboptimal control involves an overloading of the joints. Some total knee replacement patients adopt suboptimal control strategies during level walking; this is particularly true for patients with co-morbidities that cause neuromotor control degeneration, such as Parkinson’s Disease (PD). The increase of joint loading increases the risk of implant failure, as reported in one study in PD patients (5.44% of failures at 9 years follow-up). One failure mode that is directly affected by joint loading is massive wear of the prosthetic articular surface. In this study we used a validated patient-specific biomechanical model to estimate how a severely suboptimal control could increase the wear rate of total knee replacements. Whereas autopsy-retrieved implants from non-PD patients typically show average polyethylene wear of 17 mm³ per year, our simulations suggested that a severely suboptimal control could cause a wear rate as high as of 69 mm³ per year. Assuming the risk of implant failure due to massive wear increase linearly with the wear rate, a severely suboptimal control could increase the risk associated to that failure mode from 0.1% to 0.5%. Based on these results, such increase would not be not sufficient to justify alone the higher incidence rate of revision in patients affected by Parkinson’s Disease, suggesting that other failure modes may be involved.

Bayesian metabolic flux analysis reveals intracellular flux couplings

Motivation

Metabolic flux balance analysis (FBA) is a standard tool in analyzing metabolic reaction rates compatible with measurements, steady-state and the metabolic reaction network stoichiometry. Flux analysis methods commonly place model assumptions on fluxes due to the convenience of formulating the problem as a linear programing model, while many methods do not consider the inherent uncertainty in flux estimates.

Results

We introduce a novel paradigm of Bayesian metabolic flux analysis that models the reactions of the whole genome-scale cellular system in probabilistic terms, and can infer the full flux vector distribution of genome-scale metabolic systems based on exchange and intracellular (e.g. 13C) flux measurements, steady-state assumptions, and objective function assumptions. The Bayesian model couples all fluxes jointly together in a simple truncated multivariate posterior distribution, which reveals informative flux couplings. Our model is a plug-in replacement to conventional metabolic balance methods, such as FBA. Our experiments indicate that we can characterize the genome-scale flux covariances, reveal flux couplings, and determine more intracellular unobserved fluxes in Clostridium acetobutylicum from 13C data than flux variability analysis.

Availability and implementation

The COBRA compatible software is available at github.com/markusheinonen/bamfa.

Supplementary information

Supplementary data are available at Bioinformatics online.

Seeing the wood for the trees: a forest of methods for optimization and omic-network integration in metabolic modelling

Metabolic modelling has entered a mature phase with dozens of methods and software implementations available to the practitioner and the theoretician. It is not easy for a modeller to be able to see the wood (or the forest) for the trees. Driven by this analogy, we here present a 'forest' of principal methods used for constraint-based modelling in systems biology. This provides a tree-based view of methods available to prospective modellers, also available in interactive version at http://modellingmetabolism.net, where it will be kept updated with new methods after the publication of the present manuscript. Our updated classification of existing methods and tools highlights the most promising in the different branches, with the aim to develop a vision of how existing methods could hybridize and become more complex. We then provide the first hands-on tutorial for multi-objective optimization of metabolic models in R. We finally discuss the implementation of multi-view machine learning approaches in poly-omic integration. Throughout this work, we demonstrate the optimization of trade-offs between multiple metabolic objectives, with a focus on omic data integration through machine learning. We anticipate that the combination of a survey, a perspective on multi-view machine learning and a step-by-step R tutorial should be of interest for both the beginner and the advanced user.

Stochastic modelling of muscle recruitment during activity

Muscle forces can be selected from a space of muscle recruitment strategies that produce stable motion and variable muscle and joint forces. However, current optimization methods provide only a single muscle recruitment strategy. We modelled the spectrum of muscle recruitment strategies while walking. The equilibrium equations at the joints, muscle constraints, static optimization solutions and 15-channel electromyography (EMG) recordings for seven walking cycles were taken from earlier studies. The spectrum of muscle forces was calculated using Bayesian statistics and Markov chain Monte Carlo (MCMC) methods, whereas EMG-driven muscle forces were calculated using EMG-driven modelling. We calculated the differences between the spectrum and EMG-driven muscle force for 1-15 input EMGs, and we identified the muscle strategy that best matched the recorded EMG pattern. The best-fit strategy, static optimization solution and EMG-driven force data were compared using correlation analysis. Possible and plausible muscle forces were defined as within physiological boundaries and within EMG boundaries. Possible muscle and joint forces were calculated by constraining the muscle forces between zero and the peak muscle force. Plausible muscle forces were constrained within six selected EMG boundaries. The spectrum to EMG-driven force difference increased from 40 to 108 N for 1-15 EMG inputs. The best-fit muscle strategy better described the EMG-driven pattern (R (2) = 0.94; RMSE = 19 N) than the static optimization solution (R (2) = 0.38; RMSE = 61 N). Possible forces for 27 of 34 muscles varied between zero and the peak muscle force, inducing a peak hip force of 11.3 body-weights. Plausible muscle forces closely matched the selected EMG patterns; no effect of the EMG constraint was observed on the remaining muscle force ranges. The model can be used to study alternative muscle recruitment strategies in both physiological and pathophysiological neuromotor conditions.

Using bioconductor package BiGGR for metabolic flux estimation based on gene expression changes in brain

Predicting the distribution of metabolic fluxes in biochemical networks is of major interest in systems biology. Several databases provide metabolic reconstructions for different organisms. Software to analyze flux distributions exists, among others for the proprietary MATLAB environment. Given the large user community for the R computing environment, a simple implementation of flux analysis in R appears desirable and will facilitate easy interaction with computational tools to handle gene expression data. We extended the R software package BiGGR, an implementation of metabolic flux analysis in R. BiGGR makes use of public metabolic reconstruction databases, and contains the BiGG database and the reconstruction of human metabolism Recon2 as Systems Biology Markup Language (SBML) objects. Models can be assembled by querying the databases for pathways, genes or reactions of interest. Fluxes can then be estimated by maximization or minimization of an objective function using linear inverse modeling algorithms. Furthermore, BiGGR provides functionality to quantify the uncertainty in flux estimates by sampling the constrained multidimensional flux space. As a result, ensembles of possible flux configurations are constructed that agree with measured data within precision limits. BiGGR also features automatic visualization of selected parts of metabolic networks using hypergraphs, with hyperedge widths proportional to estimated flux values. BiGGR supports import and export of models encoded in SBML and is therefore interoperable with different modeling and analysis tools. As an application example, we calculated the flux distribution in healthy human brain using a model of central carbon metabolism. We introduce a new algorithm termed Least-squares with equalities and inequalities Flux Balance Analysis (Lsei-FBA) to predict flux changes from gene expression changes, for instance during disease. Our estimates of brain metabolic flux pattern with Lsei-FBA for Alzheimer’s disease agree with independent measurements of cerebral metabolism in patients. This second version of BiGGR is available from Bioconductor.

Development and application of efficient pathway enumeration algorithms for metabolic engineering applications

Metabolic Engineering (ME) aims to design microbial cell factories towards the production of valuable compounds. In this endeavor, one important task relates to the search for the most suitable heterologous pathway(s) to add to the selected host. Different algorithms have been developed in the past towards this goal, following distinct approaches spanning constraint-based modeling, graph-based methods and knowledge-based systems based on chemical rules. While some of these methods search for pathways optimizing specific objective functions, here the focus will be on methods that address the enumeration of pathways that are able to convert a set of source compounds into desired targets and their posterior evaluation according to different criteria. Two pathway enumeration algorithms based on (hyper)graph-based representations are selected as the most promising ones and are analyzed in more detail: the Solution Structure Generation and the Find Path algorithms. Their capabilities and limitations are evaluated when designing novel heterologous pathways, by applying these methods on three case studies of synthetic ME related to the production of non-native compounds in E. coli and S. cerevisiae: 1-butanol, curcumin and vanillin. Some targeted improvements are implemented, extending both methods to address limitations identified that impair their scalability, improving their ability to extract potential pathways over large-scale databases. In all case-studies, the algorithms were able to find already described pathways for the production of the target compounds, but also alternative pathways that can represent novel ME solutions after further evaluation.

Quantitative in silico Analysis of Neurotransmitter Pathways Under Steady State Conditions

The modeling of glutamate/GABA-glutamine cycling in the brain tissue involving astrocytes, glutamatergic and GABAergic neurons leads to a complex compartmentalized metabolic network that comprises neurotransmitter synthesis, shuttling, and degradation. Without advanced computational tools, it is difficult to quantitatively track possible scenarios and identify viable ones. In this article, we follow a sampling-based computational paradigm to analyze the biochemical network in a multi-compartment system modeling astrocytes, glutamatergic, and GABAergic neurons, and address some questions about the details of transmitter cycling, with particular emphasis on the ammonia shuttling between astrocytes and neurons, and the synthesis of transmitter GABA. More specifically, we consider the joint action of the alanine-lactate shuttle, the branched chain amino acid shuttle, and the glutamine-glutamate cycle, as well as the role of glutamate dehydrogenase (GDH) activity. When imposing a minimal amount of bound constraints on reaction and transport fluxes, a preferred stoichiometric steady state equilibrium requires an unrealistically high reductive GDH activity in neurons, indicating the need for additional bound constants which were included in subsequent computer simulations. The statistical flux balance analysis also suggests a stoichiometrically viable role for leucine transport as an alternative to glutamine for replenishing the glutamate pool in neurons.

Genome-scale model management and comparison

Spurred by recent innovations in genome sequencing, the reconstruction of genome-scale models has increased in recent years. Genome-scale models are now available for a wide range of organisms, and models have been successfully applied to a number of research topics including metabolic engineering, genome annotation, biofuel production, and interpretation of omics data sets. The challenge is how to manage the large amount of data in genome-scale models and perform comparative analysis to gain new biological insights. In this chapter, important standards for genome-scale modeling are outlined. Furthermore, management strategies as well as existing repository and construction tools are discussed. As the comparison of models is an important aspect during the development and analysis stages, available methods are presented and existing software solutions are reviewed.

The metabolism of neurons and astrocytes through mathematical models

Mathematical modeling of the energy metabolism of brain cells plays a central role in understanding data collected with different imaging modalities, and in making predictions based on them. During the last decade, several sophisticated brain metabolism models have appeared. Unfortunately, the picture of the metabolic details that emerges from them is far from coherent: while each model has its justification and is in agreement with some experimental data, some of the predictions of different models can diverge from each other significantly. In this article, we review some of the recent published models, emphasizing similarities and differences between them to understand where the differences in predictions stem from. In that context we present a probabilistic approach, which rather than assigning fixed values to the model parameters, regard them as random variables whose distributions are inferred on in the light of stoichiometric information and different observations. The probabilistic approach reveals how much intrinsic variability a metabolic system may contain, which in turn may be a valid explanation of the different findings.

Ménage à trois: the role of neurotransmitters in the energy metabolism of astrocytes, glutamatergic, and GABAergic neurons

This work is a computational study based on a new detailed metabolic network model comprising well-mixed compartments representing separate cytosol and mitochondria of astrocytes, glutamatergic and gamma aminobutyric acid (GABA)ergic neurons, communicating through an extracellular space compartment and fed by arterial blood flow. Our steady-state analysis assumes statistical mass balance of both carbons and amino groups. The study is based on Bayesian flux balance analysis, which uses Markov chain Monte Carlo sampling techniques and provides a quantitative description of steady states when the two exchangers aspartate-glutamate carrier (AGC1) and oxoglutarate carrier (OGC) in the malate-aspartate shuttle in astrocyte are not in equilibrium, as recent studies suggest. It also highlights the importance of anaplerotic reactions, pyruvate carboxylase in astrocyte and malic enzyme in neurons, for neurotransmitter synthesis and recycling. The model is unbiased with respect to the glucose partitioning between cell types, and shows that determining the partitioning cannot be done by stoichiometric constraints alone. Furthermore, the intercellular lactate trafficking is found to depend directly on glucose partitioning, suggesting that a steady state may support different scenarios. At inhibitory steady state, characterized by high rate of GABA release, there is elevated oxidative activity in astrocyte, not in response to specific energetic needs.

A tentative taxonomy for predictive models in relation to their falsifiability

The growing importance of predictive models in biomedical research raises some concerns on the correct methodological approach to the falsification of such models, as they are developed in interdisciplinary research contexts between physics, biology and medicine. In each of these research sectors, there are established methods to develop cause-effect explanations for observed phenomena, which can be used to predict: epidemiological models, biochemical models, biophysical models, Bayesian models, neural networks, etc. Each research sector has accepted processes to verify how correct these models are (falsification). But interdisciplinary research imposes a broader perspective, which encompasses all possible models in a general methodological framework of falsification. The present paper proposes a general definition of 'scientific model' that makes it possible to categorize predictive models into broad categories. For each of these categories, generic falsification strategies are proposed, except for the so-called 'abductive' models. For this category, which includes artificial neural networks, Bayesian models and integrative models, the definition of a generic falsification strategy requires further investigation by researchers and philosophers of science.

Dynamic activation model for a glutamatergic neurovascular unit

This article considers a dynamic spatially lumped model for brain energy metabolism and proposes to use the results of a Markov chain Monte Carlo (MCMC) based flux balance analysis to estimate the kinetic model parameters. By treating steady state reaction fluxes and transport rates as random variables we are able to propagate the uncertainty in the steady state configurations to the predictions of the dynamic model, whose responses are no longer individual but ensembles of time courses. The kinetic model consists of five compartments and is governed by kinetic mass balance equations with Michaelis-Menten type expressions for reaction rates and transports between the compartments. The neuronal activation is implemented in terms of the effect of neuronal activity on parameters controlling the blood flow and neurotransmitter transport, and a feedback mechanism coupling the glutamate concentration in the synaptic cleft and the ATP hydrolysis, thus accounting for the energetic cost of the membrane potential restoration in the postsynaptic neurons. The changes in capillary volume follow the balloon model developed for BOLD MRI. The model follows the time course of the saturation levels of the blood hemoglobin, which link metabolism and BOLD FMRI signal. Analysis of the model predictions suggest that stoichiometry alone is not enough to determine glucose partitioning between neuron and astrocyte. Lactate exchange between neuron and astrocyte is supported by the model predictions, but the uncertainty on the direction and rate is rather elevated. By and large, the model suggests that astrocyte produces and effluxes lactate, while neuron may switch from using to producing lactate. The level of ATP hydrolysis in astrocyte is substantially higher than strictly required for neurotransmitter cycling, in agreement with the literature.

Energetics of inhibition: insights with a computational model of the human GABAergic neuron-astrocyte cellular complex

We investigate metabolic interactions between astrocytes and GABAergic neurons at steady states corresponding to different activity levels using a six-compartment model and a new methodology based on Bayesian statistics. Many questions about the energetics of inhibition are still waiting for definite answers, including the role of glutamine and lactate effluxed by astrocytes as precursors for γ-aminobutyric acid (GABA), and whether metabolic coupling applies to the inhibitory neurotransmitter GABA. Our identification and quantification of metabolic pathways describing the interaction between GABAergic neurons and astrocytes in connection with the release of GABA makes a contribution to this important problem. Lactate released by astrocytes and its neuronal uptake is found to be coupled with neuronal activity, unlike glucose consumption, suggesting that in astrocytes, the metabolism of GABA does not require increased glycolytic activity. Negligible glutamine trafficking between the two cell types at steady state questions glutamine as a precursor of GABA, not excluding glutamine cycling as a transient dynamic phenomenon, or a prominent role of GABA reuptake. Redox balance is proposed as an explanation for elevated oxidative phosphorylation and adenosine triphosphate hydrolysis in astrocytes, decoupled from energy requirements.

Production of biofuels and biochemicals: in need of an ORACLE

The engineering of cells for the production of fuels and chemicals involves simultaneous optimization of multiple objectives, such as specific productivity, extended substrate range and improved tolerance - all under a great degree of uncertainty. The achievement of these objectives under physiological and process constraints will be impossible without the use of mathematical modeling. However, the limited information and the uncertainty in the available information require new methods for modeling and simulation that will characterize the uncertainty and will quantify, in a statistical sense, the expectations of success of alternative metabolic engineering strategies. We discuss these considerations toward developing a framework for the Optimization and Risk Analysis of Complex Living Entities (ORACLE) - a computational method that integrates available information into a mathematical structure to calculate control coefficients.