Coracle-a machine learning framework to identify bacteria associated with continuous variables.

SUMMARY: We present Coracle, an artificial intelligence (AI) framework that can identify associations between bacterial communities and continuous variables. Coracle uses an ensemble approach of prominent feature selection methods and machine learning (ML) models to identify features, i.e. bacteria, associated with a continuous variable, e.g. host thermal tolerance. The results are aggregated into a score that incorporates the performances of the different ML models and the respective feature importance, while also considering the robustness of feature selection. Additionally, regression coefficients provide first insights into the direction of the association. We show the utility of Coracle by analyzing associations between bacterial composition data (i.e. 16S rRNA Amplicon Sequence Variants, ASVs) and coral thermal tolerance (i.e. standardized short-term heat stress-derived diagnostics). This analysis identified high-scoring bacterial taxa that were previously found associated with coral thermal tolerance. Coracle scales with feature number and performs well with hundreds to thousands of features, corresponding to the typical size of current datasets. Coracle performs best if run at a higher taxonomic level first (e.g. order or family) to identify groups of interest that can subsequently be run at the ASV level. AVAILABILITY AND IMPLEMENTATION: Coracle can be accessed via a dedicated web server that allows free and simple access: http://www.micportal.org/coracle/index. The underlying code is open-source and available via GitHub https://github.com/SebastianStaab/coracle.git.

PathwayNexus: a tool for interactive metabolic data analysis.

MOTIVATION: High-throughput omics methods increasingly result in large datasets including metabolomics data, which are often difficult to analyse. RESULTS: To help researchers to handle and analyse those datasets by mapping and investigating metabolomics data of multiple sampling conditions (e.g. different time points or treatments) in the context of pathways, PathwayNexus has been developed, which presents the mapping results in a matrix format, allowing users to easily observe the relations between the compounds and the pathways. It also offers functionalities like ranking, sorting, clustering, pathway views, and further analytical tools. Its primary objective is to condense large sets of pathways into smaller, more relevant subsets that align with the specific interests of the user. AVAILABILITY AND IMPLEMENTATION: The methodology presented here is implemented in PathwayNexus, an open-source add-on for Vanted available at www.cls.uni-konstanz.de/software/pathway-nexus. CONTACT: falk.schreiber@unikonstanz.de. SUPPLEMENTARY INFORMATION: Website: www.cls.uni-konstanz.de/software/pathway-nexus.

Addressing barriers in comprehensiveness, accessibility, reusability, interoperability and reproducibility of computational models in systems biology.

Computational models are often employed in systems biology to study the dynamic behaviours of complex systems. With the rise in the number of computational models, finding ways to improve the reusability of these models and their ability to reproduce virtual experiments becomes critical. Correct and effective model annotation in community-supported and standardised formats is necessary for this improvement. Here, we present recent efforts toward a common framework for annotated, accessible, reproducible and interoperable computational models in biology, and discuss key challenges of the field.

APL@voro-interactive visualization and analysis of cell membrane simulations.

SUMMARY: Molecular dynamics (MD) simulations of cell membranes allow for a better understanding of complex processes such as changing membrane dynamics, lipid rafts and the incorporation/passing of macromolecules into/through membranes. To explore and understand cell membrane compositions, dynamics and processes, visual analytics can help to interpret MD simulation data. APL@Voro is a software for the interactive visualization and analysis of cell membrane simulations. Here, we present the new APL@Voro, which has been continuously developed since its initial release in 2013. We discuss newly implemented algorithms, methodologies and features, such as the interactive comparison of related simulations and methods to assign lipids to either the upper or lower leaflet. AVAILABILITY AND IMPLEMENTATION: The current open-source version of APL@Voro can be downloaded from http://aplvoro.com.

Visual exploration of large metabolic models.

MOTIVATION: Large metabolic models, including genome-scale metabolic models, are nowadays common in systems biology, biotechnology and pharmacology. They typically contain thousands of metabolites and reactions and therefore methods for their automatic visualization and interactive exploration can facilitate a better understanding of these models. RESULTS: We developed a novel method for the visual exploration of large metabolic models and implemented it in LMME (Large Metabolic Model Explorer), an add-on for the biological network analysis tool VANTED. The underlying idea of our method is to analyze a large model as follows. Starting from a decomposition into several subsystems, relationships between these subsystems are identified and an overview is computed and visualized. From this overview, detailed subviews may be constructed and visualized in order to explore subsystems and relationships in greater detail. Decompositions may either be predefined or computed, using built-in or self-implemented methods. Realized as add-on for VANTED, LMME is embedded in a domain-specific environment, allowing for further related analysis at any stage during the exploration. We describe the method, provide a use case and discuss the strengths and weaknesses of different decomposition methods. AVAILABILITY AND IMPLEMENTATION: The methods and algorithms presented here are implemented in LMME, an open-source add-on for VANTED. LMME can be downloaded from www.cls.uni-konstanz.de/software/lmme and VANTED can be downloaded from www.vanted.org. The source code of LMME is available from GitHub, at https://github.com/LSI-UniKonstanz/lmme.

COVID19 Disease Map, a computational knowledge repository of virus-host interaction mechanisms.

We need to effectively combine the knowledge from surging literature with complex datasets to propose mechanistic models of SARS-CoV-2 infection, improving data interpretation and predicting key targets of intervention. Here, we describe a large-scale community effort to build an open access, interoperable and computable repository of COVID-19 molecular mechanisms. The COVID-19 Disease Map (C19DMap) is a graphical, interactive representation of disease-relevant molecular mechanisms linking many knowledge sources. Notably, it is a computational resource for graph-based analyses and disease modelling. To this end, we established a framework of tools, platforms and guidelines necessary for a multifaceted community of biocurators, domain experts, bioinformaticians and computational biologists. The diagrams of the C19DMap, curated from the literature, are integrated with relevant interaction and text mining databases. We demonstrate the application of network analysis and modelling approaches by concrete examples to highlight new testable hypotheses. This framework helps to find signatures of SARS-CoV-2 predisposition, treatment response or prioritisation of drug candidates. Such an approach may help deal with new waves of COVID-19 or similar pandemics in the long-term perspective.

Simulations of octapeptin-outer membrane interactions reveal conformational flexibility is linked to antimicrobial potency.

The octapeptins are lipopeptide antibiotics that are structurally similar to polymyxins yet retain activity against polymyxin-resistant Gram-negative pathogens, suggesting they might be used to treat recalcitrant infections. However, the basis of their unique activity is unclear because of the difficulty in generating high-resolution experimental data of the interaction of antimicrobial peptides with lipid membranes. To elucidate these structure-activity relationships, we employed all-atom molecular dynamics simulations with umbrella sampling to investigate the conformational and energetic landscape of octapeptins interacting with bacterial outer membrane (OM). Specifically, we examined the interaction of octapeptin C4 and FADDI-115, lacking a single hydroxyl group compared with octapeptin C4, with the lipid A-phosphoethanolamine modified OM of Acinetobacter baumannii Octapeptin C4 and FADDI-115 both penetrated into the OM hydrophobic center but experienced different conformational transitions from an unfolded to a folded state that was highly dependent on the structural flexibility of their respective N-terminal fatty acyl groups. The additional hydroxyl group present in the fatty acyl group of octapeptin C4 resulted in the molecule becoming trapped in a semifolded state, leading to a higher free energy barrier for OM penetration. The free energy barrier for the translocation through the OM hydrophobic layer was ∼72 kcal/mol for octapeptin C4 and 62 kcal/mol for FADDI-115. Our results help to explain the lower antimicrobial activity previously observed for octapeptin C4 compared with FADDI-115 and more broadly improve our understanding of the structure-function relationships of octapeptins. These findings may facilitate the discovery of next-generation octapeptins against polymyxin-resistant Gram-negative 'superbugs.'

Molecular dynamics simulations informed by membrane lipidomics reveal the structure-interaction relationship of polymyxins with the lipid A-based outer membrane of Acinetobacter baumannii.

BACKGROUND: MDR bacteria represent an urgent threat to human health globally. Polymyxins are a last-line therapy against life-threatening Gram-negative 'superbugs', including Acinetobacter baumannii. Polymyxins exert antimicrobial activity primarily via permeabilizing the bacterial outer membrane (OM); however, the mechanism of interaction between polymyxins and the OM remains unclear at the atomic level. METHODS: We constructed a lipid A-based OM model of A. baumannii using quantitative membrane lipidomics data and employed all-atom molecular dynamics simulations with umbrella sampling techniques to elucidate the structure-interaction relationship and thermodynamics governing the penetration of polymyxins [B1 and E1 (i.e. colistin A) representing the two clinically used polymyxins] into the OM. RESULTS: Polymyxin B1 and colistin A bound to the A. baumannii OM by the initial electrostatic interactions between the Dab residues of polymyxins and the phosphates of lipid A, competitively displacing the cations from the headgroup region of the OM. Both polymyxin B1 and colistin A formed a unique folded conformation upon approaching the hydrophobic centre of the OM, consistent with previous experimental observations. Polymyxin penetration induced reorientation of the headgroups of the OM lipids near the penetration site and caused local membrane disorganization, thereby significantly increasing membrane permeability and promoting the subsequent penetration of polymyxin molecules into the OM and periplasmic space. CONCLUSIONS: The thermodynamics governing the penetration of polymyxins through the outer leaflet of the A. baumannii OM were examined and novel structure-interaction relationship information was obtained at the atomic and membrane level. Our findings will facilitate the discovery of novel polymyxins against MDR Gram-negative pathogens.

Stage-specific metabolic features of differentiating neurons: Implications for toxicant sensitivity.

Developmental neurotoxicity (DNT) may be induced when chemicals disturb a key neurodevelopmental process, and many tests focus on this type of toxicity. Alternatively, DNT may occur when chemicals are cytotoxic only during a specific neurodevelopmental stage. The toxicant sensitivity is affected by the expression of toxicant targets and by resilience factors. Although cellular metabolism plays an important role, little is known how it changes during human neurogenesis, and how potential alterations affect toxicant sensitivity of mature vs. immature neurons. We used immature (d0) and mature (d6) LUHMES cells (dopaminergic human neurons) to provide initial answers to these questions. Transcriptome profiling and characterization of energy metabolism suggested a switch from predominantly glycolytic energy generation to a more pronounced contribution of the tricarboxylic acid cycle (TCA) during neuronal maturation. Therefore, we used pulsed stable isotope-resolved metabolomics (pSIRM) to determine intracellular metabolite pool sizes (concentrations), and isotopically non-stationary 13C-metabolic flux analysis (INST 13C-MFA) to calculate metabolic fluxes. We found that d0 cells mainly use glutamine to fuel the TCA. Furthermore, they rely on extracellular pyruvate to allow continuous growth. This metabolic situation does not allow for mitochondrial or glycolytic spare capacity, i.e. the ability to adapt energy generation to altered needs. Accordingly, neuronal precursor cells displayed a higher sensitivity to several mitochondrial toxicants than mature neurons differentiated from them. In summary, this study shows that precursor cells lose their glutamine dependency during differentiation while they gain flexibility of energy generation and thereby increase their resistance to low concentrations of mitochondrial toxicants.

Plant metabolic modeling: achieving new insight into metabolism and metabolic engineering.

Models are used to represent aspects of the real world for specific purposes, and mathematical models have opened up new approaches in studying the behavior and complexity of biological systems. However, modeling is often time-consuming and requires significant computational resources for data development, data analysis, and simulation. Computational modeling has been successfully applied as an aid for metabolic engineering in microorganisms. But such model-based approaches have only recently been extended to plant metabolic engineering, mainly due to greater pathway complexity in plants and their highly compartmentalized cellular structure. Recent progress in plant systems biology and bioinformatics has begun to disentangle this complexity and facilitate the creation of efficient plant metabolic models. This review highlights several aspects of plant metabolic modeling in the context of understanding, predicting and modifying complex plant metabolism. We discuss opportunities for engineering photosynthetic carbon metabolism, sucrose synthesis, and the tricarboxylic acid cycle in leaves and oil synthesis in seeds and the application of metabolic modeling to the study of plant acclimation to the environment. The aim of the review is to offer a current perspective for plant biologists without requiring specialized knowledge of bioinformatics or systems biology.

The Metabolic Signature of Biomass Formation in Barley.

The network analysis of genome-wide transcriptome responses, metabolic signatures and enzymes' relationship to biomass formation has been studied in a diverse panel of 12 barley accessions during vegetative and reproductive stages. The primary metabolites and enzymes involved in central metabolism that determine the accumulation of shoot biomass at the vegetative stage of barley development are primarily being linked to sucrose accumulation and sucrose synthase activity. Interestingly, the metabolic and enzyme links which are strongly associated with biomass accumulation during reproductive stages are related to starch accumulation and tricarboxylic acid (TCA) cycle intermediates citrate, malate, trans-aconitate and isocitrate. Additional significant associations were also found for UDP glucose, ATP and the amino acids isoleucine, valine, glutamate and histidine during the reproductive stage. A network analysis resulted in a combined identification of metabolite and enzyme signatures indicative for grain weight accumulation that was correlated with the activity of ADP-glucose pyrophosphorylase (AGPase), a rate-limiting enzyme involved in starch biosynthesis, and with that of alanine amino transferase involved in the synthesis of storage proteins. We propose that the mechanism related to vegetative and reproductive biomass formation vs. seed biomass formation is being linked to distinct fluxes regulating sucrose, starch, sugars and amino acids as central resources. These distinct biomarkers can be used to engineer biomass production and grain weight in barley.

Metabolic Architecture of the Cereal Grain and Its Relevance to Maximize Carbon Use Efficiency.

Here, we have characterized the spatial heterogeneity of the cereal grain's metabolism and demonstrated how, by integrating a distinct set of metabolic strategies, the grain has evolved to become an almost perfect entity for carbon storage. In vivo imaging revealed light-induced cycles in assimilate supply toward the ear/grain of barley (Hordeum vulgare) and wheat (Triticum aestivum). In silico modeling predicted that, in the two grain storage organs (the endosperm and embryo), the light-induced shift in solute influx does cause adjustment in metabolic flux without changes in pathway utilization patterns. The enveloping, leaf-like pericarp, in contrast, shows major shifts in flux distribution (starch metabolism, photosynthesis, remobilization, and tricarboxylic acid cycle activity) allow to refix 79% of the CO2 released by the endosperm and embryo, allowing the grain to achieve an extraordinary high carbon conversion efficiency of 95%. Shading experiments demonstrated that ears are autonomously able to raise the influx of solutes in response to light, but with little effect on the steady-state levels of metabolites or transcripts or on the pattern of sugar distribution within the grain. The finding suggests the presence of a mechanism(s) able to ensure metabolic homeostasis in the face of short-term environmental fluctuation. The proposed multicomponent modeling approach is informative for predicting the metabolic effects of either an altered level of incident light or a momentary change in the supply of sucrose. It is therefore of potential value for assessing the impact of either breeding and/or biotechnological interventions aimed at increasing grain yield.

Quantitative Multilevel Analysis of Central Metabolism in Developing Oilseeds of Oilseed Rape during in Vitro Culture.

Seeds provide the basis for many food, feed, and fuel products. Continued increases in seed yield, composition, and quality require an improved understanding of how the developing seed converts carbon and nitrogen supplies into storage. Current knowledge of this process is often based on the premise that transcriptional regulation directly translates via enzyme concentration into flux. In an attempt to highlight metabolic control, we explore genotypic differences in carbon partitioning for in vitro cultured developing embryos of oilseed rape (Brassica napus). We determined biomass composition as well as 79 net fluxes, the levels of 77 metabolites, and 26 enzyme activities with specific focus on central metabolism in nine selected germplasm accessions. Overall, we observed a tradeoff between the biomass component fractions of lipid and starch. With increasing lipid content over the spectrum of genotypes, plastidic fatty acid synthesis and glycolytic flux increased concomitantly, while glycolytic intermediates decreased. The lipid/starch tradeoff was not reflected at the proteome level, pointing to the significance of (posttranslational) metabolic control. Enzyme activity/flux and metabolite/flux correlations suggest that plastidic pyruvate kinase exerts flux control and that the lipid/starch tradeoff is most likely mediated by allosteric feedback regulation of phosphofructokinase and ADP-glucose pyrophosphorylase. Quantitative data were also used to calculate in vivo mass action ratios, reaction equilibria, and metabolite turnover times. Compounds like cyclic 3',5'-AMP and sucrose-6-phosphate were identified to potentially be involved in so far unknown mechanisms of metabolic control. This study provides a rich source of quantitative data for those studying central metabolism.

Alterations in cancer cell metabolism: the Warburg effect and metabolic adaptation.

The Warburg effect means higher glucose uptake of cancer cells compared to normal tissues, whereas a smaller fraction of this glucose is employed for oxidative phosphorylation. With the advent of high throughput technologies and computational systems biology, cancer cell metabolism has been reinvestigated over the last decades toward identifying various events underlying "how" and "why" a cancer cell employs aerobic glycolysis. Significant progress has been shaped to revise the Warburg effect. In this study, we have integrated the gene expression of 13 different cancer cells with the genome-scale metabolic network of human (Recon1) based on the E-Flux method, and analyzed them based on constraint-based modeling. Results show that regardless of significant up- and down-regulated metabolic genes, the distribution of metabolic changes is similar in different cancer types. These findings support the theory that the Warburg effect is a consequence of metabolic adaptation in cancer cells.

A unified framework for estimating parameters of kinetic biological models.

BACKGROUND: Utilizing kinetic models of biological systems commonly require computational approaches to estimate parameters, posing a variety of challenges due to their highly non-linear and dynamic nature, which is further complicated by the issue of non-identifiability. We propose a novel parameter estimation framework by combining approaches for solving identifiability with a recently introduced filtering technique that can uniquely estimate parameters where conventional methods fail. This framework first conducts a thorough analysis to identify and classify the non-identifiable parameters and provides a guideline for solving them. If no feasible solution can be found, the framework instead initializes the filtering technique with informed prior to yield a unique solution. RESULTS: This framework has been applied to uniquely estimate parameter values for the sucrose accumulation model in sugarcane culm tissue and a gene regulatory network. In the first experiment the results show the progression of improvement in reliable and unique parameter estimation through the use of each tool to reduce and remove non-identifiability. The latter experiment illustrates the common situation where no further measurement data is available to solve the non-identifiability. These results show the successful application of the informed prior as well as the ease with which parallel data sources may be utilized without increasing the model complexity. CONCLUSION: The proposed unified framework is distinct from other approaches by providing a robust and complete solution which yields reliable and unique parameter estimation even in the face of non-identifiability.

An improved constraint filtering technique for inferring hidden states and parameters of a biological model.

MOTIVATION: In systems biology, kinetic models represent the biological system using a set of ordinary differential equations (ODEs). The correct values of the parameters within these ODEs are critical for a reliable study of the dynamic behaviour of such systems. Typically, it is only possible to experimentally measure a fraction of these parameter values. The rest must be indirectly determined from measurements of other quantities. In this article, we propose a novel statistical inference technique to computationally estimate these unknown parameter values. By characterizing the ODEs with non-linear state-space equations, this inference technique models the unknown parameters as hidden states, which can then be estimated from noisy measurement data. RESULTS: Here we extended the square-root unscented Kalman filter SR-UKF proposed by Merwe and Wan to include constraints with the state estimation process. We developed the constrained square-root unscented Kalman filter (CSUKF) to estimate parameters of non-linear state-space models. This probabilistic inference technique was successfully used to estimate parameters of a glycolysis model in yeast and a gene regulatory network. We showed that our method is numerically stable and can reliably estimate parameters within a biologically meaningful parameter space from noisy observations. When compared with the two common non-linear extensions of Kalman filter in addition to four widely used global optimization algorithms, CSUKF is shown to be both accurate and computationally efficient. With CSUKF, statistical analysis is straightforward, as it directly provides the uncertainty on the estimation result. AVAILABILITY AND IMPLEMENTATION: Matlab code available upon request from the author. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Multiscale metabolic modeling: dynamic flux balance analysis on a whole-plant scale.

Plant metabolism is characterized by a unique complexity on the cellular, tissue, and organ levels. On a whole-plant scale, changing source and sink relations accompanying plant development add another level of complexity to metabolism. With the aim of achieving a spatiotemporal resolution of source-sink interactions in crop plant metabolism, a multiscale metabolic modeling (MMM) approach was applied that integrates static organ-specific models with a whole-plant dynamic model. Allowing for a dynamic flux balance analysis on a whole-plant scale, the MMM approach was used to decipher the metabolic behavior of source and sink organs during the generative phase of the barley (Hordeum vulgare) plant. It reveals a sink-to-source shift of the barley stem caused by the senescence-related decrease in leaf source capacity, which is not sufficient to meet the nutrient requirements of sink organs such as the growing seed. The MMM platform represents a novel approach for the in silico analysis of metabolism on a whole-plant level, allowing for a systemic, spatiotemporally resolved understanding of metabolic processes involved in carbon partitioning, thus providing a novel tool for studying yield stability and crop improvement.

Combined noninvasive imaging and modeling approaches reveal metabolic compartmentation in the barley endosperm.

The starchy endosperm of cereals is a priori taken as a metabolically uniform tissue. By applying a noninvasive assay based on (13)C/(1)H-magnetic resonance imaging (MRI) to barley (Hordeum vulgare) grains, we uncovered metabolic compartmentation in the endosperm. (13)C-Suc feeding during grain filling showed that the primary site of Ala synthesis was the central region of the endosperm, the part of the caryopsis experiencing the highest level of hypoxia. Region-specific metabolism in the endosperm was characterized by flux balance analysis (FBA) and metabolite profiling. FBA predicts that in the central region of the endosperm, the tricarboxylic acid cycle shifts to a noncyclic mode, accompanied by elevated glycolytic flux and the accumulation of Ala. The metabolic compartmentation within the endosperm is advantageous for the grain's carbon and energy economy, with a prominent role being played by Ala aminotransferase. An investigation of caryopses with a genetically perturbed tissue pattern demonstrated that Ala accumulation is a consequence of oxygen status, rather than being either tissue specific or dependent on the supply of Suc. Hence the (13)C-Ala gradient can be used as an in vivo marker for hypoxia. The combination of MRI and metabolic modeling offers opportunities for the noninvasive analysis of metabolic compartmentation in plants.