PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration.

As terabytes of multi-omics data are being generated, there is an ever-increasing need for methods facilitating the integration and interpretation of such data. Current multi-omics integration methods typically output lists, clusters, or subnetworks of molecules related to an outcome. Even with expert domain knowledge, discerning the biological processes involved is a time-consuming activity. Here we propose PathIntegrate, a method for integrating multi-omics datasets based on pathways, designed to exploit knowledge of biological systems and thus provide interpretable models for such studies. PathIntegrate employs single-sample pathway analysis to transform multi-omics datasets from the molecular to the pathway-level, and applies a predictive single-view or multi-view model to integrate the data. Model outputs include multi-omics pathways ranked by their contribution to the outcome prediction, the contribution of each omics layer, and the importance of each molecule in a pathway. Using semi-synthetic data we demonstrate the benefit of grouping molecules into pathways to detect signals in low signal-to-noise scenarios, as well as the ability of PathIntegrate to precisely identify important pathways at low effect sizes. Finally, using COPD and COVID-19 data we showcase how PathIntegrate enables convenient integration and interpretation of complex high-dimensional multi-omics datasets. PathIntegrate is available as an open-source Python package.

Genome scale metabolic network modelling for metabolic profile predictions.

Metabolic profiling (metabolomics) aims at measuring small molecules (metabolites) in complex samples like blood or urine for human health studies. While biomarker-based assessment often relies on a single molecule, metabolic profiling combines several metabolites to create a more complex and more specific fingerprint of the disease. However, in contrast to genomics, there is no unique metabolomics setup able to measure the entire metabolome. This challenge leads to tedious and resource consuming preliminary studies to be able to design the right metabolomics experiment. In that context, computer assisted metabolic profiling can be of strong added value to design metabolomics studies more quickly and efficiently. We propose a constraint-based modelling approach which predicts in silico profiles of metabolites that are more likely to be differentially abundant under a given metabolic perturbation (e.g. due to a genetic disease), using flux simulation. In genome-scale metabolic networks, the fluxes of exchange reactions, also known as the flow of metabolites through their external transport reactions, can be simulated and compared between control and disease conditions in order to calculate changes in metabolite import and export. These import/export flux differences would be expected to induce changes in circulating biofluid levels of those metabolites, which can then be interpreted as potential biomarkers or metabolites of interest. In this study, we present SAMBA (SAMpling Biomarker Analysis), an approach which simulates fluxes in exchange reactions following a metabolic perturbation using random sampling, compares the simulated flux distributions between the baseline and modulated conditions, and ranks predicted differentially exchanged metabolites as potential biomarkers for the perturbation. We show that there is a good fit between simulated metabolic exchange profiles and experimental differential metabolites detected in plasma, such as patient data from the disease database OMIM, and metabolic trait-SNP associations found in mGWAS studies. These biomarker recommendations can provide insight into the underlying mechanism or metabolic pathway perturbation lying behind observed metabolite differential abundances, and suggest new metabolites as potential avenues for further experimental analyses.

PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration.

As terabytes of multi-omics data are being generated, there is an ever-increasing need for methods facilitating the integration and interpretation of such data. Current multi-omics integration methods typically output lists, clusters, or subnetworks of molecules related to an outcome. Even with expert domain knowledge, discerning the biological processes involved is a time-consuming activity. Here we propose PathIntegrate, a method for integrating multi-omics datasets based on pathways, designed to exploit knowledge of biological systems and thus provide interpretable models for such studies. PathIntegrate employs single-sample pathway analysis to transform multi-omics datasets from the molecular to the pathway-level, and applies a predictive single-view or multi-view model to integrate the data. Model outputs include multi-omics pathways ranked by their contribution to the outcome prediction, the contribution of each omics layer, and the importance of each molecule in a pathway. Using semi-synthetic data we demonstrate the benefit of grouping molecules into pathways to detect signals in low signal-to-noise scenarios, as well as the ability of PathIntegrate to precisely identify important pathways at low effect sizes. Finally, using COPD and COVID-19 data we showcase how PathIntegrate enables convenient integration and interpretation of complex high-dimensional multi-omics datasets. The PathIntegrate Python package is available at https://github.com/cwieder/PathIntegrate.

APEX: an Annotation Propagation Workflow through Multiple Experimental Networks to Improve the Annotation of New Metabolite Classes in Caenorhabditis elegans.

Spectral similarity networks, also known as molecular networks, are crucial in non-targeted metabolomics to aid identification of unknowns aiming to establish a potential structural relation between different metabolite features. However, too extensive differences in compound structures can lead to separate clusters, complicating annotation. To address this challenge, we developed an automated Annotation Propagation through multiple EXperimental Networks (APEX) workflow, which integrates spectral similarity networks with mass difference networks and homologous series. The incorporation of multiple network tools improved annotation quality, as evidenced by high matching rates of the molecular formula derived by SIRIUS. The selection of manual annotations as the Seed Nodes Set (SNS) significantly influenced APEX annotations, with a higher number of seed nodes enhancing the annotation process. We applied APEX to different Caenorhabditis elegans metabolomics data sets as a proof-of-principle for the effective and comprehensive annotation of glycerophospho N-acyl ethanolamides (GPNAEs) and their glyco-variants. Furthermore, we demonstrated the workflow's applicability to two other, well-described metabolite classes in C. elegans, specifically ascarosides and modular glycosides (MOGLs), using an additional publicly available data set. In summary, the APEX workflow presents a powerful approach for metabolite annotation and identification by leveraging multiple experimental networks. By refining the SNS selection and integrating diverse networks, APEX holds promise for comprehensive annotation in metabolomics research, enabling a deeper understanding of the metabolome.

Networks and Graphs Discovery in Metabolomics Data Analysis and Interpretation.

Both targeted and untargeted mass spectrometry-based metabolomics approaches are used to understand the metabolic processes taking place in various organisms, from prokaryotes, plants, fungi to animals and humans. Untargeted approaches allow to detect as many metabolites as possible at once, identify unexpected metabolic changes, and characterize novel metabolites in biological samples. However, the identification of metabolites and the biological interpretation of such large and complex datasets remain challenging. One approach to address these challenges is considering that metabolites are connected through informative relationships. Such relationships can be formalized as networks, where the nodes correspond to the metabolites or features (when there is no or only partial identification), and edges connect nodes if the corresponding metabolites are related. Several networks can be built from a single dataset (or a list of metabolites), where each network represents different relationships, such as statistical (correlated metabolites), biochemical (known or putative substrates and products of reactions), or chemical (structural similarities, ontological relations). Once these networks are built, they can subsequently be mined using algorithms from network (or graph) theory to gain insights into metabolism. For instance, we can connect metabolites based on prior knowledge on enzymatic reactions, then provide suggestions for potential metabolite identifications, or detect clusters of co-regulated metabolites. In this review, we first aim at settling a nomenclature and formalism to avoid confusion when referring to different networks used in the field of metabolomics. Then, we present the state of the art of network-based methods for mass spectrometry-based metabolomics data analysis, as well as future developments expected in this area. We cover the use of networks applications using biochemical reactions, mass spectrometry features, chemical structural similarities, and correlations between metabolites. We also describe the application of knowledge networks such as metabolic reaction networks. Finally, we discuss the possibility of combining different networks to analyze and interpret them simultaneously.

Suggesting disease associations for overlooked metabolites using literature from metabolic neighbors.

In human health research, metabolic signatures extracted from metabolomics data have a strong added value for stratifying patients and identifying biomarkers. Nevertheless, one of the main challenges is to interpret and relate these lists of discriminant metabolites to pathological mechanisms. This task requires experts to combine their knowledge with information extracted from databases and the scientific literature. However, we show that most compounds (>99%) in the PubChem database lack annotated literature. This dearth of available information can have a direct impact on the interpretation of metabolic signatures, which is often restricted to a subset of significant metabolites. To suggest potential pathological phenotypes related to overlooked metabolites that lack annotated literature, we extend the "guilt-by-association" principle to literature information by using a Bayesian framework. The underlying assumption is that the literature associated with the metabolic neighbors of a compound can provide valuable insights, or an a priori, into its biomedical context. The metabolic neighborhood of a compound can be defined from a metabolic network and correspond to metabolites to which it is connected through biochemical reactions. With the proposed approach, we suggest more than 35,000 associations between 1,047 overlooked metabolites and 3,288 diseases (or disease families). All these newly inferred associations are freely available on the FORUM ftp server (see information at https://github.com/eMetaboHUB/Forum-LiteraturePropagation).

Pathway analysis in metabolomics: Recommendations for the use of over-representation analysis.

Over-representation analysis (ORA) is one of the commonest pathway analysis approaches used for the functional interpretation of metabolomics datasets. Despite the widespread use of ORA in metabolomics, the community lacks guidelines detailing its best-practice use. Many factors have a pronounced impact on the results, but to date their effects have received little systematic attention. Using five publicly available datasets, we demonstrated that changes in parameters such as the background set, differential metabolite selection methods, and pathway database used can result in profoundly different ORA results. The use of a non-assay-specific background set, for example, resulted in large numbers of false-positive pathways. Pathway database choice, evaluated using three of the most popular metabolic pathway databases (KEGG, Reactome, and BioCyc), led to vastly different results in both the number and function of significantly enriched pathways. Factors that are specific to metabolomics data, such as the reliability of compound identification and the chemical bias of different analytical platforms also impacted ORA results. Simulated metabolite misidentification rates as low as 4% resulted in both gain of false-positive pathways and loss of truly significant pathways across all datasets. Our results have several practical implications for ORA users, as well as those using alternative pathway analysis methods. We offer a set of recommendations for the use of ORA in metabolomics, alongside a set of minimal reporting guidelines, as a first step towards the standardisation of pathway analysis in metabolomics.

FORUM: building a Knowledge Graph from public databases and scientific literature to extract associations between chemicals and diseases.

MOTIVATION: Metabolomics studies aim at reporting a metabolic signature (list of metabolites) related to a particular experimental condition. These signatures are instrumental in the identification of biomarkers or classification of individuals, however their biological and physiological interpretation remains a challenge. To support this task, we introduce FORUM: a Knowledge Graph (KG) providing a semantic representation of relations between chemicals and biomedical concepts, built from a federation of life science databases and scientific literature repositories. RESULTS: The use of a Semantic Web framework on biological data allows us to apply ontological-based reasoning to infer new relations between entities. We show that these new relations provide different levels of abstraction and could open the path to new hypotheses. We estimate the statistical relevance of each extracted relation, explicit or inferred, using an enrichment analysis, and instantiate them as new knowledge in the KG to support results interpretation/further inquiries. AVAILABILITY AND IMPLEMENTATION: A web interface to browse and download the extracted relations, as well as a SPARQL endpoint to directly probe the whole FORUM KG, are available at https://forum-webapp.semantic-metabolomics.fr. The code needed to reproduce the triplestore is available at https://github.com/eMetaboHUB/Forum-DiseasesChem. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Atopic dermatitis or eczema? Consequences of ambiguity in disease name for biomedical literature mining.

BACKGROUND: Biomedical research increasingly relies on computational approaches to extract relevant information from large corpora of publications. OBJECTIVE: To investigate the consequence of the ambiguity between the use of terms "Eczema" and "Atopic Dermatitis" (AD) from the Information Retrieval perspective, and its impact on meta-analyses, systematic reviews and text mining. METHODS: Articles were retrieved by querying the PubMed using terms 'eczema' (D003876) and "dermatitis, atopic" (D004485). We used machine learning to investigate the differences between the contexts in which each term is used. We used a decision tree approach and trained model to predict if an article would be indexed with eczema or AD tags. We used text-mining tools to extract biological entities associated with eczema and AD, and investigated the discrepancy regarding the retrieval of key findings according to the terminology used. RESULTS: Atopic dermatitis query yielded more articles related to veterinary science, biochemistry, cellular and molecular biology; the eczema query linked to public health, infectious disease and respiratory system. Medical Subject Headings terms associated with "AD" or "Eczema" differed, with an agreement between the top 40 lists of 52%. The presence of terms related to cellular mechanisms, especially allergies and inflammation, characterized AD literature. The metabolites mentioned more frequently than expected in articles with AD tag differed from those indexed with eczema. Fewer enriched genes were retrieved when using eczema compared to AD query. CONCLUSIONS AND CLINICAL RELEVANCE: There is a considerable discrepancy when using text mining to extract bio-entities related to eczema or AD. Our results suggest that any systematic approach (particularly when looking for metabolites or genes related to the condition) should be performed using both terms jointly. We propose to use decision tree learning as a tool to spot and characterize ambiguity, and provide the source code for disambiguation at https://github.com/cfrainay/ResearchCodeBase.

MetaboRank: network-based recommendation system to interpret and enrich metabolomics results.

Motivation: Metabolomics has shown great potential to improve the understanding of complex diseases, potentially leading to therapeutic target identification. However, no single analytical method allows monitoring all metabolites in a sample, resulting in incomplete metabolic fingerprints. This incompleteness constitutes a stumbling block to interpretation, raising the need for methods that can enrich those fingerprints. We propose MetaboRank, a new solution inspired by social network recommendation systems for the identification of metabolites potentially related to a metabolic fingerprint. Results: MetaboRank method had been used to enrich metabolomics data obtained on cerebrospinal fluid samples from patients suffering from hepatic encephalopathy (HE). MetaboRank successfully recommended metabolites not present in the original fingerprint. The quality of recommendations was evaluated by using literature automatic search, in order to check that recommended metabolites could be related to the disease. Complementary mass spectrometry experiments and raw data analysis were performed to confirm these suggestions. In particular, MetaboRank recommended the overlooked α-ketoglutaramate as a metabolite which should be added to the metabolic fingerprint of HE, thus suggesting that metabolic fingerprints enhancement can provide new insight on complex diseases. Availability and implementation: Method is implemented in the MetExplore server and is available at www.metexplore.fr. A tutorial is available at https://metexplore.toulouse.inra.fr/com/tutorials/MetaboRank/2017-MetaboRank.pdf. Supplementary information: Supplementary data are available at Bioinformatics online.

Large-Scale Modeling Approach Reveals Functional Metabolic Shifts during Hepatic Differentiation.

Being able to explore the metabolism of broad metabolizing cells is of critical importance in many research fields. This article presents an original modeling solution combining metabolic network and omics data to identify modulated metabolic pathways and changes in metabolic functions occurring during differentiation of a human hepatic cell line (HepaRG). Our results confirm the activation of hepato-specific functionalities and newly evidence modulation of other metabolic pathways, which could not be evidenced from transcriptomic data alone. Our method takes advantage of the network structure to detect changes in metabolic pathways that do not have gene annotations and exploits flux analyses techniques to identify activated metabolic functions. Compared to the usual cell-specific metabolic network reconstruction approaches, it limits false predictions by considering several possible network configurations to represent one phenotype rather than one arbitrarily selected network. Our approach significantly enhances the comprehensive and functional assessment of cell metabolism, opening further perspectives to investigate metabolic shifts occurring within various biological contexts.

Machine learning to identify pairwise interactions between specific IgE antibodies and their association with asthma: A cross-sectional analysis within a population-based birth cohort.

BACKGROUND: The relationship between allergic sensitisation and asthma is complex; the data about the strength of this association are conflicting. We propose that the discrepancies arise in part because allergic sensitisation may not be a single entity (as considered conventionally) but a collection of several different classes of sensitisation. We hypothesise that pairings between immunoglobulin E (IgE) antibodies to individual allergenic molecules (components), rather than IgE responses to 'informative' molecules, are associated with increased risk of asthma. METHODS AND FINDINGS: In a cross-sectional analysis among 461 children aged 11 years participating in a population-based birth cohort, we measured serum-specific IgE responses to 112 allergen components using a multiplex array (ImmunoCAP Immuno­Solid phase Allergy Chip [ISAC]). We characterised sensitivity to 44 active components (specific immunoglobulin E [sIgE] > 0.30 units in at least 5% of children) among the 213 (46.2%) participants sensitised to at least one of these 44 components. We adopted several machine learning methodologies that offer a powerful framework to investigate the highly complex sIgE-asthma relationship. Firstly, we applied network analysis and hierarchical clustering (HC) to explore the connectivity structure of component-specific IgEs and identify clusters of component-specific sensitisation ('component clusters'). Of the 44 components included in the model, 33 grouped in seven clusters (C.sIgE-1-7), and the remaining 11 formed singleton clusters. Cluster membership mapped closely to the structural homology of proteins and/or their biological source. Components in the pathogenesis-related (PR)-10 proteins cluster (C.sIgE-5) were central to the network and mediated connections between components from grass (C.sIgE-4), trees (C.sIgE-6), and profilin clusters (C.sIgE-7) with those in mite (C.sIgE-1), lipocalins (C.sIgE-3), and peanut clusters (C.sIgE-2). We then used HC to identify four common 'sensitisation clusters' among study participants: (1) multiple sensitisation (sIgE to multiple components across all seven component clusters and singleton components), (2) predominantly dust mite sensitisation (IgE responses mainly to components from C.sIgE-1), (3) predominantly grass and tree sensitisation (sIgE to multiple components across C.sIgE-4-7), and (4) lower-grade sensitisation. We used a bipartite network to explore the relationship between component clusters, sensitisation clusters, and asthma, and the joint density-based nonparametric differential interaction network analysis and classification (JDINAC) to test whether pairwise interactions of component-specific IgEs are associated with asthma. JDINAC with pairwise interactions provided a good balance between sensitivity (0.84) and specificity (0.87), and outperformed penalised logistic regression with individual sIgE components in predicting asthma, with an area under the curve (AUC) of 0.94, compared with 0.73. We then inferred the differential network of pairwise component-specific IgE interactions, which demonstrated that 18 pairs of components predicted asthma. These findings were confirmed in an independent sample of children aged 8 years who participated in the same birth cohort but did not have component-resolved diagnostics (CRD) data at age 11 years. The main limitation of our study was the exclusion of potentially important allergens caused by both the ISAC chip resolution as well as the filtering step. Clustering and the network analyses might have provided different solutions if additional components had been available. CONCLUSIONS: Interactions between pairs of sIgE components are associated with increased risk of asthma and may provide the basis for designing diagnostic tools for asthma.

Classification of Pediatric Asthma: From Phenotype Discovery to Clinical Practice.

Advances in big data analytics have created an opportunity for a step change in unraveling mechanisms underlying the development of complex diseases such as asthma, providing valuable insights that drive better diagnostic decision-making in clinical practice, and opening up paths to individualized treatment plans. However, translating findings from data-driven analyses into meaningful insights and actionable solutions requires approaches and tools which move beyond mining and patterning longitudinal data. The purpose of this review is to summarize recent advances in phenotyping of asthma, to discuss key hurdles currently hampering the translation of phenotypic variation into mechanistic insights and clinical setting, and to suggest potential solutions that may address these limitations and accelerate moving discoveries into practice. In order to advance the field of phenotypic discovery, greater focus should be placed on investigating the extent of within-phenotype variation. We advocate a more cautious modeling approach by "supervising" the findings to delineate more precisely the characteristics of the individual trajectories assigned to each phenotype. Furthermore, it is important to employ different methods within a study to compare the stability of derived phenotypes, and to assess the immutability of individual assignments to phenotypes. If we are to make a step change toward precision (stratified or personalized) medicine and capitalize on the available big data assets, we have to develop genuine cross-disciplinary collaborations, wherein data scientists who turn data into information using algorithms and machine learning, team up with medical professionals who provide deep insights on specific subjects from a clinical perspective.

Mind the Gap: Mapping Mass Spectral Databases in Genome-Scale Metabolic Networks Reveals Poorly Covered Areas.

The use of mass spectrometry-based metabolomics to study human, plant and microbial biochemistry and their interactions with the environment largely depends on the ability to annotate metabolite structures by matching mass spectral features of the measured metabolites to curated spectra of reference standards. While reference databases for metabolomics now provide information for hundreds of thousands of compounds, barely 5% of these known small molecules have experimental data from pure standards. Remarkably, it is still unknown how well existing mass spectral libraries cover the biochemical landscape of prokaryotic and eukaryotic organisms. To address this issue, we have investigated the coverage of 38 genome-scale metabolic networks by public and commercial mass spectral databases, and found that on average only 40% of nodes in metabolic networks could be mapped by mass spectral information from standards. Next, we deciphered computationally which parts of the human metabolic network are poorly covered by mass spectral libraries, revealing gaps in the eicosanoids, vitamins and bile acid metabolism. Finally, our network topology analysis based on the betweenness centrality of metabolites revealed the top 20 most important metabolites that, if added to MS databases, may facilitate human metabolome characterization in the future.

Integrated transcriptomics and metabolomics reveal signatures of lipid metabolism dysregulation in HepaRG liver cells exposed to PCB 126.

Chemical pollutant exposure is a risk factor contributing to the growing epidemic of non-alcoholic fatty liver disease (NAFLD) affecting human populations that consume a western diet. Although it is recognized that intoxication by chemical pollutants can lead to NAFLD, there is limited information available regarding the mechanism by which typical environmental levels of exposure can contribute to the onset of this disease. Here, we describe the alterations in gene expression profiles and metabolite levels in the human HepaRG liver cell line, a validated model for cellular steatosis, exposed to the polychlorinated biphenyl (PCB) 126, one of the most potent chemical pollutants that can induce NAFLD. Sparse partial least squares classification of the molecular profiles revealed that exposure to PCB 126 provoked a decrease in polyunsaturated fatty acids as well as an increase in sphingolipid levels, concomitant with a decrease in the activity of genes involved in lipid metabolism. This was associated with an increased oxidative stress reflected by marked disturbances in taurine metabolism. A gene ontology analysis showed hallmarks of an activation of the AhR receptor by dioxin-like compounds. These changes in metabolome and transcriptome profiles were observed even at the lowest concentration (100 pM) of PCB 126 tested. A decrease in docosatrienoate levels was the most sensitive biomarker. Overall, our integrated multi-omics analysis provides mechanistic insight into how this class of chemical pollutant can cause NAFLD. Our study lays the foundation for the development of molecular signatures of toxic effects of chemicals causing fatty liver diseases to move away from a chemical risk assessment based on in vivo animal experiments.

MetExplore: collaborative edition and exploration of metabolic networks.

Metabolism of an organism is composed of hundreds to thousands of interconnected biochemical reactions responding to environmental or genetic constraints. This metabolic network provides a rich knowledge to contextualize omics data and to elaborate hypotheses on metabolic modulations. Nevertheless, performing this kind of integrative analysis is challenging for end users with not sufficiently advanced computer skills since it requires the use of various tools and web servers. MetExplore offers an all-in-one online solution composed of interactive tools for metabolic network curation, network exploration and omics data analysis. In particular, it is possible to curate and annotate metabolic networks in a collaborative environment. The network exploration is also facilitated in MetExplore by a system of interactive tables connected to a powerful network visualization module. Finally, the contextualization of metabolic elements in the network and the calculation of over-representation statistics make it possible to interpret any kind of omics data. MetExplore is a sustainable project maintained since 2010 freely available at https://metexplore.toulouse.inra.fr/metexplore2/.

Targeted versus untargeted omics - the CAFSA story.

BACKGROUND: In 2009, untargeted metabolomics led to the delineation of a new clinico-biological entity called cerebellar ataxia with elevated cerebrospinal free sialic acid, or CAFSA. In order to elucidate CAFSA, we applied sequentially targeted and untargeted omic approaches. METHODS AND RESULTS: First, we studied five of the six CAFSA patients initially described. Besides increased CSF free sialic acid concentrations, three patients presented with markedly decreased 5-methyltetrahydrofolate (5-MTHF) CSF concentrations. Exome sequencing identified a homozygous POLG mutation in two affected sisters, but failed to identify a causative gene in the three sporadic patients with high sialic acid but low 5-MTHF. Using targeted mass spectrometry, we confirmed that free sialic acid was increased in the CSF of a third known POLG-mutated patient. We then pursued pathophysiological analyses of CAFSA using mass spectrometry-based metabolomics on CSF from two sporadic CAFSA patients as well as 95 patients with an unexplained encephalopathy and 39 controls. This led to the identification of a common metabotype between the two initial CAFSA patients and three additional patients, including one patient with Kearns-Sayre syndrome. Metabolites of the CSF metabotype were positioned in a reconstruction of the human metabolic network, which highlighted the proximity of the metabotype with acetyl-CoA and carnitine, two key metabolites regulating mitochondrial energy homeostasis. CONCLUSION: Our genetic and metabolomics analyses suggest that CAFSA is a heterogeneous entity related to mitochondrial DNA alterations either through POLG mutations or a mechanism similar to what is observed in Kearns-Sayre syndrome.

MetExploreViz: web component for interactive metabolic network visualization.

SUMMARY: MetExploreViz is an open source web component that can be easily embedded in any web site. It provides features dedicated to the visualization of metabolic networks and pathways and thus offers a flexible solution to analyse omics data in a biochemical context. AVAILABILITY AND IMPLEMENTATION: Documentation and link to GIT code repository (GPL 3.0 license) are available at this URL: http://metexplore.toulouse.inra.fr/metexploreViz/doc/.

Computational methods to identify metabolic sub-networks based on metabolomic profiles.

Untargeted metabolomics makes it possible to identify compounds that undergo significant changes in concentration in different experimental conditions. The resulting metabolomic profile characterizes the perturbation concerned, but does not explain the underlying biochemical mechanisms. Bioinformatics methods make it possible to interpret results in light of the whole metabolism. This knowledge is modelled into a network, which can be mined using algorithms that originate in graph theory. These algorithms can extract sub-networks related to the compounds identified. Several attempts have been made to adapt them to obtain more biologically meaningful results. However, there is still no consensus on this kind of analysis of metabolic networks. This review presents the main graph approaches used to interpret metabolomic data using metabolic networks. Their advantages and drawbacks are discussed, and the impacts of their parameters are emphasized. We also provide some guidelines for relevant sub-network extraction and also suggest a range of applications for most methods.