Global Prioritization of Disease Candidate Metabolites Based on a Multi-omics Composite Network

Identification of metabolite-disease associations based on knowledge graph

BACKGROUND

Despite the insights that metabolite analysis can provide into the onset, development, and progression of diseases-thus offering new concepts and methodologies for prevention, diagnosis, and treatment-traditional wet lab experiments are often time-consuming and labor-intensive. Consequently, this study aimed to develop a machine learning model named COM-RAN, which is based on a knowledge graph and random forest algorithm, to identify potential associations between metabolites and diseases.

METHODS

Firstly, we integrated the known associations between diseases and metabolites. Secondly, we provided a synthesis of the extant data regarding diseases and metabolites, accompanied by supplementary information pertinent to these entities. Thirdly, knowledge graph-based embedded features were used to characterize disease-metabolite associations. Finally, a random forest algorithm was employed to construct a model for identifying potential disease-metabolite associations.

RESULTS

The experimental results demonstrated that the proposed model achieved an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.968 in 5-fold cross-validations, while the Area Under the Precision-Recall Curve (AUPR) was 0.901, outperforming the vast majority of existing prediction methods. The case studies corroborated the majority of the novel associations identified by COM-RAN, thereby further demonstrating the reliability of the current method in predicting the potential relationship between metabolites and diseases.

CONCLUSION

The COM-RAN model demonstrated promise in predicting associations between diseases and metabolites, suggesting that integrating knowledge graphs with machine learning methodologies can significantly improve the accuracy and reliability of predictions related to disease-associated metabolites.

Patterns of infant fecal metabolite concentrations and social behavioral development in toddlers

BACKGROUND

Gut-derived metabolites, products of microbial and host co-metabolism, may inform mechanisms underlying children's neurodevelopment. We investigated whether infant fecal metabolites were related to toddler social behavior.

METHODS

Stool samples collected from 6-week-olds (n = 86) and 1-year-olds (n = 209) in the New Hampshire Birth Cohort Study (NHBCS) were analyzed using nuclear magnetic resonance spectroscopy metabolomics. Autism-related behavior in 3-year-olds was assessed by caregivers using the Social Responsiveness Scale (SRS-2). To assess the association between metabolites and SRS-2 scores, we used a traditional single-metabolite approach, quantitative metabolite set enrichment (QEA), and self-organizing maps (SOMs).

RESULTS

Using a single-metabolite approach and QEA, no individual fecal metabolite or metabolite set at either age was associated with SRS-2 scores. Using the SOM method, fecal metabolites of six-week-olds organized into four profiles, which were unrelated to SRS-2 scores. In 1-year-olds, one of twelve fecal metabolite profiles was associated with fewer autism-related behaviors, with SRS-2 scores 3.4 (95%CI: -7, 0.2) points lower than the referent group. This profile had higher concentrations of lactate and lower concentrations of short chain fatty acids than the reference.

CONCLUSIONS

We uncovered metabolic profiles in infant stool associated with subsequent social behavior, highlighting one potential mechanism by which gut bacteria may influence neurobehavior.

IMPACT

Differences in host and microbial metabolism may explain variability in neurobehavioral phenotypes, but prior studies do not have consistent results. We applied three statistical techniques to explore fecal metabolite differences related to social behavior, including self-organizing maps (SOMs), a novel machine learning algorithm. A 1-year-old fecal metabolite pattern characterized by high lactate and low short-chain fatty acid concentrations, identified using SOMs, was associated with social behavior less indicative of autism spectrum disorder. Our findings suggest that social behavior may be related to metabolite profiles and that future studies may uncover novel findings by applying the SOM algorithm.

Implications of metabolism on multi-systems healthy aging across the lifespan

Aging is increasingly thought to involve dysregulation of metabolism in multiple organ systems that culminate in decreased functional capacity and morbidity. Here, we seek to understand complex interactions among metabolism, aging, and systems-wide phenotypes across the lifespan. Among 2469 adults (mean age 74.7 years; 38% Black) in the Health, Aging and Body Composition study we identified metabolic cross-sectionally correlates across 20 multi-dimensional aging-related phenotypes spanning seven domains. We used LASSO-PCA and bioinformatic techniques to summarize metabolome-phenome relationships and derive metabolic scores, which were subsequently linked to healthy aging, mortality, and incident outcomes (cardiovascular disease, disability, dementia, and cancer) over 9 years. To clarify the relationship of metabolism in early adulthood to aging, we tested association of these metabolic scores with aging phenotypes/outcomes in 2320 participants (mean age 32.1, 44% Black) of the Coronary Artery Risk Development in Young Adults (CARDIA) study. We observed significant overlap in metabolic correlates across the seven aging domains, specifying pathways of mitochondrial/cellular energetics, host-commensal metabolism, inflammation, and oxidative stress. Across four metabolic scores (body composition, mental-physical performance, muscle strength, and physical activity), we found strong associations with healthy aging and incident outcomes, robust to adjustment for risk factors. Metabolic scores for participants four decades younger in CARDIA were related to incident cardiovascular, metabolic, and neurocognitive performance, as well as long-term cardiovascular disease and mortality over three decades. Conserved metabolic states are strongly related to domain-specific aging and outcomes over the life-course relevant to energetics, host-commensal interactions, and mechanisms of innate immunity.

Phomactinine, the First Nitrogen-Bearing Phomactin, Produced by Biatriospora sp. CBMAI 1333

Metabolomics analyses and improvement of growth conditions were applied toward diversification of phomactin terpenoids by the fungus Biatriospora sp. CBMAI 1333. Visualization of molecular networking results on Gephi assisted the observation of phomactin diversification and guided the isolation of new phomactin variants by applying a modified version of chemometrics based on a fractional factorial design. Consequentially, the first nitrogen-bearing phomactin, phomactinine (1), with a new rearranged carbon skeleton, was isolated and identified. The strategy combining metabolomics and chemometrics can be extended to include bioassay potency, structure novelty, and metabolic diversification connected or not to genomic analyses.

Metabolite-disease interaction prediction based on logistic matrix factorization and local neighborhood constraints

Background

Increasing evidence indicates that metabolites are closely related to human diseases. Identifying disease-related metabolites is especially important for the diagnosis and treatment of disease. Previous works have mainly focused on the global topological information of metabolite and disease similarity networks. However, the local tiny structure of metabolites and diseases may have been ignored, leading to insufficiency and inaccuracy in the latent metabolite-disease interaction mining.

Methods

To solve the aforementioned problem, we propose a novel metabolite-disease interaction prediction method with logical matrix factorization and local nearest neighbor constraints (LMFLNC). First, the algorithm constructs metabolite-metabolite and disease-disease similarity networks by integrating multi-source heterogeneous microbiome data. Then, the local spectral matrices based on these two networks are established and used as the input of the model, together with the known metabolite-disease interaction network. Finally, the probability of metabolite-disease interaction is calculated according to the learned latent representations of metabolites and diseases.

Results

Extensive experiments on the metabolite-disease interaction data were conducted. The results show that the proposed LMFLNC method outperformed the second-best algorithm by 5.28 and 5.61% in the AUPR and F1, respectively. The LMFLNC method also exhibited several potential metabolite-disease interactions, such as "Cortisol" (HMDB0000063), relating to "21-Hydroxylase deficiency," and "3-Hydroxybutyric acid" (HMDB0000011) and "Acetoacetic acid" (HMDB0000060), both relating to "3-Hydroxy-3-methylglutaryl-CoA lyase deficiency."

Conclusion

The proposed LMFLNC method can well preserve the geometrical structure of original data and can thus effectively predict the underlying associations between metabolites and diseases. The experimental results show its effectiveness in metabolite-disease interaction prediction.

MultiNEP: a multi-omics network enhancement framework for prioritizing disease genes and metabolites simultaneously

MOTIVATION

Many studies have successfully used network information to prioritize candidate omics profiles associated with diseases. The metabolome, as the link between genotypes and phenotypes, has accumulated growing attention. Using a "multi-omics" network constructed with a gene-gene network, a metabolite-metabolite network, and a gene-metabolite network to simultaneously prioritize candidate disease-associated metabolites and gene expressions could further utilize gene-metabolite interactions that are not used when prioritizing them separately. However, the number of metabolites is usually 100 times fewer than that of genes. Without accounting for this imbalance issue, we cannot effectively use gene-metabolite interactions when simultaneously prioritizing disease-associated metabolites and genes.

RESULTS

Here, we developed a Multi-omics Network Enhancement Prioritization (MultiNEP) framework with a weighting scheme to reweight contributions of different sub-networks in a multi-omics network to effectively prioritize candidate disease-associated metabolites and genes simultaneously. In simulation studies, MultiNEP outperforms competing methods that do not address network imbalances and identifies more true signal genes and metabolites simultaneously when we down-weight relative contributions of the gene-gene network and up-weight that of the metabolite-metabolite network to the gene-metabolite network. Applications to two human cancer cohorts show that MultiNEP prioritizes more cancer-related genes by effectively using both within- and between-omics interactions after handling network imbalance.

AVAILABILITY AND IMPLEMENTATION

The developed MultiNEP framework is implemented in an R package and available at: https://github.com/Karenxzr/MultiNep.

Multi-omics profiling approach in food allergy

The prevalence of food allergy (FA) among children is increasing, affecting nearly 8% of children, and FA is the most common cause of anaphylaxis and anaphylaxis-related emergency department visits in children. Importantly, FA is a complex, multi-system, multifactorial disease mediated by food-specific immunoglobulin E (IgE) and type 2 immune responses and involving environmental and genetic factors and gene-environment interactions. Early exposure to external and internal environmental factors largely influences the development of immune responses to allergens. Genetic factors and gene-environment interactions have established roles in the FA pathophysiology. To improve diagnosis and identification of FA therapeutic targets, high-throughput omics approaches have emerged and been applied over the past decades to screen for potential FA biomarkers, such as genes, transcripts, proteins, and metabolites. In this article, we provide an overview of the current status of FA omics studies, namely genomic, transcriptomic, epigenomic, proteomic, exposomic, and metabolomic. The current development of multi-omics integration of FA studies is also briefly discussed. As individual omics technologies only provide limited information on the multi-system biological processes of FA, integration of population-based multi-omics data and clinical data may lead to robust biomarker discovery that could translate into advances in disease management and clinical care and ultimately lead to precision medicine approaches.

Integrated Analysis of Gut Microbiome and Lipid Metabolism in Mice Infected with Carbapenem-Resistant Enterobacteriaceae

The disturbance in gut microbiota composition and metabolism has been implicated in the process of pathogenic bacteria infection. However, the characteristics of the microbiota and the metabolic interaction of commensals−host during pathogen invasion remain more than vague. In this study, the potential associations of gut microbes with disturbed lipid metabolism in mice upon carbapenem-resistant Escherichia coli (CRE) infection were explored by the biochemical and multi-omics approaches including metagenomics, metabolomics and lipidomics, and then the key metabolites−reaction−enzyme−gene interaction network was constructed. Results showed that intestinal Erysipelotrichaceae family was strongly associated with the hepatic total cholesterol and HDL-cholesterol, as well as a few sera and fecal metabolites involved in lipid metabolism such as 24, 25-dihydrolanosterol. A high-coverage lipidomic analysis further demonstrated that a total of 529 lipid molecules was significantly enriched and 520 were depleted in the liver of mice infected with CRE. Among them, 35 lipid species showed high correlations (|r| > 0.8 and p < 0.05) with the Erysipelotrichaceae family, including phosphatidylglycerol (42:2), phosphatidylglycerol (42:3), phosphatidylglycerol (38:5), phosphatidylcholine (42:4), ceramide (d17:1/16:0), ceramide (d18:1/16:0) and diacylglycerol (20:2), with correlation coefficients higher than 0.9. In conclusion, the systematic multi-omics study improved the understanding of the complicated connection between the microbiota and the host during pathogen invasion, which thereby is expected to lead to the future discovery and establishment of novel control strategies for CRE infection.

Deciphering the Core Metabolites of Fanconi Anemia by Using a Multi-Omics Composite Network

Deciphering the metabolites of human diseases is an important objective of biomedical research. Here, we aimed to capture the core metabolites of Fanconi anemia (FA) using the bioinformatics method of a multi-omics composite network. Based on the assumption that metabolite levels can directly mirror the physiological state of the human body, we used a multi-omics composite network that integrates six types of interactions in humans (gene-gene, disease phenotype-phenotype, disease-related metabolite-metabolite, gene-phenotype, gene-metabolite, and metabolite-phenotype) to procure the core metabolites of FA. This method is applicable in predicting and prioritizing disease candidate metabolites and is effective in a network without known disease metabolites. In this report, we first singled out the differentially expressed genes upon different groups that were related with FA and then constructed the multi-omics composite network of FA by integrating the aforementioned six networks. Ultimately, we utilized random walk with restart (RWR) to screen the prioritized candidate metabolites of FA, and meanwhile the co-expression gene network of FA was also obtained. As a result, the top 5 metabolites of FA were tenormin (TN), guanosine 5'-triphosphate, guanosine 5'-diphosphate, triphosadenine (DCF) and adenosine 5'-diphosphate, all of which were reported to have a direct or indirect relationship with FA. Furthermore, the top 5 co-expressed genes were CASP3, BCL2, HSPD1, RAF1 and MMP9. By prioritizing the metabolites, the multi-omics composite network may provide us with additional indicators closely linked to FA.

Interpretation of network-based integration from multi-omics longitudinal data

Multi-omics integration is key to fully understand complex biological processes in an holistic manner. Furthermore, multi-omics combined with new longitudinal experimental design can unreveal dynamic relationships between omics layers and identify key players or interactions in system development or complex phenotypes. However, integration methods have to address various experimental designs and do not guarantee interpretable biological results. The new challenge of multi-omics integration is to solve interpretation and unlock the hidden knowledge within the multi-omics data. In this paper, we go beyond integration and propose a generic approach to face the interpretation problem. From multi-omics longitudinal data, this approach builds and explores hybrid multi-omics networks composed of both inferred and known relationships within and between omics layers. With smart node labelling and propagation analysis, this approach predicts regulation mechanisms and multi-omics functional modules. We applied the method on 3 case studies with various multi-omics designs and identified new multi-layer interactions involved in key biological functions that could not be revealed with single omics analysis. Moreover, we highlighted interplay in the kinetics that could help identify novel biological mechanisms. This method is available as an R package netOmics to readily suit any application.

Heterogeneous information network and its application to human health and disease

The molecular components with the functional interdependencies in human cell form complicated biological network. Diseases are mostly caused by the perturbations of the composite of the interaction multi-biomolecules, rather than an abnormality of a single biomolecule. Furthermore, new biological functions and processes could be revealed by discovering novel biological entity relationships. Hence, more and more biologists focus on studying the complex biological system instead of the individual biological components. The emergence of heterogeneous information network (HIN) offers a promising way to systematically explore complicated and heterogeneous relationships between various molecules for apparently distinct phenotypes. In this review, we first present the basic definition of HIN and the biological system considered as a complex HIN. Then, we discuss the topological properties of HIN and how these can be applied to detect network motif and functional module. Afterwards, methodologies of discovering relationships between disease and biomolecule are presented. Useful insights on how HIN aids in drug development and explores human interactome are provided. Finally, we analyze the challenges and opportunities for uncovering combinatorial patterns among pharmacogenomics and cell-type detection based on single-cell genomic data.

Omics Multi-Layers Networks Provide Novel Mechanistic and Functional Insights Into Fat Storage and Lipid Metabolism in Poultry

Fatty acid metabolism in poultry has a major impact on production and disease resistance traits. According to the high rate of interactions between lipid metabolism and its regulating properties, a holistic approach is necessary. To study omics multilayers of adipose tissue and identification of genes and miRNAs involved in fat metabolism, storage and endocrine signaling pathways in two groups of broiler chickens with high and low abdominal fat, as well as high-throughput techniques, were used. The gene-miRNA interacting bipartite and metabolic-signaling networks were reconstructed using their interactions. In the analysis of microarray and RNA-Seq data, 1,835 genes were detected by comparing the identified genes with significant expression differences (p.adjust < 0.01, fold change ≥ 2 and ≤ -2). Then, by comparing between different data sets, 34 genes and 19 miRNAs were detected as common and main nodes. A literature mining approach was used, and seven genes were identified and added to the common gene set. Module finding revealed three important and functional modules, which were involved in the peroxisome proliferator-activated receptor (PPAR) signaling pathway, biosynthesis of unsaturated fatty acids, Alzheimer's disease metabolic pathway, adipocytokine, insulin, PI3K-Akt, mTOR, and AMPK signaling pathway. This approach revealed a new insight to better understand the biological processes associated with adipose tissue.

Network Pharmacology and Metabolomics Studies on Antimigraine Mechanisms of Da Chuan Xiong Fang (DCXF)

Background

Da Chuan Xiong Fang (DCXF) is a traditional Chinese medicine (TCM) formula used to treat migraines. Previously, we uncovered partial mechanisms involved in the therapeutic actions of DCXF on migraines.

Methods

In this study, we further elucidated its antimigraine mechanisms in vivo by using an integrated strategy coupling with network pharmacology and metabolomics techniques.

Results

Network pharmacology identified 33 genes linked with both migraine and DCXF, most of which were 5-hydroxytryptamine receptors, dopamine, and peptide receptors. The results of GO and KEGG enrichment analysis showed that DCXF significantly regulated tyrosine metabolism, tryptophan metabolism, dopamine metabolic process, glucose transmembrane transport, lipid metabolism, and fatty acid transport. The results of metabolomics analysis found that the metabolism of tryptophan and tyrosine in the brain tissue and energy and lipid metabolism of rats tended towards normal and reached normal levels after administering DCXF. The metabolomics and network pharmacology approaches demonstrated similar antimigraine effects of DCXF on endogenous neurotransmitters and overall trends in serum and brain tissue. Using both approaches, 62 hub genes were identified from the protein-protein interaction (PPI) network of DCXF and gene-metabolite interaction network, with hub genes and different metabolites in serum and brain tissue. The hub genes of DCXF, which were mostly linked with inflammation, might affect mainly neurotransmitters in serum and brain tissue metabolisms.

Conclusion

Network pharmacology and metabolomics study may help identify hub genes, metabolites, and possible pathways of disease and treatment. Additionally, two parts of the results were integrated to confirm each other. Their combination may help elucidate the relationship between hub genes and metabolites and provide the further understanding of TCM mechanisms.

Comprehensive Analysis Strategy of Nervous-Endocrine-Immune-Related Metabolites to Evaluate Arachidonic Acid as a Novel Diagnostic Biomarker in Depression

Depression is one of the most complex multifactorial diseases affected by genetic and environmental factors. The molecular mechanism underlying depression remains largely unclear. To address this issue, a novel nervous-endocrine-immune (NEI) network module was used to find the metabolites and evaluate the diagnostic ability of patients with depression. During this process, metabolites were acquired from a professional depression metabolism database. Over-representation analysis was performed using IMPaLA. Then, the metabolite-metabolite interaction (MMI) network of the NEI system was used to select key metabolites. Finally, the receiver operating characteristic curve analysis was evaluated for the diagnostic ability of arachidonic acid. The results show that the numbers of the nervous system, endocrine system, and immune system pathways are 10, 19, and 12 and the numbers of metabolites are 38, 52, and 13, respectively. The selected shared metabolite-enriched pathways can be 97.56% of the NEI-related pathways. Arachidonic acid was extracted from the NEI system network by using an optimization formula and validated by in vivo experiments. It was indicated that the proposed model was good at screening arachidonic acid for the diagnosis of depression. This method provides reliable evidences and references for the diagnosis and mechanism research of other related diseases.

An Inductive Logistic Matrix Factorization Model for Predicting Drug-Metabolite Association With Vicus Regularization

Metabolites are closely related to human disease. The interaction between metabolites and drugs has drawn increasing attention in the field of pharmacomicrobiomics. However, only a small portion of the drug-metabolite interactions were experimentally observed due to the fact that experimental validation is labor-intensive, costly, and time-consuming. Although a few computational approaches have been proposed to predict latent associations for various bipartite networks, such as miRNA-disease, drug-target interaction networks, and so on, to our best knowledge the associations between drugs and metabolites have not been reported on a large scale. In this study, we propose a novel algorithm, namely inductive logistic matrix factorization (ILMF) to predict the latent associations between drugs and metabolites. Specifically, the proposed ILMF integrates drug-drug interaction, metabolite-metabolite interaction, and drug-metabolite interaction into this framework, to model the probability that a drug would interact with a metabolite. Moreover, we exploit inductive matrix completion to guide the learning of projection matrices U and V that depend on the low-dimensional feature representation matrices of drugs and metabolites: Fm and Fd . These two matrices can be obtained by fusing multiple data sources. Thus, Fd U and Fm V can be viewed as drug-specific and metabolite-specific latent representations, different from classical LMF. Furthermore, we utilize the Vicus spectral matrix that reveals the refined local geometrical structure inherent in the original data to encode the relationships between drugs and metabolites. Extensive experiments are conducted on a manually curated "DrugMetaboliteAtlas" dataset. The experimental results show that ILMF can achieve competitive performance compared with other state-of-the-art approaches, which demonstrates its effectiveness in predicting potential drug-metabolite associations.

A Multi-Omics Study Revealing the Metabolic Effects of Estrogen in Liver Cancer Cells HepG2

Hepatocellular carcinoma (HCC) that is triggered by metabolic defects is one of the most malignant liver cancers. A much higher incidence of HCC among men than women suggests the protective roles of estrogen in HCC development and progression. To begin to understand the mechanisms involving estrogenic metabolic effects, we compared cell number, viability, cytotoxicity, and apoptosis among HCC-derived HepG2 cells that were treated with different concentrations of 2-deoxy-d-glucose (2-DG) that blocks glucose metabolism, oxamate that inhibits lactate dehydrogenase and glycolysis, or oligomycin that blocks ATP synthesis and mitochondrial oxidative phosphorylation. We confirmed that HepG2 cells primarily utilized glycolysis followed by lactate fermentation, instead of mitochondrial oxidative phosphorylation, for cell growth. We hypothesized that estrogen altered energy metabolism via its receptors to carry out its anticancer effects in HepG2 cells. We treated cells with 17β-estradiol (E2), 1,3,5-tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole (PPT) an estrogen receptor (ER) α (ERα) agonist, or 2,3-bis(4-hydroxyphenyl)-propionitrile (DPN), an ERβ agonist. We then used transcriptomic and metabolomic analyses and identified differentially expressed genes and unique metabolite fingerprints that are produced by each treatment. We further performed integrated multi-omics analysis, and identified key genes and metabolites in the gene–metabolite interaction contributed by E2 and ER agonists. This integrated transcriptomic and metabolomic study suggested that estrogen acts on estrogen receptors to suppress liver cancer cell growth via altering metabolism. This is the first exploratory study that comprehensively investigated estrogen and its receptors, and their roles in regulating gene expression, metabolites, metabolic pathways, and gene–metabolite interaction in HCC cells using bioinformatic tools. Overall, this study provides potential therapeutic targets for future HCC treatment.

Multi-omics integration in biomedical research - A metabolomics-centric review

Recent advances in high-throughput technologies have enabled the profiling of multiple layers of a biological system, including DNA sequence data (genomics), RNA expression levels (transcriptomics), and metabolite levels (metabolomics). This has led to the generation of vast amounts of biological data that can be integrated in so-called multi-omics studies to examine the complex molecular underpinnings of health and disease. Integrative analysis of such datasets is not straightforward and is particularly complicated by the high dimensionality and heterogeneity of the data and by the lack of universal analysis protocols. Previous reviews have discussed various strategies to address the challenges of data integration, elaborating on specific aspects, such as network inference or feature selection techniques. Thereby, the main focus has been on the integration of two omics layers in their relation to a phenotype of interest. In this review we provide an overview over a typical multi-omics workflow, focusing on integration methods that have the potential to combine metabolomics data with two or more omics. We discuss multiple integration concepts including data-driven, knowledge-based, simultaneous and step-wise approaches. We highlight the application of these methods in recent multi-omics studies, including large-scale integration efforts aiming at a global depiction of the complex relationships within and between different biological layers without focusing on a particular phenotype.

Advancing from protein interactomes and gene co-expression networks towards multi-omics-based composite networks: approaches for predicting and extracting biological knowledge

Prediction of biological interaction networks from single-omics data has been extensively implemented to understand various aspects of biological systems. However, more recently, there is a growing interest in integrating multi-omics datasets for the prediction of interactomes that provide a global view of biological systems with higher descriptive capability, as compared to single omics. In this review, we have discussed various computational approaches implemented to infer and analyze two of the most important and well studied interactomes: protein-protein interaction networks and gene co-expression networks. We have explicitly focused on recent methods and pipelines implemented to infer and extract biologically important information from these interactomes, starting from utilizing single-omics data and then progressing towards multi-omics data. Accordingly, recent examples and case studies are also briefly discussed. Overall, this review will provide a proper understanding of the latest developments in protein and gene network modelling and will also help in extracting practical knowledge from them.

Using MetaboAnalyst 4.0 for Comprehensive and Integrative Metabolomics Data Analysis

MetaboAnalyst (https://www.metaboanalyst.ca) is an easy-to-use web-based tool suite for comprehensive metabolomic data analysis, interpretation, and integration with other omics data. Since its first release in 2009, MetaboAnalyst has evolved significantly to meet the ever-expanding bioinformatics demands from the rapidly growing metabolomics community. In addition to providing a variety of data processing and normalization procedures, MetaboAnalyst supports a wide array of functions for statistical, functional, as well as data visualization tasks. Some of the most widely used approaches include PCA (principal component analysis), PLS-DA (partial least squares discriminant analysis), clustering analysis and visualization, MSEA (metabolite set enrichment analysis), MetPA (metabolic pathway analysis), biomarker selection via ROC (receiver operating characteristic) curve analysis, as well as time series and power analysis. The current version of MetaboAnalyst (4.0) features a complete overhaul of the user interface and significantly expanded underlying knowledge bases (compound database, pathway libraries, and metabolite sets). Three new modules have been added to support pathway activity prediction directly from mass peaks, biomarker meta-analysis, and network-based multi-omics data integration. To enable more transparent and reproducible analysis of metabolomic data, we have released a companion R package (MetaboAnalystR) to complement the web-based application. This article provides an overview of the main functional modules and the general workflow of MetaboAnalyst 4.0, followed by 12 detailed protocols: © 2019 by John Wiley & Sons, Inc. Basic Protocol 1: Data uploading, processing, and normalization Basic Protocol 2: Identification of significant variables Basic Protocol 3: Multivariate exploratory data analysis Basic Protocol 4: Functional interpretation of metabolomic data Basic Protocol 5: Biomarker analysis based on receiver operating characteristic (ROC) curves Basic Protocol 6: Time-series and two-factor data analysis Basic Protocol 7: Sample size estimation and power analysis Basic Protocol 8: Joint pathway analysis Basic Protocol 9: MS peaks to pathway activities Basic Protocol 10: Biomarker meta-analysis Basic Protocol 11: Knowledge-based network exploration of multi-omics data Basic Protocol 12: MetaboAnalystR introduction.

Construction and Analysis of Human Diseases and Metabolites Network

The relationship between aberrant metabolism and the initiation and progression of diseases has gained considerable attention in recent years. To gain insights into the global relationship between diseases and metabolites, here we constructed a human diseases-metabolites network (HDMN). Through analyses based on network biology, the metabolites associated with the same disorder tend to participate in the same metabolic pathway or cascade. In addition, the shortest distance between disease-related metabolites was shorter than that of all metabolites in the Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic network. Both disease and metabolite nodes in the HDMN displayed slight clustering phenomenon, resulting in functional modules. Furthermore, a significant positive correlation was observed between the degree of metabolites and the proportion of disease-related metabolites in the KEGG metabolic network. We also found that the average degree of disease metabolites is larger than that of all metabolites. Depicting a comprehensive characteristic of HDMN could provide great insights into understanding the global relationship between disease and metabolites.

Multi-network logistic matrix factorization for metabolite-disease interaction prediction

Identifying disease-related metabolites is of great significance for the diagnosis, prevention, and treatment of disease. In this study, we propose a novel computational model of multiple-network logistic matrix factorization (MN-LMF) for predicting metabolite-disease interactions, which is especially relevant for new diseases and new metabolites. First, MN-LMF builds disease (or metabolite) similarity network by integrating heterogeneous omics data. Second, it combines these similarities with known metabolite-disease interaction networks, using modified logistic matrix factorization to predict potential metabolite-disease interactions. Experimental results show that MN-LMF accurately predicts metabolite-disease interactions, and outperforms other state-of-the-art methods. Moreover, case studies also demonstrated the effectiveness of the model to infer unknown metabolite-disease interactions for novel diseases without any known associations.

Heterogeneous Multi-Layered Network Model for Omics Data Integration and Analysis

Advances in next-generation sequencing and high-throughput techniques have enabled the generation of vast amounts of diverse omics data. These big data provide an unprecedented opportunity in biology, but impose great challenges in data integration, data mining, and knowledge discovery due to the complexity, heterogeneity, dynamics, uncertainty, and high-dimensionality inherited in the omics data. Network has been widely used to represent relations between entities in biological system, such as protein-protein interaction, gene regulation, and brain connectivity (i.e. network construction) as well as to infer novel relations given a reconstructed network (aka link prediction). Particularly, heterogeneous multi-layered network (HMLN) has proven successful in integrating diverse biological data for the representation of the hierarchy of biological system. The HMLN provides unparalleled opportunities but imposes new computational challenges on establishing causal genotype-phenotype associations and understanding environmental impact on organisms. In this review, we focus on the recent advances in developing novel computational methods for the inference of novel biological relations from the HMLN. We first discuss the properties of biological HMLN. Then we survey four categories of state-of-the-art methods (matrix factorization, random walk, knowledge graph, and deep learning). Thirdly, we demonstrate their applications to omics data integration and analysis. Finally, we outline strategies for future directions in the development of new HMLN models.

Prioritizing candidate diseases-related metabolites based on literature and functional similarity

Background

As the terminal products of cellular regulatory process, functional related metabolites have a close relationship with complex diseases, and are often associated with the same or similar diseases. Therefore, identification of disease related metabolites play a critical role in understanding comprehensively pathogenesis of disease, aiming at improving the clinical medicine. Considering that a large number of metabolic markers of diseases need to be explored, we propose a computational model to identify potential disease-related metabolites based on functional relationships and scores of referred literatures between metabolites. First, obtaining associations between metabolites and diseases from the Human Metabolome database, we calculate the similarities of metabolites based on modified recommendation strategy of collaborative filtering utilizing the similarities between diseases. Next, a disease-associated metabolite network (DMN) is built with similarities between metabolites as weight. To improve the ability of identifying disease-related metabolites, we introduce scores of text mining from the existing database of chemicals and proteins into DMN and build a new disease-associated metabolite network (FLDMN) by fusing functional associations and scores of literatures. Finally, we utilize random walking with restart (RWR) in this network to predict candidate metabolites related to diseases.

Results

We construct the disease-associated metabolite network and its improved network (FLDMN) with 245 diseases, 587 metabolites and 28,715 disease-metabolite associations. Subsequently, we extract training sets and testing sets from two different versions of the Human Metabolome database and assess the performance of DMN and FLDMN on 19 diseases, respectively. As a result, the average AUC (area under the receiver operating characteristic curve) of DMN is 64.35%. As a further improved network, FLDMN is proven to be successful in predicting potential metabolic signatures for 19 diseases with an average AUC value of 76.03%.

Conclusion

In this paper, a computational model is proposed for exploring metabolite-disease pairs and has good performance in predicting potential metabolites related to diseases through adequate validation. This result suggests that integrating literature and functional associations can be an effective way to construct disease associated metabolite network for prioritizing candidate diseases-related metabolites.

MetaboAnalyst 4.0: towards more transparent and integrative metabolomics analysis

We present a new update to MetaboAnalyst (version 4.0) for comprehensive metabolomic data analysis, interpretation, and integration with other omics data. Since the last major update in 2015, MetaboAnalyst has continued to evolve based on user feedback and technological advancements in the field. For this year's update, four new key features have been added to MetaboAnalyst 4.0, including: (1) real-time R command tracking and display coupled with the release of a companion MetaboAnalystR package; (2) a MS Peaks to Pathways module for prediction of pathway activity from untargeted mass spectral data using the mummichog algorithm; (3) a Biomarker Meta-analysis module for robust biomarker identification through the combination of multiple metabolomic datasets and (4) a Network Explorer module for integrative analysis of metabolomics, metagenomics, and/or transcriptomics data. The user interface of MetaboAnalyst 4.0 has been reengineered to provide a more modern look and feel, as well as to give more space and flexibility to introduce new functions. The underlying knowledgebases (compound libraries, metabolite sets, and metabolic pathways) have also been updated based on the latest data from the Human Metabolome Database (HMDB). A Docker image of MetaboAnalyst is also available to facilitate download and local installation of MetaboAnalyst. MetaboAnalyst 4.0 is freely available at http://metaboanalyst.ca.

Identification of lung-adenocarcinoma-related long non-coding RNAs by random walking on a competing endogenous RNA network

Background

Identification of novel risk long non-coding RNAs (lncRNAs) in lung adenocarcinoma (LUAD) is still a significant challenge in cancer research.

Methods

In this study, we first constructed a LUAD-specific competing endogenous RNA (ceRNA) network using both experimental- and computational-supported datasets. Then, a random walking with restart method was performed to predict LUAD-associated risk lncRNAs based on the ceRNA network. The role of lncRNA MAPKAPK5-AS1 was assessed by siRNA transfection, followed by a colony formation assay, the CCK-8 assay, and immunofluorescence on A549 cells.

Results

Our method achieved an area under the curve (AUC) value of over 0.83. Of the several potential novel LUAD-related lncRNAs identified, the highest ranked lncRNA was SNHG12, which, interestingly, was also shown to promote tumorigenesis and metastasis in LUAD in a recent study. Furthermore, we found that the expression of MAPKAPK5-AS1, which was ranked second, was higher in both LUAD tissues and three LUAD cell lines. After the silencing of MAPKAPK5-AS1 by siRNA transfection, a colony formation assay revealed fewer colonies, and a CCK-8 assay revealed significantly suppressed growth of A549 cells. Moreover, immunofluorescence staining of Ki-67, a proliferation marker, revealed that the proliferation capability of A549 was dramatically reduced following MAPKAPK5-AS1 downregulation. AO/EB staining showed an increased proportion of apoptotic cells among A549 cells depleted of MAPKAPK5-AS1.

Conclusions

In brief, the lncRNAs were predicted to serve as potential biomarkers for the diagnosis, treatment, and prognosis of LUAD.

Combined metabolic, phenomic and genomic data to prioritize atrial fibrillation-related metabolites

Metabolites in atrial fibrillation (AF) were characterized to further explore the molecular mechanisms of AF by integrating metabolic, phenomic and genomic data. Gene expression data on AF (E-GEOD-79768) were downloaded from the EMBL-EBI database, followed by identification of differentially expressed genes (DEGs) which were used to construct gene-gene network. Then, multi-omics composite networks were constructed. Subsequently, random walk with restart was expanded to a multi-omics composite network to identify and prioritize the metabolites according to the AF-related seed genes deposited in the OMIM database, the whole metabolome as candidates and the phenotype of AF. Using the interaction score among metabolites, we extracted the top 50 metabolites, and identified the top 100 co-expressed genes interacted with the top 50 metabolites. Based on the FDR <0.05, 622 DEGs were extracted. In order to demonstrate the intrinsic mode of this method, we sorted the metabolites of the composite network in descending order based on the interaction scores. The top 5 metabolites were respectively weighed potassium, sodium ion, chitin, benzo[a]pyrene-7,8-dihydrodiol-9,10-oxide, and celebrex (TN). Potassium and sodium ion possessed higher degrees in the subnetwork of the entire composite network and the co-expressed network. Metabolites such as potassium and sodium ion may provide valuable clues for early diagnostic and therapeutic targets for AF.

Long Noncoding RNA and Protein Interactions: From Experimental Results to Computational Models Based on Network Methods

Non-coding RNAs with a length of more than 200 nucleotides are long non-coding RNAs (lncRNAs), which have gained tremendous attention in recent decades. Many studies have confirmed that lncRNAs have important influence in post-transcriptional gene regulation; for example, lncRNAs affect the stability and translation of splicing factor proteins. The mutations and malfunctions of lncRNAs are closely related to human disorders. As lncRNAs interact with a variety of proteins, predicting the interaction between lncRNAs and proteins is a significant way to depth exploration functions and enrich annotations of lncRNAs. Experimental approaches for lncRNA–protein interactions are expensive and time-consuming. Computational approaches to predict lncRNA–protein interactions can be grouped into two broad categories. The first category is based on sequence, structural information and physicochemical property. The second category is based on network method through fusing heterogeneous data to construct lncRNA related heterogeneous network. The network-based methods can capture the implicit feature information in the topological structure of related biological heterogeneous networks containing lncRNAs, which is often ignored by sequence-based methods. In this paper, we summarize and discuss the materials, interaction score calculation algorithms, advantages and disadvantages of state-of-the-art algorithms of lncRNA–protein interaction prediction based on network methods to assist researchers in selecting a suitable method for acquiring more dependable results. All the related different network data are also collected and processed in convenience of users, and are available at https://github.com/HAN-Siyu/APINet/.

Identification of atherosclerosis-related prioritizing metabolites based on a multi-omics composite network

Metabolites are the final products of cellular regulation processes, their level is the ultimate response of biological systems to environmental and genetic changes. Therefore, the identification of key metabolites is required for the diagnosis and therapy of diseases. In this study, atherosclerosis-related gene expression profile information was extracted from ArrayExpress database (GEOD-57691), and analyzed with limma package. Furthermore, we constructed an intricate multi-omics network involved in genes, phenotypes, metabolites and their associations. To identify the prioritization of atherosclerosis-related metabolites, the relation score of each metabolite in the composite network was computed with the random walk with restart (RWR) method. The top 50 metabolites and top 100 genes were chosen based on the score in the weighted composite network. Consequently, several key metabolites that were ranked in the top 5 of relation score or degree greater than 70 were confirmed. Particularly, metabolites Tretinoin and Estraderm not only have high relation scores, but also contain more degrees. Moreover, we obtained 24 co-expression genes that may be regarded as the targets of atherosclerosis therapy. Therefore, identification of metabolite prioritizations by the composite network integrated the information of genes, phenotypes and metabolites may be available to diagnose atherosclerosis, and can provide the potential therapeutic strategies for atherosclerosis.

Prioritization of candidate metabolites for postmenopausal osteoporosis using multi-omics composite network

Risk metabolites of postmenopausal osteoporosis (PO) were explored to offer a theoretical basis for future therapy. The data E-GEOD-7429 were downloaded from ArrayExpress database. In total 20 samples deprived from postmenopausal women having low or high bone mineral density (BMD) were covered in this expression profile. After screening of differentially expressed genes (DEGs), gene-gene network was constructed taking the intersection between the DEGs and genes in the seed protein-protein interaction network. Then, the other five networks were established, including metabolite, phenotype, gene-metabolite, phenotype-gene, and phenotype-metabolite networks. Next, these 6 networks were integrated into one weighted multi-omics network to further identify the candidate metabolites using random walk with restart based on the PO-related seed genes, seed metabolites and phenotype. Using the score among nodes of the weighted composite network, the top 50 metabolites, and the top 100 co-expressed genes interacting with the top 50 metabolites were detected. A set of 601 DEGs between low BMD and high BMD samples were selected. Significantly, the top 5 metabolites were respectively glucosylgalactosyl hydroxylysine, all-trans-5,6-epoxyretinoic acid, tretinoin, colecalciferol, and rocaltrol. Moreover, 3 metabolites (estraderm, triphosadenine, and tretinoin) had a degree >50 in the co-expression network. Tretinoin was the member of the top 5 metabolites, and estraderm was a metabolite with the seventh interaction score. A series of metabolites, tretinoin and estraderm might be closely associated with the onset and progression of PO.

Identification of Key Metabolites for Acute Lung Injury in Patients with Sepsis

BACKGROUND

The study aimed to detect critical metabolites in acute lung injury (ALI).

METHODS

A comparative analysis of microarray profile of patients with sepsis-induced ALI compared with sepsis patients with was conducted using bioinformatic tools through constructing multi-omics network. Multi-omics composite networks (gene network, metabolite network, phenotype network, gene-metabolite association network, phenotype-gene association network, and phenotype-metabolite association network) were constructed, following by integration of these composite networks to establish a heterogeneous network. Next, seed genes, and ALI phenotype were mapped into the heterogeneous network to further obtain a weighted composite network. Random walk with restart (RWR) was used for the weighted composite network to extract and prioritize the metabolites. On the basis of the distance proximity among metabolites, the top 50 metabolites with the highest proximity were identified, and the top 100 co-expressed genes interacted with the top 50 metabolites were also screened out.

RESULTS

Totally, there were 9363 nodes and 10,226,148 edges in the integrated composite network. There were 4 metabolites with the scores > 0.009, including CHITIN, Tretinoin, sodium ion, and Celebrex. Adenosine 5'-diphosphate, triphosadenine, and tretinoin had higher degrees in the composite network and the co-expressed network.

CONCLUSION

Adenosine 5'-diphosphate, triphosadenine, and tretinoin may be potential biomarkers for diagnosis and treatment of ALI.

MetabolitePredict: A de novo human metabolomics prediction system and its applications in rheumatoid arthritis

Human metabolomics has great potential in disease mechanism understanding, early diagnosis, and therapy. Existing metabolomics studies are often based on profiling patient biofluids and tissue samples and are difficult owing to the challenges of sample collection and data processing. Here, we report an alternative approach and developed a computation-based prediction system, MetabolitePredict, for disease metabolomics biomarker prediction. We applied MetabolitePredict to identify metabolite biomarkers and metabolite targeting therapies for rheumatoid arthritis (RA), a last-lasting complex disease with multiple genetic and environmental factors involved. MetabolitePredict is a de novo prediction system. It first constructs a disease-specific genetic profile using genes and pathways data associated with an input disease. It then constructs genetic profiles for a total of 259,170 chemicals/metabolites using known chemical genetics and human metabolomic data. MetabolitePredict prioritizes metabolites for a given disease based on the genetic profile similarities between disease and metabolites. We evaluated MetabolitePredict using 63 known RA-associated metabolites. MetabolitePredict found 24 of the 63 metabolites (recall: 0.38) and ranked them highly (mean ranking: top 4.13%, median ranking: top 1.10%, P-value: 5.08E-19). MetabolitePredict performed better than an existing metabolite prediction system, PROFANCY, in predicting RA-associated metabolites (PROFANCY: recall: 0.31, mean ranking: 20.91%, median ranking: 16.47%, P-value: 3.78E-7). Short-chain fatty acids (SCFAs), the abundant metabolites of gut microbiota in the fermentation of fiber, ranked highly (butyrate, 0.03%; acetate, 0.05%; propionate, 0.38%). Finally, we established MetabolitePredict's potential in novel metabolite targeting for disease treatment: MetabolitePredict ranked highly three known metabolite inhibitors for RA treatments (methotrexate:0.25%; leflunomide: 0.56%; sulfasalazine: 0.92%). MetabolitePredict is a generalizable disease metabolite prediction system. The only required input to the system is a disease name or a set of disease-associated genes. The web-based MetabolitePredict is available at:http://xulab.

CASE

edu/MetabolitePredict.

Multi-OMICS: a critical technical perspective on integrative lipidomics approaches

During the past decades, high-throughput approaches for analyzing different molecular classes such as nucleic acids, proteins, metabolites, and lipids have grown rapidly. These approaches became powerful tools for getting a fundamental understanding of biological systems. Considering each approach and its results separately, relations and causal connections between these classes have no chance to be revealed, since only separate molecular snapshots are provided. Only a combined approach, not fully established yet, with the integration of the corresponding data, might yield a comprehensive and complete understanding of biological processes, such as crosstalk and interactions in signaling pathways. Taking two or more omics-methods into consideration for analysis is referred to as a multi-omics approach, which is gradually evolving. In this critical note, we briefly discuss the relevance, challenges, current state, and potential of data integration from multi-omics approaches, with a special focus on lipidomics analysis, listing the advantages and gaps in this field. This article is part of a Special Issue entitled: BBALIP_Lipidomics Opinion Articles edited by Sepp Kohlwein.

Global Prioritizing Disease Candidate lncRNAs via a Multi-level Composite Network

LncRNAs play pivotal roles in many important biological processes, but research on the functions of lncRNAs in human disease is still in its infancy. Therefore, it is urgent to prioritize lncRNAs that are potentially associated with diseases. In this work, we developed a novel algorithm, LncPriCNet, that uses a multi-level composite network to prioritize candidate lncRNAs associated with diseases. By integrating genes, lncRNAs, phenotypes and their associations, LncPriCNet achieves an overall performance superior to that of previous methods, with high AUC values of up to 0.93. Notably, LncPriCNet still performs well when information on known disease lncRNAs is lacking. When applied to breast cancer, LncPriCNet identified known breast cancer-related lncRNAs, revealed novel lncRNA candidates and inferred their functions via pathway analysis. We further constructed the human disease-lncRNA landscape, revealed the modularity of the disease-lncRNA network and identified several lncRNA hotspots. In summary, LncPriCNet is a useful tool for prioritizing disease-related lncRNAs and may facilitate understanding of the molecular mechanisms of human disease at the lncRNA level.

RWCFusion: identifying phenotype-specific cancer driver gene fusions based on fusion pair random walk scoring method

While gene fusions have been increasingly detected by next-generation sequencing (NGS) technologies based methods in human cancers, these methods have limitations in identifying driver fusions. In addition, the existing methods to identify driver gene fusions ignored the specificity among different cancers or only considered their local rather than global topology features in networks. Here, we proposed a novel network-based method, called RWCFusion, to identify phenotype-specific cancer driver gene fusions. To evaluate its performance, we used leave-one-out cross-validation in 35 cancers and achieved a high AUC value 0.925 for overall cancers and an average 0.929 for signal cancer. Furthermore, we classified 35 cancers into two classes: haematological and solid, of which the haematological got a highly AUC which is up to 0.968. Finally, we applied RWCFusion to breast cancer and found that top 13 gene fusions, such as BCAS3-BCAS4, NOTCH-NUP214, MED13-BCAS3 and CARM-SMARCA4, have been previously proved to be drivers for breast cancer. Additionally, 8 among the top 10 of the remaining candidate gene fusions, such as SULF2-ZNF217, MED1-ACSF2, and ACACA-STAC2, were inferred to be potential driver gene fusions of breast cancer by us.