A causal network analysis in an observational study identifies metabolomics pathways influencing plasma triglyceride levels

Metabolomic profiles in Jamaican children with and without autism spectrum disorder

Background

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition with a wide range of behavioral and cognitive impairments. While genetic and environmental factors are known to contribute to its etiology, the underlying metabolic perturbations associated with ASD which can potentially connect genetic and environmental factors, remain poorly understood. Therefore, we conducted a metabolomic case-control study and performed a comprehensive analysis to identify significant alterations in metabolite profiles between children with ASD and typically developing (TD) controls.

Objective

To elucidate potential metabolomic signatures associated with ASD in children and identify specific metabolites that may serve as biomarkers for the disorder.

Methods

We conducted metabolomic profiling on plasma samples from participants in the second phase of Epidemiological Research on Autism in Jamaica (ERAJ-2), which was a 1:1 age (±6 months)-and sex-matched cohort of 200 children with ASD and 200 TD controls (2-8 years old). Using high-throughput liquid chromatography-mass spectrometry techniques, we performed a targeted metabolite analysis, encompassing amino acids, lipids, carbohydrates, and other key metabolic compounds. After quality control and imputation of missing values, we performed univariable and multivariable analysis using normalized metabolites while adjusting for covariates, age, sex, socioeconomic status, and child's parish of birth.

Results

Our findings revealed unique metabolic patterns in children with ASD for four metabolites compared to TD controls. Notably, three of these metabolites were fatty acids, including myristoleic acid, eicosatetraenoic acid, and octadecenoic acid. Additionally, the amino acid sarcosine exhibited a significant association with ASD.

Conclusions

These findings highlight the role of metabolites in the etiology of ASD and suggest opportunities for the development of targeted interventions.

A fat- and sucrose-enriched diet causes metabolic alterations in mdx mice

Duchenne muscular dystrophy (DMD), a progressive muscle disease caused by the absence of functional dystrophin protein, is associated with multiple cellular, physiological, and metabolic dysfunctions. As an added complication to the primary insult, obesity/insulin resistance (O/IR) is frequently reported in patients with DMD; however, how IR impacts disease severity is unknown. We hypothesized a high-fat, high-sucrose diet (HFHSD) would induce O/IR, exacerbate disease severity, and cause metabolic alterations in dystrophic mice. To test this hypothesis, we treated 7-wk-old mdx (disease model) and C57 mice with a control diet (CD) or an HFHSD for 15 wk. The HFHSD induced insulin resistance, glucose intolerance, and hyperglycemia in C57 and mdx mice. Of note, mdx mice on CD were also insulin resistant. In addition, visceral adipose tissue weights were increased with HFHSD in C57 and mdx mice though differed by genotype. Serum creatine kinase activity and histopathological analyses using Masson's trichrome staining in the diaphragm indicated muscle damage was driven by dystrophin deficiency but was not augmented by diet. In addition, markers of inflammatory signaling, mitochondrial abundance, and autophagy were impacted by disease but not diet. Despite this, in addition to disease signatures in CD-fed mice, metabolomic and lipidomic analyses demonstrated a HFHSD caused some common changes in C57 and mdx mice and some unique signatures of O/IR within the context of dystrophin deficiency. In total, these data revealed that in mdx mice, 15 wk of HFHSD did not overtly exacerbate muscle injury but further impaired the metabolic status of dystrophic muscle.

Inference of causal metabolite networks in the presence of invalid instrumental variables with GWAS summary data

We propose structural equation models (SEMs) as a general framework to infer causal networks for metabolites and other complex traits. Traditionally SEMs are used only for individual-level data under the assumption that all instrumental variables (IVs) are valid. To overcome these limitations, we propose both one- and two-sample approaches for causal network inference based on SEMs that can: (1) perform causal analysis and discover causal relationships among multiple traits; (2) account for the possible presence of some invalid IVs; (3) allow for data analysis using only genome-wide association studies (GWAS) summary statistics when individual-level data are not available; (4) consider the possibility of bidirectional relationships between traits. Our method employs a simple stepwise selection to identify invalid IVs, thus avoiding false positives while possibly increasing true discoveries based on two-stage least squares (2SLS). We use both real GWAS data and simulated data to demonstrate the superior performance of our method over the standard 2SLS/SEMs. For real data analysis, our proposed approach is applied to a human blood metabolite GWAS summary data set to uncover putative causal relationships among the metabolites; we also identify some metabolites (putative) causal to Alzheimer's disease (AD), which, along with the inferred causal metabolite network, suggest some possible pathways of metabolites involved in AD.

Deep neural networks with knockoff features identify nonlinear causal relations and estimate effect sizes in complex biological systems

BACKGROUND

Learning the causal structure helps identify risk factors, disease mechanisms, and candidate therapeutics for complex diseases. However, although complex biological systems are characterized by nonlinear associations, existing bioinformatic methods of causal inference cannot identify the nonlinear relationships and estimate their effect size.

RESULTS

To overcome these limitations, we developed the first computational method that explicitly learns nonlinear causal relations and estimates the effect size using a deep neural network approach coupled with the knockoff framework, named causal directed acyclic graphs using deep learning variable selection (DAG-deepVASE). Using simulation data of diverse scenarios and identifying known and novel causal relations in molecular and clinical data of various diseases, we demonstrated that DAG-deepVASE consistently outperforms existing methods in identifying true and known causal relations. In the analyses, we also illustrate how identifying nonlinear causal relations and estimating their effect size help understand the complex disease pathobiology, which is not possible using other methods.

CONCLUSIONS

With these advantages, the application of DAG-deepVASE can help identify driver genes and therapeutic agents in biomedical studies and clinical trials.

The UK Biobank: A Shining Example of Genome-Wide Association Study Science with the Power to Detect the Murky Complications of Real-World Epidemiology

Genome-wide association studies (GWASs) have successfully identified thousands of genetic variants that are reliably associated with human traits. Although GWASs are restricted to certain variant frequencies, they have improved our understanding of the genetic architecture of complex traits and diseases. The UK Biobank (UKBB) has brought substantial analytical opportunity and performance to association studies. The dramatic expansion of many GWAS sample sizes afforded by the inclusion of UKBB data has improved the power of estimation of effect sizes but, critically, has done so in a context where phenotypic depth and precision enable outcome dissection and the application of epidemiological approaches. However, at the same time, the availability of such a large, well-curated, and deeply measured population-based collection has the capacity to increase our exposure to the many complications and inferential complexities associated with GWASs and other analyses. In this review, we discuss the impact that UKBB has had in the GWAS era, some of the opportunities that it brings, and exemplar challenges that illustrate the reality of using data from this world-leading resource.

Bounded Fuzzy Possibilistic Method Reveals Information about Lung Cancer through Analysis of Metabolomics

Learning methods, such as conventional clustering and classification, have been applied in diagnosing diseases to categorize samples based on their features. Going beyond clustering samples, membership degrees represent to what degree each sample belongs to a cluster. Variation of membership degrees in each cluster provides information about the cluster as a whole and each sample individually which enables us to have insights toward precision medicine. Membership degrees are measured more accurately through removing restrictions from clustering samples. Bounded Fuzzy Possibilistic Method (BFPM) introduces a membership function that keeps the search space flexible to cluster samples with higher accuracy. The method evaluates samples for their movement from one cluster to another. This technique allows us to find critical samples in advance those with the potential ability to belong to other clusters in the near future. BFPM was applied on metabolomics of individuals in a lung cancer case-control study. Metabolomics as proximal molecular signals to the actual disease processes may serve as strong biomarkers of current disease process. The goal is to know whether serum metabolites of a healthy human can be differentiated from those with lung cancer. Using BFPM, some differences were observed, the pathology data were evaluated, and critical samples were recognized.

Insulin Resistance during normal child growth and development is associated with a distinct blood metabolic phenotype (Earlybird 72)

BACKGROUND

While insulin resistance (IR) is associated with specific metabolite signatures in adults, there have been few truly longitudinal studies in healthy children, either to confirm which abnormalities are present, or to determine whether they precede or result from IR. Therefore, we investigated the association of serum metabolites with IR in childhood in the Earlybird cohort.

METHODS

The Earlybird cohort is a well-characterized cohort of healthy children with annual measurements from age 5 to 16 years. For the first time, longitudinal association analyses between individual serum metabolites and homeostatic model assessment (HOMA) of insulin resistance (HOMA-IR) have been performed taking into account the effects of age, growth, puberty, adiposity, and physical activity.

RESULTS

IR was higher in girls than in boys and was associated with increasing body mass index (BMI). In longitudinal analysis IR was associated with reduced concentrations of branched-chain amino acids (BCAA), 2-ketobutyrate, citrate and 3-hydroxybutyrate, and higher concentrations of lactate and alanine. These findings demonstrate the widespread biochemical consequences of IR for intermediary metabolism, ketogenesis, and pyruvate oxidation during normal child growth and development.

CONCLUSIONS

Longitudinal analysis can differentiate metabolite signatures that precede or follow the development of greater levels of IR. In healthy normal weight children, higher levels of IR are associated with reduced levels of BCAA, ketogenesis, and fuel oxidation. In contrast, elevated lactate concentrations preceded the rise in IR. These changes reveal the metabolite signature of insulin action during normal growth, and they contrast with previous findings in obese children and adults that represent the consequences of IR and obesity.

Comparison of metabolite networks from four German population-based studies

Background

Metabolite networks are suggested to reflect biological pathways in health and disease. However, it is unknown whether such metabolite networks are reproducible across different populations. Therefore, the current study aimed to investigate similarity of metabolite networks in four German population-based studies.

Methods

One hundred serum metabolites were quantified in European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam (n = 2458), EPIC-Heidelberg (n = 812), KORA (Cooperative Health Research in the Augsburg Region) (n = 3029) and CARLA (Cardiovascular Disease, Living and Ageing in Halle) (n = 1427) with targeted metabolomics. In a cross-sectional analysis, Gaussian graphical models were used to construct similar networks of 100 edges each, based on partial correlations of these metabolites. The four metabolite networks of the top 100 edges were compared based on (i) common features, i.e. number of common edges, Pearson correlation (r) and hamming distance (h); and (ii) meta-analysis of the four networks.

Results

Among the four networks, 57 common edges and 66 common nodes (metabolites) were identified. Pairwise network comparisons showed moderate to high similarity (r = 63–0.96, h = 7–72), among the networks. Meta-analysis of the networks showed that, among the 100 edges and 89 nodes of the meta-analytic network, 57 edges and 66 metabolites were present in all the four networks, 58–76 edges and 75–89 nodes were present in at least three networks, and 63–84 edges and 76–87 edges were present in at least two networks. The meta-analytic network showed clear grouping of 10 sphingolipids, 8 lyso-phosphatidylcholines, 31 acyl-alkyl-phosphatidylcholines, 30 diacyl-phosphatidylcholines, 8 amino acids and 2 acylcarnitines.

Conclusions

We found structural similarity in metabolite networks from four large studies. Using a meta-analytic network, as a new approach for combining metabolite data from different studies, closely related metabolites could be identified, for some of which the biological relationships in metabolic pathways have been previously described. They are candidates for further investigation to explore their potential role in biological processes.

Genome analysis and pleiotropy assessment using causal networks with loss of function mutation and metabolomics

BACKGROUND

Many genome-wide association studies have detected genomic regions associated with traits, yet understanding the functional causes of association often remains elusive. Utilizing systems approaches and focusing on intermediate molecular phenotypes might facilitate biologic understanding.

RESULTS

The availability of exome sequencing of two populations of African-Americans and European-Americans from the Atherosclerosis Risk in Communities study allowed us to investigate the effects of annotated loss-of-function (LoF) mutations on 122 serum metabolites. To assess the findings, we built metabolomic causal networks for each population separately and utilized structural equation modeling. We then validated our findings with a set of independent samples. By use of methods based on concepts of Mendelian randomization of genetic variants, we showed that some of the affected metabolites are risk predictors in the causal pathway of disease. For example, LoF mutations in the gene KIAA1755 were identified to elevate the levels of eicosapentaenoate (p-value = 5E-14), an essential fatty acid clinically identified to increase essential hypertension. We showed that this gene is in the pathway to triglycerides, where both triglycerides and essential hypertension are risk factors of metabolomic disorder and heart attack. We also identified that the gene CLDN17, harboring loss-of-function mutations, had pleiotropic actions on metabolites from amino acid and lipid pathways.

CONCLUSION

Using systems biology approaches for the analysis of metabolomics and genetic data, we integrated several biological processes, which lead to findings that may functionally connect genetic variants with complex diseases.

Metabolic Biomarkers and Neurodegeneration: A Pathway Enrichment Analysis of Alzheimer's Disease, Parkinson's Disease, and Amyotrophic Lateral Sclerosis

Neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and amyotrophic lateral sclerosis (ALS) lack robust diagnostics and prognostic biomarkers. Metabolomics is a postgenomics field that offers fresh insights for biomarkers of common complex as well as rare diseases. Using data on metabolite-disease associations published in the previous decade (2006-2016) in PubMed, ScienceDirect, Scopus, and Web of Science, we identified 101 metabolites as putative biomarkers for these three neurodegenerative diseases. Notably, uric acid, choline, creatine, L-glutamine, alanine, creatinine, and N-acetyl-L-aspartate were the shared metabolite signatures among the three diseases. The disease-metabolite-pathway associations pointed out the importance of membrane transport (through ATP binding cassette transporters), particularly of arginine and proline amino acids in all three neurodegenerative diseases. When disease-specific and common metabolic pathways were queried by using the pathway enrichment analyses, we found that alanine, aspartate, glutamate, and purine metabolism might act as alternative pathways to overcome inadequate glucose supply and energy crisis in neurodegeneration. These observations underscore the importance of metabolite-based biomarker research in deciphering the elusive pathophysiology of neurodegenerative diseases. Future research investments in metabolomics of complex diseases might provide new insights on AD, PD, and ALS that continue to place a significant burden on global health.

A comparison of methods for inferring causal relationships between genotype and phenotype using additional biological measurements

Genome wide association studies (GWAS) have been very successful over the last decade at identifying genetic variants associated with disease phenotypes. However, interpretation of the results obtained can be challenging. Incorporation of further relevant biological measurements (e.g. ‘omics’ data) measured in the same individuals for whom we have genotype and phenotype data may help us to learn more about the mechanism and pathways through which causal genetic variants affect disease. We review various methods for causal inference that can be used for assessing the relationships between genetic variables, other biological measures, and phenotypic outcome, and present a simulation study assessing the performance of the methods under different conditions. In general, the methods we considered did well at inferring the causal structure for data simulated under simple scenarios. However, the presence of an unknown and unmeasured common environmental effect could lead to spurious inferences, with the methods we considered displaying varying degrees of robustness to this confounder. The use of causal inference techniques to integrate omics and GWAS data has the potential to improve biological understanding of the pathways leading to disease. Our study demonstrates the suitability of various methods for performing causal inference under several biologically plausible scenarios.

Identification, analysis, and interpretation of a human serum metabolomics causal network in an observational study

Untargeted metabolomics, measurement of large numbers of metabolites irrespective of their chemical or biologic characteristics, has proven useful for identifying novel biomarkers of health and disease. Of particular importance is the analysis of networks of metabolites, as opposed to the level of an individual metabolite. The aim of this study is to achieve causal inference among serum metabolites in an observational setting. A metabolomics causal network is identified using the genome granularity directed acyclic graph (GDAG) algorithm where information across the genome in a deeper level of granularity is extracted to create strong instrumental variables and identify causal relationships among metabolites in an upper level of granularity. Information from 1,034,945 genetic variants distributed across the genome was used to identify a metabolomics causal network among 122 serum metabolites. We introduce individual properties within the network, such as strength of a metabolite. Based on these properties, hypothesized targets for intervention and prediction are identified. Four nodes corresponding to the metabolites leucine, arichidonoyl-glycerophosphocholine, N-acyelyalanine, and glutarylcarnitine had high impact on the entire network by virtue of having multiple arrows pointing out, which propagated long distances. Five modules, largely corresponding to functional metabolite categories (e.g. amino acids), were identified over the network and module boundaries were determined using directionality and causal effect sizes. Two families, each consists of a triangular motif identified in the network had essential roles in the network by virtue of influencing a large number of other nodes. We discuss causal effect measurement while confounders and mediators are identified graphically.

A Causal Network Analysis of the Fatty Acid Metabolome in African-Americans Reveals a Critical Role for Palmitoleate and Margarate

Fatty acids are important sources of energy and possible predictors and etiologic factors in many common complex pathologies such as cardiovascular disease, diabetes, and certain forms of cancers. While fatty acids are thought to covary with each other, their underlying causal networks have not been fully elucidated. This study reports the identification and analysis of a statistical causal network among 15 mostly long-chain fatty acids. In an African-American population sample and using the Genome granularity-Directed Acyclic Graph (GDAG) algorithm, we determined directions or causal relationships in the fatty acid metabolome. A directed causal network was constructed that revealed 29 significant edges among the 15 nodes (p < 0.001). We report that two fatty acid metabolites, palmitoleate and margarate, which originate from dietary intake, together influence every other fatty acid in the network. On the other hand, despite its high connectivity, dihomo-linoleate did not appear to play an important role over the whole fatty acid network. These findings collectively suggest possible strategic entry points for new treatments or preventive modalities against diseases affected by fatty acid metabolites such as cardiovascular disease, diabetes, and obesity. Further studies examining the embedded substructure of the fatty acid metabolite networks in independent population samples would be timely and warranted as we move toward novel postgenomic diagnostics and therapeutics.