Untargeted Metabolomics for Metabolic Diagnostic Screening with Automated Data Interpretation Using a Knowledge-Based Algorithm

The Multi-Omic Approach to Newborn Screening: Opportunities and Challenges

Newborn screening programs have seen significant evolution since their initial implementation more than 60 years ago, with the primary goal of detecting treatable conditions within the earliest possible timeframe to ensure the optimal treatment and outcomes for the newborn. New technologies have driven the expansion of screening programs to cover additional conditions. In the current era, the breadth of screened conditions could be further expanded by integrating omic technologies such as untargeted metabolomics and genomics. Genomic screening could offer opportunities for lifelong care beyond the newborn period. For genomic newborn screening to be effective and ready for routine adoption, it must overcome barriers such as implementation cost, public acceptability, and scalability. Metabolomics approaches, on the other hand, can offer insight into disease phenotypes and could be used to identify known and novel biomarkers of disease. Given recent advances in metabolomic technologies, alongside advances in genomics including whole-genome sequencing, the combination of complementary multi-omic approaches may provide an exciting opportunity to leverage the best of both approaches and overcome their respective limitations. These techniques are described, along with the current outlook on multi-omic-based NBS research.

Viral, Bacterial, Metabolic, and Autoimmune Causes of Severe Acute Encephalopathy in Sub-Saharan Africa: A Multicenter Cohort Study

OBJECTIVES

To assess whether viral, bacterial, metabolic, and autoimmune diseases are missed by conventional diagnostics among children with severe acute encephalopathy in sub-Saharan Africa.

STUDY DESIGN

One hundred thirty-four children (6 months to 18 years) presenting with nontraumatic coma or convulsive status epilepticus to 1 of 4 medical referral centers in Uganda, Malawi, and Rwanda were enrolled between 2015 and 2016. Locally available diagnostic tests could be supplemented in 117 patients by viral, bacterial, and 16s quantitative polymerase chain reaction testing, metagenomics, untargeted metabolomics, and autoimmune immunohistochemistry screening.

RESULTS

Fourteen (12%) cases of viral encephalopathies, 8 (7%) cases of bacterial central nervous system (CNS) infections, and 4 (4%) cases of inherited metabolic disorders (IMDs) were newly identified by additional diagnostic testing as the most likely cause of encephalopathy. No confirmed cases of autoimmune encephalitis were found. Patients for whom additional diagnostic testing aided causal evaluation (aOR 3.59, 90% CI 1.57-8.36), patients with a viral CNS infection (aOR 7.91, 90% CI 2.49-30.07), and patients with an IMD (aOR 9.10, 90% CI 1.37-110.45) were at increased risk for poor outcome of disease.

CONCLUSIONS

Viral and bacterial CNS infections and IMDs are prevalent causes of severe acute encephalopathy in children in Uganda, Malawi, and Rwanda that are missed by conventional diagnostics and are associated with poor outcome of disease. Improved diagnostic capacity may increase diagnostic yield and might improve outcome of disease.

Ex vivo proton spectroscopy (1 H-NMR) analysis of inborn errors of metabolism: Automatic and computer-assisted analyses

There are about 1500 genetic metabolic diseases. A small number of treatable diseases are diagnosed by newborn screening programs, which are continually being developed. However, most diseases can only be diagnosed based on clinical symptoms or metabolic findings. The main biological fluids used are urine, plasma and, in special situations, cerebrospinal fluid. In contrast to commonly used methods such as gas chromatography and high performance liquid chromatography mass spectrometry, ex vivo proton spectroscopy (¹ H-NMR) is not yet used in routine clinical practice, although it has been recommended for more than 30 years. Automatic analysis and improved NMR technology have also expanded the applications used for the diagnosis of inborn errors of metabolism. We provide a mini-overview of typical applications, especially in urine but also in plasma, used to diagnose common but also rare genetic metabolic diseases with ¹ H-NMR. The use of computer-assisted diagnostic suggestions can facilitate interpretation of the profiles. In a proof of principle, to date, 182 reports of 59 different diseases and 500 reports of healthy children are stored. The percentage of correct automatic diagnoses was 74%. Using the same ¹ H-NMR profile-targeted analysis, it is possible to apply an untargeted approach that distinguishes profile differences from healthy individuals. Thus, additional conditions such as lysosomal storage diseases or drug interferences are detectable. Furthermore, because ¹ H-NMR is highly reproducible and can detect a variety of different substance categories, the metabolomic approach is suitable for monitoring patient treatment and revealing additional factors such as nutrition and microbiome metabolism. Besides the progress in analytical techniques, a multiomics approach is most effective to combine metabolomics with, for example, whole exome sequencing, to also diagnose patients with nondetectable metabolic abnormalities in biological fluids. In this mini review we also provide our own data to demonstrate the role of NMR in a multiomics platform in the field of inborn errors of metabolism.

Metabolomic Approach to Screening Homozygotes in Chinese Patients with Severe Familial Hypercholesterolemia

Homozygous familial hypercholesterolemia (HoFH) is a rare inborn-errors-of-metabolism disorder characterized by devastatingly elevated low-density lipoprotein cholesterol (LDL-C) and premature cardiovascular disease. The gold standard for screening and diagnosing HoFH is genetic testing. In China, it is expensive and is always recommended for the most likely HoFH subjects with aggressive LDL-C phenotype. However, the LDL-C levels of HoFH patients and a substantial proportion of heterozygous FH (HeFH) patients overlapped considerably. Here, we performed a cost-effective metabolomic profiling on genetically diagnosed HoFH (n = 69) and HeFH patients (n = 101) with overlapping LDL-C levels, aiming to discovery a unique metabolic pattern for screening homozygotes in patients with severe FH. We demonstrated a differential serum metabolome profile in HoFH patients compared to HeFH patients. Twenty-one metabolomic alterations showed independent capability in differentiating HoFH from severe HeFH. The combined model based on seven identified metabolites yielded a corrected diagnosis in 91.3% of HoFH cases with an area under the curve value of 0.939. Collectively, this study demonstrated that metabolomic profiling serves as a useful and economical approach to preselecting homozygotes in FH patients with severe hypercholesterolemia and may help clinicians to conduct selective genetic confirmation testing and familial cascade screening.

Diagnosing, discarding, or de-VUSsing: A practical guide to (un)targeted metabolomics as variant-transcending functional tests

PURPOSE

For patients with inherited metabolic disorders (IMDs), any diagnostic delay should be avoided because early initiation of personalized treatment could prevent irreversible health damage. To improve diagnostic interpretation of genetic data, gene function tests can be valuable assets. For IMDs, variant-transcending functional tests are readily available through (un)targeted metabolomics assays. To support the application of metabolomics for this purpose, we developed a gene-based guide to select functional tests to either confirm or exclude an IMD diagnosis.

METHODS

Using information from a diagnostic IMD exome panel, Kyoto Encyclopedia of Genes and Genomes, and Inborn Errors of Metabolism Knowledgebase, we compiled a guide for metabolomics-based gene function tests. From our practical experience with this guide, we retrospectively selected illustrative cases for whom combined metabolomic/genomic testing improved diagnostic success and evaluated the effect hereof on clinical management.

RESULTS

The guide contains 2047 metabolism-associated genes for which a validated or putative variant-transcending gene function test is available. We present 16 patients for whom metabolomic testing either confirmed or ruled out the presence of a second pathogenic variant, validated or ruled out pathogenicity of variants of uncertain significance, or identified a diagnosis initially missed by genetic analysis.

CONCLUSION

Metabolomics-based gene function tests provide additional value in the diagnostic trajectory of patients with suspected IMD by enhancing and accelerating diagnostic success.

A narrative review of metabolomics in the era of "-omics": integration into clinical practice for inborn errors of metabolism

BACKGROUND AND OBJECTIVE

Traditional targeted metabolomic investigations identify a pre-defined list of analytes in samples and have been widely used for decades in the diagnosis and monitoring of inborn errors of metabolism (IEMs). Recent technological advances have resulted in the development and maturation of untargeted metabolomics: a holistic, unbiased, analytical approach to detecting metabolic disturbances in human disease. We aim to provide a summary of untargeted metabolomics [focusing on tandem mass spectrometry (MS-MS)] and its application in the field of IEMs.

METHODS

Data for this review was identified through a literature search using PubMed, Google Scholar, and personal repositories of articles collected by the authors. Findings are presented within several sections describing the metabolome, the current use of targeted metabolomics in the diagnostic pathway of patients with IEMs, the more recent integration of untargeted metabolomics into clinical care, and the limitations of this newly employed analytical technique.

KEY CONTENT AND FINDINGS

Untargeted metabolomic investigations are increasingly utilized in screening for rare disorders, improving understanding of cellular and subcellular physiology, discovering novel biomarkers, monitoring therapy, and functionally validating genomic variants. Although the untargeted metabolomic approach has some limitations, this "next generation metabolic screening" platform is becoming increasingly affordable and accessible.

CONCLUSIONS

When used in conjunction with genomics and the other promising "-omic" technologies, untargeted metabolomics has the potential to revolutionize the diagnostics of IEMs (and other rare disorders), improving both clinical and health economic outcomes.

Integration of metabolomics with genomics: Metabolic gene prioritization using metabolomics data and genomic variant (CADD) scores

The integration of metabolomics data with sequencing data is a key step towards improving the diagnostic process for finding the disease-causing genetic variant(s) in patients suspected of having an inborn error of metabolism (IEM). The measured metabolite levels could provide additional phenotypical evidence to elucidate the degree of pathogenicity for variants found in genes associated with metabolic processes. We present a computational approach, called Reafect, that calculates for each reaction in a metabolic pathway a score indicating whether that reaction is deficient or not. When calculating this score, Reafect takes multiple factors into account: the magnitude and sign of alterations in the metabolite levels, the reaction distances between metabolites and reactions in the pathway, and the biochemical directionality of the reactions. We applied Reafect to untargeted metabolomics data of 72 patient samples with a known IEM and found that in 81% of the cases the correct deficient enzyme was ranked within the top 5% of all considered enzyme deficiencies. Next, we integrated Reafect with Combined Annotation Dependent Depletion (CADD) scores (a measure for gene variant deleteriousness) and ranked the metabolic genes of 27 IEM patients. We observed that this integrated approach significantly improved the prioritization of the genes containing the disease-causing variant when compared with the two approaches individually. For 15/27 IEM patients the correct affected gene was ranked within the top 0.25% of the set of potentially affected genes. Together, our findings suggest that metabolomics data improves the identification of affected genes in patients suffering from IEM.

Clinical diagnosis of metabolic disorders using untargeted metabolomic profiling and disease-specific networks learned from profiling data

Untargeted metabolomics is a global molecular profiling technology that can be used to screen for inborn errors of metabolism (IEMs). Metabolite perturbations are evaluated based on current knowledge of specific metabolic pathway deficiencies, a manual diagnostic process that is qualitative, has limited scalability, and is not equipped to learn from accumulating clinical data. Our purpose was to improve upon manual diagnosis of IEMs in the clinic by developing novel computational methods for analyzing untargeted metabolomics data. We employed CTD, an automated computational diagnostic method that "connects the dots" between metabolite perturbations observed in individual metabolomics profiling data and modules identified in disease-specific metabolite co-perturbation networks learned from prior profiling data. We also extended CTD to calculate distances between any two individuals (CTDncd) and between an individual and a disease state (CTDdm), to provide additional network-quantified predictors for use in diagnosis. We show that across 539 plasma samples, CTD-based network-quantified measures can reproduce accurate diagnosis of 16 different IEMs, including adenylosuccinase deficiency, argininemia, argininosuccinic aciduria, aromatic L-amino acid decarboxylase deficiency, cerebral creatine deficiency syndrome type 2, citrullinemia, cobalamin biosynthesis defect, GABA-transaminase deficiency, glutaric acidemia type 1, maple syrup urine disease, methylmalonic aciduria, ornithine transcarbamylase deficiency, phenylketonuria, propionic acidemia, rhizomelic chondrodysplasia punctata, and the Zellweger spectrum disorders. Our approach can be used to supplement information from biochemical pathways and has the potential to significantly enhance the interpretation of variants of uncertain significance uncovered by exome sequencing. CTD, CTDdm, and CTDncd can serve as an essential toolset for biological interpretation of untargeted metabolomics data that overcomes limitations associated with manual diagnosis to assist diagnosticians in clinical decision-making. By automating and quantifying the interpretation of perturbation patterns, CTD can improve the speed and confidence by which clinical laboratory directors make diagnostic and treatment decisions, while automatically improving performance with new case data.

Targeted urine metabolomics with a graphical reporting tool for rapid diagnosis of inborn errors of metabolism

The current diagnostic work-up of inborn errors of metabolism (IEM) is rapidly moving toward integrative analytical approaches. We aimed to develop an innovative, targeted urine metabolomics (TUM) screening procedure to accelerate the diagnosis of patients with IEM. Urinary samples, spiked with three stable isotope-labeled internal standards, were analyzed for 258 diagnostic metabolites with an ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UHPLC-QTOF-MS) configuration run in positive and negative ESI modes. The software automatically annotated peaks, corrected for peak overloading, and reported peak quality and shifting. Robustness and reproducibility were satisfactory for most metabolites. Z-scores were calculated against four age-group-matched control cohorts. Disease phenotypes were scored based on database metabolite matching. Graphical reports comprised a needle plot, annotating abnormal metabolites, and a heatmap showing the prioritized disease phenotypes. In the clinical validation, we analyzed samples of 289 patients covering 78 OMIM phenotypes from 12 of the 15 society for the study of inborn errors of metabolism (SSIEM) disease groups. The disease groups include disorders in the metabolism of amino acids, fatty acids, ketones, purines and pyrimidines, carbohydrates, porphyrias, neurotransmitters, vitamins, cofactors, and creatine. The reporting tool easily and correctly diagnosed most samples. Even subtle aberrant metabolite patterns as seen in mild multiple acyl-CoA dehydrogenase deficiency (GAII) and maple syrup urine disease (MSUD) were correctly called without difficulty. Others, like creatine transporter deficiency, are illustrative of IEM that remain difficult to diagnose. We present TUM as a powerful diagnostic screening tool that merges most urinary diagnostic assays expediting the diagnostics for patients suspected of an IEM.

A Siamese neural network model for the prioritization of metabolic disorders by integrating real and simulated data

MOTIVATION

Untargeted metabolomic approaches hold a great promise as a diagnostic tool for inborn errors of metabolisms (IEMs) in the near future. However, the complexity of the involved data makes its application difficult and time consuming. Computational approaches, such as metabolic network simulations and machine learning, could significantly help to exploit metabolomic data to aid the diagnostic process. While the former suffers from limited predictive accuracy, the latter is normally able to generalize only to IEMs for which sufficient data are available. Here, we propose a hybrid approach that exploits the best of both worlds by building a mapping between simulated and real metabolic data through a novel method based on Siamese neural networks (SNN).

RESULTS

The proposed SNN model is able to perform disease prioritization for the metabolic profiles of IEM patients even for diseases that it was not trained to identify. To the best of our knowledge, this has not been attempted before. The developed model is able to significantly outperform a baseline model that relies on metabolic simulations only. The prioritization performances demonstrate the feasibility of the method, suggesting that the integration of metabolic models and data could significantly aid the IEM diagnosis process in the near future.

AVAILABILITY AND IMPLEMENTATION

Metabolic datasets used in this study are publicly available from the cited sources. The original data produced in this study, including the trained models and the simulated metabolic profiles, are also publicly available (Messa et al., 2020).

The tip of the iceberg for diagnostic dilemmas: Performance of current diagnostics and future complementary screening approaches

Genetic testing is currently the leading edge of clinical care when it comes to diagnostics. However, many questions remain unanswered even when employing next-generation sequencing techniques due to our inability to decode genetic variations and our limited repertoire of available diagnoses. Accordingly, diagnostic yields for current genomic screenings are <50% and fail to provide the whole picture, leaving the remaining patients without a definitive diagnosis. Human phenotypic/disease expression is explained by alterations not only at the genome, but also at the transcriptome, proteome and metabolome levels. These "higher" complexity levels represent at wealth of information, and diagnostic screenings tests at these levels have been shown to significantly improve diagnostic yields in specific populations compared to conventional diagnostic workup or gold standards in use (7-30% increase in diagnostic yields, depending on the population, approach and gold standard being compared against). However, these are not yet routinely available to clinicians. Due to their dynamic and modifiable nature, tapping into data from different omics will improve our understanding of the pathophysiological bases underlying (many yet to characterize) human disorders. We herein review how alterations at these levels (e.g. post-transcriptional and post-translational) may be pathogenic, how such tests may be implemented and in which situations they are of significant utility.

Cross-Omics: Integrating Genomics with Metabolomics in Clinical Diagnostics

Next-generation sequencing and next-generation metabolic screening are, independently, increasingly applied in clinical diagnostics of inborn errors of metabolism (IEM). Integrated into a single bioinformatic method, these two –omics technologies can potentially further improve the diagnostic yield for IEM. Here, we present cross-omics: a method that uses untargeted metabolomics results of patient’s dried blood spots (DBSs), indicated by Z-scores and mapped onto human metabolic pathways, to prioritize potentially affected genes. We demonstrate the optimization of three parameters: (1) maximum distance to the primary reaction of the affected protein, (2) an extension stringency threshold reflecting in how many reactions a metabolite can participate, to be able to extend the metabolite set associated with a certain gene, and (3) a biochemical stringency threshold reflecting paired Z-score thresholds for untargeted metabolomics results. Patients with known IEMs were included. We performed untargeted metabolomics on 168 DBSs of 97 patients with 46 different disease-causing genes, and we simulated their whole-exome sequencing results in silico. We showed that for accurate prioritization of disease-causing genes in IEM, it is essential to take into account not only the primary reaction of the affected protein but a larger network of potentially affected metabolites, multiple steps away from the primary reaction.