RRmix: A method for simultaneous batch effect correction and analysis of metabolomics data in the absence of internal standards

Sialic acids Neu5Ac and KDN in adipose tissue samples from individuals following habitual vegetarian or non-vegetarian dietary patterns

Sialic acids (Sias) are a class of sugar molecules with a parent nine-carbon neuraminic acid, generally present at the ends of carbohydrate chains, either attached to cellular surfaces or as secreted glycoconjugates. Given their position and structural diversity, Sias modulate a wide variety of biological processes. However, little is known about the role of Sias in human adipose tissue, or their implications for health and disease, particularly among individuals following different dietary patterns. The goal of this study was to measure N-Acetylneuraminic acid (Neu5Ac), N-Glycolylneuraminic acid (Neu5Gc), and 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (KDN) concentrations in adipose tissue samples from participants in the Adventist Health Study-2 (AHS-2) and to compare the abundance of these Sias in individuals following habitual, long-term vegetarian or non-vegetarian dietary patterns. A method was successfully developed for the extraction and detection of Sias in adipose tissue. Sias levels were quantified in 52 vegans, 56 lacto-vegetarians, and 48 non-vegetarians using LC-MS/MS with Neu5Ac-D-1,2,3-13C3 as an internal standard. Dietary groups were compared using linear regression. Vegans and lacto-ovo-vegetarians had significantly higher concentrations of Neu5Ac relative to non-vegetarians. While KDN levels tended to be higher in vegans and lacto-ovo-vegetarians, these differences were not statistically significant. However, KDN levels were significantly inversely associated with body mass index. In contrast, Neu5Gc was not detected in human adipose samples. It is plausible that different Neu5Ac concentrations in adipose tissues of vegetarians, compared to those of non-vegetarians, reflect a difference in the baseline inflammatory status between the two groups. Epidemiologic studies examining levels of Sias in human adipose tissue and other biospecimens will help to further explore their roles in development and progression of inflammatory conditions and chronic diseases.

Quantifying Cell Heterogeneity and Subpopulations Using Single Cell Metabolomics

Mass spectrometry (MS) has become an indispensable tool for metabolomics studies. However, due to the lack of applicable experimental platforms, suitable algorithm, software, and quantitative analyses of cell heterogeneity and subpopulations, investigating global metabolomics profiling at the single cell level remains challenging. We combined the Single-probe single cell MS (SCMS) experimental technique with a bioinformatics software package, SinCHet-MS (Single Cell Heterogeneity for Mass Spectrometry), to characterize changes of tumor heterogeneity, quantify cell subpopulations, and prioritize the metabolite biomarkers of each subpopulation. As proof of principle studies, two melanoma cancer cell lines, the primary (WM115; with a lower drug resistance) and the metastatic (WM266-4; with a higher drug resistance), were used as models. Our results indicate that after the treatment of the anticancer drug vemurafenib, a new subpopulation emerged in WM115 cells, while the proportion of the existing subpopulations was changed in the WM266-4 cells. In addition, metabolites for each subpopulation can be prioritized. Combining the SCMS experimental technique with a bioinformatics tool, our label-free approach can be applied to quantitatively study cell heterogeneity, prioritize markers for further investigation, and improve the understanding of cell metabolism in human diseases and response to therapy.

Alignment and Analysis of a Disparately Acquired Multibatch Metabolomics Study of Maternal Pregnancy Samples

Untargeted liquid chromatography-mass spectrometry metabolomics studies are typically performed under roughly identical experimental settings. Measurements acquired with different LC-MS protocols or following extended time intervals harbor significant variation in retention times and spectral abundances due to altered chromatographic, spectrometric, and other factors, raising many data analysis challenges. We developed a computational workflow for merging and harmonizing metabolomics data acquired under disparate LC-MS conditions. Plasma metabolite profiles were collected from two sets of maternal subjects three years apart using distinct instruments and LC-MS procedures. Metabolomics features were aligned using metabCombiner to generate lists of compounds detected across all experimental batches. We applied data set-specific normalization methods to remove interbatch and interexperimental variation in spectral intensities, enabling statistical analysis on the assembled data matrix. Bioinformatics analyses revealed large-scale metabolic changes in maternal plasma between the first and third trimesters of pregnancy and between maternal plasma and umbilical cord blood. We observed increases in steroid hormones and free fatty acids from the first trimester to term of gestation, along with decreases in amino acids coupled to increased levels in cord blood. This work demonstrates the viability of integrating nonidentically acquired LC-MS metabolomics data and its utility in unconventional metabolomics study designs.

Evaluating and minimizing batch effects in metabolomics

Determining metabolomic differences among samples of different phenotypes is a critical component of metabolomics research. With the rapid advances in analytical tools such as ultrahigh-resolution chromatography and mass spectrometry, an increasing number of metabolites can now be profiled with high quantification accuracy. The increased detectability and accuracy raise the level of stringiness required to reduce or control any experimental artifacts that can interfere with the measurement of phenotype-related metabolome changes. One of the artifacts is the batch effect that can be caused by multiple sources. In this review, we discuss the origins of batch effects, approaches to detect interbatch variations, and methods to correct unwanted data variability due to batch effects. We recognize that minimizing batch effects is currently an active research area, yet a very challenging task from both experimental and data processing perspectives. Thus, we try to be critical in describing the performance of a reported method with the hope of stimulating further studies for improving existing methods or developing new methods.

A hierarchical approach to removal of unwanted variation for large-scale metabolomics data

Liquid chromatography-mass spectrometry-based metabolomics studies are increasingly applied to large population cohorts, which run for several weeks or even years in data acquisition. This inevitably introduces unwanted intra- and inter-batch variations over time that can overshadow true biological signals and thus hinder potential biological discoveries. To date, normalisation approaches have struggled to mitigate the variability introduced by technical factors whilst preserving biological variance, especially for protracted acquisitions. Here, we propose a study design framework with an arrangement for embedding biological sample replicates to quantify variance within and between batches and a workflow that uses these replicates to remove unwanted variation in a hierarchical manner (hRUV). We use this design to produce a dataset of more than 1000 human plasma samples run over an extended period of time. We demonstrate significant improvement of hRUV over existing methods in preserving biological signals whilst removing unwanted variation for large scale metabolomics studies. Our tools not only provide a strategy for large scale data normalisation, but also provides guidance on the design strategy for large omics studies.

Mass spectrometry-based metabolomics is a powerful method for profiling large clinical cohorts but batch variations can obscure biologically meaningful differences. Here, the authors develop a computational workflow that removes unwanted data variation while preserving biologically relevant information.

The evolving landscape of untargeted metabolomics

AIMS

Untargeted Metabolomics is a "hypothesis-generating discovery strategy" that compares groups of samples (e.g., cases vs controls); identifies the metabolome and establishes (early signs of) perturbations. Targeted Metabolomics helped gather key information in life sciences and disclosed novel strategies for the treatment of major clinical entities (e.g., malignancy, cardiovascular diabetes mellitus, drug toxicity). Because of its relevance in biomarker discovery, attention is now devoted to improving the translational potential of untargeted Metabolomics.

DATA SYNTHESIS

Expertise in laboratory medicine and in bioinformatics helps solve challenges/pitfalls that may bias metabolite profiling in untargeted Metabolomics. Clinical validation (availability/reliability of analytical instruments) and profitability (how many people will use the test) are mandatory steps for potential biomarkers. Biomarkers to predict individual patient response, patient populations that will best respond to specific strategies and/or approaches for an optimal response to treatment are now being developed. Additional help is expected from professional, and regulatory Agencies as to guidelines for study design and data acquisition and analysis, to be applied from the very beginning of a project. Evidence from food, plant, human, environmental, and animal research argues for the need of miniaturized approaches that employ low-cost, easy to use, mobile devices. ELISA kits with such characteristics that employ targeted metabolites are already available.

CONCLUSIONS

Improving knowledge of the mechanisms behind the disease status (pathophysiology) will help untargeted Metabolomics gather a direct positive impact on welfare and industrial advancements, and fade uncertainties perceived by regulators/payers and patients concerning variables related to miniaturised instruments and user-friendly software and databases.

Dietary Fructose Alters the Composition, Localization, and Metabolism of Gut Microbiota in Association With Worsening Colitis

BACKGROUND & AIMS

The incidence of inflammatory bowel diseases has increased over the last half century, suggesting a role for dietary factors. Fructose consumption has increased in recent years. Recently, a high fructose diet (HFrD) was shown to enhance dextran sodium sulfate (DSS)-induced colitis in mice. The primary objectives of the current study were to elucidate the mechanism(s) underlying the pro-colitic effects of dietary fructose and to determine whether this effect occurs in both microbially driven and genetic models of colitis.

METHODS

Antibiotics and germ-free mice were used to determine the relevance of microbes for HFrD-induced worsening of colitis. Mucus thickness and quality were determined by histologic analyses. 16S rRNA profiling, in situ hybridization, metatranscriptomic analyses, and fecal metabolomics were used to determine microbial composition, spatial distribution, and metabolism. The significance of HFrD on pathogen and genetic-driven models of colitis was determined by using Citrobacter rodentium infection and Il10^-/- mice, respectively.

RESULTS

Reducing or eliminating bacteria attenuated HFrD-mediated worsening of DSS-induced colitis. HFrD feeding enhanced access of gut luminal microbes to the colonic mucosa by reducing thickness and altering the quality of colonic mucus. Feeding a HFrD also altered gut microbial populations and metabolism including reduced protective commensal and bile salt hydrolase-expressing microbes and increased luminal conjugated bile acids. Administration of conjugated bile acids to mice worsened DSS-induced colitis. The HFrD also worsened colitis in Il10^-/- mice and mice infected with C rodentium.

CONCLUSIONS

Excess dietary fructose consumption has a pro-colitic effect that can be explained by changes in the composition, distribution, and metabolic function of resident enteric microbiota.

Untargeted metabolomics profiling and hemoglobin normalization for archived newborn dried blood spots from a refrigerated biorepository

Archived dried blood spots (DBS) following newborn screening are an attractive resource for interrogating early-life biology using untargeted metabolomics. Therefore, they have the potential to substantially aid etiological studies, particularly for rare and low-frequency childhood diseases and disorders. However, metabolite quantification in DBS is hindered by variation sources not present in serum and plasma samples such as the hematocrit effect and unknown initial blood volumes. Hemoglobin (Hb) is an appropriate correlate for hematocrit in experimentally-generated DBS punches. However, since many biorepositories worldwide archive DBS at 4-5 °C, there is a need to validate the utility of Hb for DBS archived under refrigeration. We evaluated two simple spectroscopic methods for measuring Hb in DBS stored at 4 +/- 2 °C for up to 21 years, obtained from the newborn screening program at the Karolinska University Hospital, Sweden. Spearman correlation analysis and Akaike Information Criterion model selection found that measurement of a Hb sodium lauryl sulfate complex at 540 nm better described nuisance variation than Hb measured at 404 nm, or using age of spot alone. This is the first study to profile metabolites and to propose a normalization factor for metabolite measurements from DBS archived for decades at 4 °C.

Addressing the batch effect issue for LC/MS metabolomics data in data preprocessing

With the growth of metabolomics research, more and more studies are conducted on large numbers of samples. Due to technical limitations of the Liquid Chromatography–Mass Spectrometry (LC/MS) platform, samples often need to be processed in multiple batches. Across different batches, we often observe differences in data characteristics. In this work, we specifically focus on data generated in multiple batches on the same LC/MS machinery. Traditional preprocessing methods treat all samples as a single group. Such practice can result in errors in the alignment of peaks, which cannot be corrected by post hoc application of batch effect correction methods. In this work, we developed a new approach that address the batch effect issue in the preprocessing stage, resulting in better peak detection, alignment and quantification. It can be combined with down-stream batch effect correction methods to further correct for between-batch intensity differences. The method is implemented in the existing workflow of the apLCMS platform. Analyzing data with multiple batches, both generated from standardized quality control (QC) plasma samples and from real biological studies, the new method resulted in feature tables with better consistency, as well as better down-stream analysis results. The method can be a useful addition to the tools available for large studies involving multiple batches. The method is available as part of the apLCMS package. Download link and instructions are at https://mypage.cuhk.edu.cn/academics/yutianwei/apLCMS/.

ESTIMATION AND INFERENCE IN METABOLOMICS WITH NON-RANDOM MISSING DATA AND LATENT FACTORS

High throughput metabolomics data are fraught with both non-ignorable missing observations and unobserved factors that influence a metabolite's measured concentration, and it is well known that ignoring either of these complications can compromise estimators. However, current methods to analyze these data can only account for the missing data or unobserved factors, but not both. We therefore developed MetabMiss, a statistically rigorous method to account for both non-random missing data and latent factors in high throughput metabolomics data. Our methodology does not require the practitioner specify a likelihood for the missing data, and makes investigating the relationship between the metabolome and tens, or even hundreds, of phenotypes computationally tractable. We demonstrate the fidelity of Metab-Miss's estimates using both simulated and real metabolomics data, and prove their asymptotic correctness when the sample size and number of metabolites grows to infinity.

Standardization procedures for real-time breath analysis by secondary electrospray ionization high-resolution mass spectrometry

Despite the attractiveness of breath analysis as a non-invasive means to retrieve relevant metabolic information, its introduction into routine clinical practice remains a challenge. Among all the different analytical techniques available to interrogate exhaled breath, secondary electrospray ionization high-resolution mass spectrometry (SESI-HRMS) offers a number of advantages (e.g., real-time, yet wide, metabolome coverage) that makes it ideal for untargeted and targeted studies. However, so far, SESI-HRMS has relied mostly on lab-built prototypes, making it difficult to standardize breath sampling and subsequent analysis, hence preventing further developments such as multi-center clinical studies. To address this issue, we present here a number of new developments. In particular, we have characterized a new SESI interface featuring real-time readout of critical exhalation parameters such as CO₂, exhalation flow rate, and exhaled volume. Four healthy subjects provided breath specimens over a period of 1 month to characterize the stability of the SESI-HRMS system. A first assessment of the repeatability of the system using a gas standard revealed a coefficient of variation (CV) of 2.9%. Three classes of aldehydes, namely 4-hydroxy-2-alkenals, 2-alkenals and 4-hydroxy-2,6-alkedienals―hypothesized to be markers of oxidative stress―were chosen as representative metabolites of interest to evaluate the repeatability and reproducibility of this breath analysis analytical platform. Median and interquartile ranges (IQRs) of CVs for CO₂, exhalation flow rate, and exhaled volume were 3.2% (1.5%), 3.1% (1.9%), and 5.0% (4.6%), respectively. Despite the high repeatability observed for these parameters, we observed a systematic decay in the signal during repeated measurements for the shorter fatty aldehydes, which eventually reached a steady state after three/four repeated exhalations. In contrast, longer fatty aldehydes showed a steady behavior, independent of the number of repeated exhalation maneuvers. We hypothesize that this highly molecule-specific and individual-independent behavior may be explained by the fact that shorter aldehydes (with higher estimated blood-to-air partition coefficients; approaching 100) mainly get exchanged in the airways of the respiratory system, whereas the longer aldehydes (with smaller estimated blood-to-air partition coefficients; approaching 10) are thought to exchange mostly in the alveoli. Exclusion of the first three exhalations from the analysis led to a median CV (IQR) of 6.7 % (5.5 %) for the said classes of aldehydes. We found that such intra-subject variability is in general much lower than inter-subject variability (median relative differences between subjects 48.2%), suggesting that the system is suitable to capture such differences. No batch effect due to sampling date was observed, overall suggesting that the intra-subject variability measured for these series of aldehydes was biological rather than technical. High correlations found among the series of aldehydes support this notion. Finally, recommendations for breath sampling and analysis for SESI-HRMS users are provided with the aim of harmonizing procedures and improving future inter-laboratory comparisons.

Tools and resources for metabolomics research community: A 2017-2018 update

The scale at which MS- and NMR-based platforms generate metabolomics datasets for both research, core, and clinical facilities to address challenges in the various sciences-ranging from biomedical to agricultural-is underappreciated. Thus, metabolomics efforts spanning microbe, environment, plant, animal, and human systems have led to continual and concomitant growth of in silico resources for analysis and interpretation of these datasets. These software tools, resources, and databases drive the field forward to help keep pace with the amount of data being generated and the sophisticated and diverse analytical platforms that are being used to generate these metabolomics datasets. To address challenges in data preprocessing, metabolite annotation, statistical interrogation, visualization, interpretation, and integration, the metabolomics and informatics research community comes up with hundreds of tools every year. The purpose of the present review is to provide a brief and useful summary of more than 95 metabolomics tools, software, and databases that were either developed or significantly improved during 2017-2018. We hope to see this review help readers, developers, and researchers to obtain informed access to these thorough lists of resources for further improvisation, implementation, and application in due course of time.

Promises and pitfalls of untargeted metabolomics

Metabolomics is one of the newer omics fields, and has enabled researchers to complement genomic and protein level analysis of disease with both semi-quantitative and quantitative metabolite levels, which are the chemical mediators that constitute a given phenotype. Over more than a decade, methodologies have advanced for both targeted (quantification of specific analytes) as well as untargeted metabolomics (biomarker discovery and global metabolite profiling). Untargeted metabolomics is especially useful when there is no a priori metabolic hypothesis. Liquid chromatography coupled to mass spectrometry (LC-MS) has been the preferred choice for untargeted metabolomics, given the versatility in metabolite coverage and sensitivity of these instruments. Resolving and profiling many hundreds to thousands of metabolites with varying chemical properties in a biological sample presents unique challenges, or pitfalls. In this review, we address the various obstacles and corrective measures available in four major aspects associated with an untargeted metabolomics experiment: (1) experimental design, (2) pre-analytical (sample collection and preparation), (3) analytical (chromatography and detection), and (4) post-analytical (data processing).

Mass Spectrometric Identification of Urinary Biomarkers of Pulmonary Tuberculosis

Background

Tuberculosis (TB) is the leading infectious cause of death worldwide. A major barrier to control of the pandemic is a lack of clinical biomarkers with the ability to distinguish active TB from healthy and sick controls and potential for development into point-of-care diagnostics.

Methods

We conducted a prospective case control study to identify candidate urine-based diagnostic biomarkers of active pulmonary TB (discovery cohort) and obtained a separate blinded “validation” cohort of confirmed cases of active pulmonary TB and controls with non-tuberculous pulmonary disease for validation. Clean-catch urine samples were collected and analyzed using high performance liquid chromatography-coupled time-of-flight mass spectrometry.

Results

We discovered ten molecules from the discovery cohort with receiver-operator characteristic (ROC) area-under-the-curve (AUC) values >85%. These 10 molecules also significantly decreased after 60 days of treatment in a subset of 20 participants followed over time. Of these, a specific combination of diacetylspermine, neopterin, sialic acid, and N-acetylhexosamine exhibited ROC AUCs >80% in a blinded validation cohort of participants with active TB and non-tuberculous pulmonary disease.

Conclusion

Urinary levels of diacetylspermine, neopterin, sialic acid, and N-acetylhexosamine distinguished patients with tuberculosis from healthy controls and patients with non-tuberculous pulmonary diseases, providing a potential noninvasive biosignature of active TB.

Funding

This study was funded by Weill Cornell Medicine, the National Institute of Allergy and Infectious Diseases, the Clinical and Translational Science Center at Weill Cornell, the NIH Fogarty International Center grants, and the NIH Tuberculosis Research Unit (Tri-I TBRU).

•

Urinary levels of small metabolites appear capable of distinguishing cases of active pulmonary tuberculosis from sick and healthy controls.

•

Levels of these biomarkers decrease after 60 days of treatment in a longitudinal cohort of 20 participants.

•

- Many of the identified biomarkers are known inflammatory intermediates that may reflect a specific immune response to tuberculosis.

Urine tests are commonly used to enable non-invasive, rapid and point-of-care diagnosis of various infectious diseases. We identified diacetylspermine, neopterin, sialic acid and N-acetylhexosamine as potential urine-based biomarkers for tuberculosis from two independent patient cohorts. These metabolites are known inflammatory intermediates and appear to decrease with anti-tuberculosis therapy in a subset of participants followed over 2 months. If validated, these metabolites have potential to both improve our understanding of the immune reaction to active tuberculosis and facilitate the development of a much-needed clinical biomarker for tuberculosis.