ACS Publications. Most Trusted. Most Cited. Most Read
Quick View

OR SEARCH CITATIONS

My Activity
Publications
CONTENT TYPES

Figure 1Loading Img
Download Hi-Res ImageDownload to MS-PowerPointCite This:J. Proteome Res. 2017, 16, 10, 3623-3633
RETURN TO ISSUEPREVArticleNEXT

Improving Visualization and Interpretation of Metabolome-Wide Association Studies: An Application in a Population-Based Cohort Using Untargeted 1H NMR Metabolic Profiling

View Author Information
† ‡ †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom
§ Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.
⊥ Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom
Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece
# Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States
∥ Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States
□ Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom
▼ Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
*E-mail: [email protected]
Cite this: J. Proteome Res. 2017, 16, 10, 3623–3633
Publication Date (Web):August 20, 2017
https://doi.org/10.1021/acs.jproteome.7b00344

Copyright © 2017 American Chemical Society. This publication is licensed under CC-BY.

  • Open Access

Article Views

1659

Altmetric

-

Citations

18
LEARN ABOUT THESE METRICS
PDF (2 MB)
Get e-Alerts
Supporting Info (1)»
SUBJECTS:

Abstract

High Resolution Image
Download MS PowerPoint Slide

1H NMR spectroscopy of biofluids generates reproducible data allowing detection and quantification of small molecules in large population cohorts. Statistical models to analyze such data are now well-established, and the use of univariate metabolome wide association studies (MWAS) investigating the spectral features separately has emerged as a computationally efficient and interpretable alternative to multivariate models. The MWAS rely on the accurate estimation of a metabolome wide significance level (MWSL) to be applied to control the family wise error rate. Subsequent interpretation requires efficient visualization and formal feature annotation, which, in-turn, call for efficient prioritization of spectral variables of interest. Using human serum 1H NMR spectroscopic profiles from 3948 participants from the Multi-Ethnic Study of Atherosclerosis (MESA), we have performed a series of MWAS for serum levels of glucose. We first propose an extension of the conventional MWSL that yields stable estimates of the MWSL across the different model parameterizations and distributional features of the outcome. We propose both efficient visualization methods and a strategy based on subsampling and internal validation to prioritize the associations. Our work proposes and illustrates practical and scalable solutions to facilitate the implementation of the MWAS approach and improve interpretation in large cohort studies.

KEYWORDS:

Introduction

ARTICLE SECTIONS
Jump To
Over the past 15 years, improvements in high-throughput technologies have accelerated the simultaneous measurement of large numbers of metabolites in a single sample using NMR and mass spectrometry (MS), which are both widely used to characterize a biofluid using either untargeted or targeted profiling approaches. (1, 2) Both technologies have emerged as efficient tools for identifying biomarkers of exposures as well as early disease manifestations, hence informing on molecular mechanisms involved in pathogenesis (e.g., in cancer, diabetes, cardiovascular, and neurological diseases). (3-6) Metabolic phenotyping uses robust and reliable analytical methods (7, 8) that are ideally suited for untargeted profiling since prior knowledge about compounds present in a sample is not required. Such an agnostic discovery approach can inform complementary targeted methods by identifying novel chemical compounds that may contribute to the molecular pathways involved in complex phenotypes. (8-10) These screening exercises call upon the efficient analyses of high-dimensional data whose exploration poses complex methodological and interpretation problems. (11) Typically, untargeted NMR experiments in a large population study generate tens of thousands of data points for thousands of individuals. By adopting univariate approaches, the concept of the metabolome wide association study (MWAS) (12) has been proposed to analyze these data, and various statistical approaches to perform such analyses were established and have been explored and reviewed. (13-17) The complex correlation structures existing in metabolic profiles result in partially redundant signals across the metabolic spectra. These need to be accounted for while correcting for multiple testing, for instance by using a permutation-based procedure to derive the metabolome-wide significance level (MWSL) controlling the family wise error rate (FWER). (13) However, the comparative applicability of these approaches, as well as the visualization of the results they produce, has not been comprehensively explored. Here we provide, using a real-life example from the COMBI-BIO project, an extension of the MWAS methodology we initially developed using simulated data sets in a class discrimination context (13) and further extended to accommodate continuous outcomes. The current extension includes the computation of a stable and reproducible statistical significance threshold and proposes ways to estimate consistently across different types of NMR spectra as well as an internal validation procedure to assess the robustness of candidate associations. Our data set comprises two versions of 1H NMR nontargeted metabolic profiles in 3948 participants from the Multiethnic Study of Atherosclerosis (MESA) cohort. (18) As a proof-of-principle example, we focus on identifying the NMR spectral features associated with fasting blood serum glucose, which was measured by an independent technique (glucose oxidase method). By examining the full NMR spectrum, our analysis was designed to identify spectral regions corresponding to other metabolites beyond glucose itself that also vary with fasting serum glucose (1H NMR glucose associated peaks).

Materials and Methods

ARTICLE SECTIONS
Jump To

Study Population and Sample Selection

The MESA cohort has been described elsewhere (18) and includes 6814 participants (53% females, 47% males) aged 44–84 years (mean = 62 years) from four different ethnic groups: Chinese-American (n = 803), African-American (n = 1893), Hispanic (n = 1496), and Caucasian (n = 2622), all recruited between 2000 and 2002 at clinical centers in the United States. Participants were free of symptomatic cardiovascular disease at baseline, and demographic, medical history, anthropometric, lifestyle data, and serum samples were collected during the first examination (July 2000–August 2002), together with information on lipid or blood pressure treatment, and diabetes, and measures of systolic blood pressure. Serum samples were stored at −80 °C after collection. At enrolment, high density lipoprotein (HDL-C) was measured in EDTA plasma on the Roche/Hitachi 911 Automatic Analyzer (Roche Diagnostics Corporation, Indianapolis, IN), and low density lipoprotein (LDL-C) was calculated using the Friedewald equation, (19) together with fasting serum glucose using the glucose oxidase method on the Vitros analyzer (Johnson and Johnson Clinical Diagnostics). Ethical approval was obtained by local ethical review boards, and subsequent analysis was conducted in full accordance with the ethical approval obtained.

Samples Preparation and 1H NMR Spectroscopic Acquisition

The full sample preparation and quality control procedure have been extensively described elsewhere. (20) Briefly, serum samples were thawed on the day prior to analysis, and 300 μL of each sample was mixed with 300 μL of phosphate buffer (NaHPO4, 0.075M, pH = 7.4). Samples were processed in two phases (each corresponding to a separate analytical batch). Eppendorfs were used for phase 1, while 96-well plates were used for phase 2. After centrifugation (12 000g at 4 °C for 5 min), 550 μL of each sample-buffer mixture was manually transferred into SampleJet 5 mm diameter NMR tubes and kept at 4 °C until analysis. Different types of quality control (QC) samples were used for each phase as described elsewhere. (20) All QC pools were aliquoted in 350 μL and stored at −80 °C prior to analysis. 1H NMR spectra were acquired using a Bruker DRX600 spectrometer (Bruker Biospin, Rheinstetten, Germany) operating at 600 MHz. A standard water suppressed one-dimensional spectrum (usually termed NOESY) and a Carr–Purcell–Meiboom–Gill (CPMG) spectrum were obtained for each sample. (20)

1H NMR Metabolite Profiling

The metabolite profiling workflow has been detailed elsewhere. (20) For each biosample, two NMR profiles were generated: (i) a 1-D spectrum (NOESY) showing resonances from all proton-containing molecules in the sample, including broad, largely undefined bands from serum proteins, sharper and well-defined bands from serum lipoproteins (with some classification into their main groups), and sharp peaks from a range of small molecule metabolites such as amino acids, simple carbohydrates, organic acids, organic bases, and a number of osmolytes, and (ii) a Carr–Purcell–Meiboom–Gill (CPMG) spectrum that attenuates the peaks from the macromolecules and allows better definition of the small molecules.
For both CPMG and NOESY NMR data, in-house written MATLAB (Mathworks Inc., USA) routines were utilized for phasing and baseline correction. Prior to spectral peak alignment, the region δ 4.400–5.100 corresponding to the H2O resonance was removed. Spectral peak alignment was performed by the Recursive Segment-wise Peak Alignment (RSPA) algorithm. (21) Regions where the peaks of different suspected contaminations (i.e., methanol) occurred were removed from the whole spectra (δ 1.180–1.240, δ 2.244–2.261 and δ 3.660–3.710). The remaining spectral regions were normalized by probabilistic quotient normalization using the median spectrum as the reference. (22) The normalized high resolution spectra contained 30 590 data points (variables) for both CPMG and NOESY data sets. To ensure comparability across batches (i.e., across studies and phases), each variable was mean-centered (23) (we will hereafter refer to this as mean corrected intensities). Score plots of the first few principal components for the resulting data set were finally used to identify and remove potential outlier samples (N = 3 for the CPMG and N = 4 for the NOESY data) from further analyses. The MESA study population was further examined for potential outliers using score plots from the first two principal components (Figure S-1), which showed strong homogeneity in the study participants.

Metabolite Assignments

Selected samples were analyzed using a range of 2-D NMR experiments such as total correlation spectroscopy (TOCSY) or heteronuclear single quantum coherence (HSQC) to aid molecular identification. We used approaches such as statistical total correlation spectroscopy (STOCSY) and STORM to help constrain possible molecular structures. (24, 25) Peak identification in the 1H NMR data was supported by a semiautomatic clustering of the full resolution 1H NMR spectra (30 590 data points) using statistical recoupling of variables, (26) where the algorithm defines a cluster as containing 10 or more variables. Each cluster is subsequently checked by NMR experts to improve the data point grouping and identify peak overlaps. From our data, 136 and 159 clusters were identified in 1H NMR NOESY and CPMG data, respectively, each of them corresponding to a single or a group of peaks. Resulting spectral information was also compared to the literature (27, 28) and to existing databases such as the Human Metabolome Database (29) (HMDB, http://www.hmdb.ca/) and the Biological Magnetic Resonance Data Bank (30) (BMRB, http://www.bmrb.wisc.edu). Resulting sets of NMR features were ultimately confirmed through spiking experiments using commercial standards. The level of assignment (LoA) we used was adapted from Sumner et al. (31) NMR features were also characterized by their peak multiplicity: singlet (s), doublet (d), triplet (t), quartet (q), doublet of doublet (dd), multiplet (m), broad peak (b), and noise (n). Ninety-two clusters were assigned to 44 unique metabolites for the NOESY data set, and 91 clusters were assigned to 48 unique metabolites for the CPMG data set.

Metabolome-Wide Association Study (MWAS)

Figure 1 presents our analytical workflow. We adopted a MWAS approach using univariate linear regression models to systematically screen the 30 590 data points assayed in both NOESY and CPMG profiles.

Figure 1

Figure 1. Analytical workflow.

High Resolution Image
Download MS PowerPoint Slide
For a given data point, the linear model can be formulated aswhere Xi represents the normalized NMR intensity, Yi is the outcome variable; here the log10 transformed blood glucose concentration and εi is the error term for an observation i. α is the intercept of the model, and ß1 measures how strongly each data point influences the outcome variable. FEi is a vector of fixed effect observations for individual i and the vector B2 compiles the regression coefficients for each adjustment covariate; for model 1, age, gender, phase, and ethnicity, and for model 2, we additionally correct for LDL and HDL cholesterol, lipids and blood pressure treatment, systolic blood pressure, smoking status, and diabetes.
Most of the recently published MWAs adopted a Bonferroni correction for multiple testing, (32-34) which ignores correlation across variables and therefore does not account for the (partial) redundancy across the statistical test performed. This may lead to an overly conservative multiple testing correction. To take into account the high degree of correlation in spectral data and prevent overcorrection for multiple testing, one computationally efficient method relies on the estimation of the effective number of tests (ENT), as defined by the virtual number of independent tests that are performed across the actual p tests performed. ENT measures the level of correlation within the spectral data and can be estimated through spectral decomposition of the correlation matrix of predictors. (35) From this estimate, the per-test significance level ensuring a FWER control can be defined as the Bonferroni-corrected threshold corresponding to that number of independent tests. However, this approach remains of limited use in real-life metabolomic data sets, notably because, for numerical reasons, the ENT is upper-bounded by the number of observations. (36)
As a scalable alternative, we used a permutation-based method to estimate the metabolome wide significance level (MWSL or α′) (13) in which the outcome (glucose levels) is randomly shuffled across observations. Each permutation mimics the null hypothesis of no association, and we performed for each permuted data set a MWAS using the linear regression model described above. In that setting, and for a given permuted data set, the minimum p-value across all variables (denoted q) represents the largest significance level to be considered to ensure no false positive findings, and the MWSL α′ controlling the FWER at a level α can be derived from the distribution of q across the N (set here to 10 000) permutations.
The effective number of tests (ENT) is then defined as the number of independent tests that would be required to obtain the same significance level using a Bonferroni correction: ENT= α/α′ and measures the level of correlation across the p tests performed.
To assess the robustness of the ENT estimates and to circumvent the strong assumptions from the generalized linear model on data structure (independence of each data points, distribution of the residuals, variance structure, and linear relationship between response and predictors), we ran our permutation-based procedure for both data sets (NOESY and CPMG) for both models (1 and 2) and used different transformations of the glucose distribution: raw concentrations, truncated concentrations (129 outlying observations with more than 2 standard deviations away from the mean glucose level were discarded), and log10-transformed glucose levels (Figure S-2).
To formally assess the sensitivity of our MWSL estimates to distributional features of the outcome, we also simulated continuous responses from gamma and several Gaussian distributions and compared resulting MWSL estimates to those obtained using measured glucose levels.

Sensitivity, Stability Analyses, and Results Prioritization

We performed further sensitivity analyses to assess the stability of the candidate associations we identified. These included a cross-validation procedure based on the independent subsampling (N = 100 times) of discovery (containing 80% of the observations) and replication (comprising the remaining 20% observation) data sets.
For each 80:20 discovery and replication split, we performed a MWAS and used the MWSL to identify the candidate associations in the discovery set. For each discovery set (80% of the observations), the number of independent signals among the candidate associations was approximated by the number of principal components needed to explain more than 99% of their variance. That number of PCs was then used to compute the Bonferroni corrected MWSL used in the replication set. (37)
Our strategy could be summarized as follows for a given split:
(1)

MWAS: identify (N0) candidate associations in the discovery set with discovery p-value < MWSL.

(2)

Estimate the number of independent signals these N0 correspond to run a PCA on the XN0, the matrix combining the N0 data points declared significant in the discovery set (a random 80% subsample or the full study population), and identify NPC, the number of PC’s needed to explain more than 99% of the variance in XN0.

(3)

Replication: identify from the candidate signals (step 1) those replicating in the 20% replication set at a Bonferroni-corrected significance level accounting for NPC tests, α/ NPC, setting α = 5%.

Steps 1–3 are repeated across the 100 independent splits and the proportion a given signal is identified in the discovery set, and replicated in the validation set is reported.
Statistical analyses were all performed using R v3.1.2. (38)

Results and Discussion

ARTICLE SECTIONS
Jump To
A total of 3948 individuals from the MESA cohort were included in the analysis, and for each individual, 30 590 serum NMR features were measured for both NOESY and CPMG spectra. The characteristics of the study population are summarized in Table 1.
Table 1. Summary Characteristics of the Study Population: Multiethnic Study of Atherosclerosis Cohort
  N% or mean (sd)
gendermen195149.4
women199750.6
age (y)all394862.9 (10.3)
phase1197650
2197250
ethnicityCaucasian152138.5
Hispanic92623.4
African-American96824.5
Chinese-American53313.5
body mass index (kg/m2)all394828.2 (5.4)
glucose (mg/dL)all394598.3 (31.1)
missing3 
LDL cholesterol (mg/dL)all3884117.3 (31.4)
missing64 
HDL cholesterol (mg/dL)all394250.7 (14.7)
missing6 
systolic blood pressure (mmHg)all3948127.1 (21.3)
height (cm)all3948166.4 (10.2)
diabetesno338785.8
yes56114.2
lipids treatmentno328683.2
yes66216.8
blood pressure treatmentno244962.1
yes149737.9
smokingnever198850.4
former48312.2
current146137
missing160.4
We first explored the sensitivity of the MWSL estimates to (i) the parametrization of the statistical model, (ii) the type of NMR data under investigation, and (iii) distributional features of the outcome of interest. As summarized in Table 2, we then ran the permutation procedure for two different models (Models 1 and 2), for both types of NMR data sets (NOESY and CPMG) and 3 versions of glucose blood concentrations: raw, log10 transformed and truncated (removing 129 outlying observations which were outside the 2 standard deviation range from the mean glucose level). Across all models investigated, MWSL estimates for CPMG are more stringent than for NOESY spectra and, correspondingly, ENT estimates for CPMG are much greater than those for NOESY data (ranging from 17 610 to 122 460 and from 3680 to 17 010 for CPMG and NOESY, respectively). This suggests stronger correlations within NOESY data, which is plausible given that NOESY NMR spectra contain stronger broad peaks from proteins and lipoproteins than CPMG spectra, such that data point intensities are highly correlated. As summarized in Table 2, irrespective of the type of spectrum, MWSL estimates (and corresponding ENT) seem to be only marginally affected by the number of confounders included in the model (i.e., comparing models 1 and 2). However, a systematic but moderate increase in the ENT is observed for the fully adjusted model (Model 2). This can be explained by the fact that the fully adjusted model removes spectral signals relating to the components of the Framingham risk scores (such as the conventionally measured lipoprotein levels), and these confounders were highly likely to be driving correlations across some data points.
Table 2. Significance Threshold α′ and Effective Number of Test (ENT) Based on a Bonferroni Correctionab
    NOESY (30 590 variables)CPMG (30 590 variables)
phenotypeNmodel MWSL (FWER = 5%)MWSL (FWER = 5%)
glucose39451α′ (×10–5)0.31 (0.26; 0.35) 0.04 (0.03; 0.05) 
ENT (×103)16.33 (14.42; 19.51)53.39122.46 (108.67; 143.67)400.32
38662α′ (×10–5)0.29 (0.25; 0.32) 0.04 (0.04; 0.05) 
ENT (×103)17.01 (15.57; 19.88)55.60121.64 (107.21; 142.23)397.64
log10(glucose)39451α′ (×10–5)1.09 (0.96; 1.17) 0.22 (0.20; 0.23) 
ENT (×103)4.59 (4.28; 5.22)15.0123.1 (21.7; 24.4)75.37
38662α′ (×10–5)1.05 (0.96; 1.13) 0.21 (0.19; 0.22) 
ENT (×103)4.75 (4.43; 5.02)15.5223.6 (22.7; 26.2)77.07
glucose without outliers38161α′ (×10–5)1.36 (1.25; 1.45) 0.28 (0.26; 0.29) 
ENT (×103)3.68 (3.45; 4.01)12.0218.02 (17.36; 19.07)58.92
37432α′ (×10–5)1.06 (1.00; 1.10) 0.28 (0.27; 0.30) 
ENT (×103)4.70 (4.53; 4.99)15.3617.61 (16.74; 18.74)57.57
a

95% confidence intervals are given in parentheses. Figures are based on 10 000 permutations for each model (1 and 2) and given for the glucose, log10(glucose) and the glucose after outliers exclusion (N = 129 excluded).

b

Bold figures are the ratios of effective/actual number of tests.

For CPMG data, estimates of the MWSL using raw glucose concentrations led to an estimated ENT greater than the actual number of tests (ratio around 4.00). This might be attributed to the parametric assumption in generalized linear regression model (e.g., normality of the outcome and equal variance) underlying our permutation procedure being violated. This is supported by the distribution of the blood levels of glucose, which is right skewed, with several outlying observations (Figure S-2A).
To account for this asymmetrical distribution, we first applied a log10 transformation to glucose levels (Figure S-2B). Although the transformed distribution remains right skewed, corresponding estimates of the MWSL were less stringent, and the effective to actual number of tests ratio dropped to around 0.75 for CPMG (and to 0.15 for NOESY). To remove the influence of outlying observations, we truncated the glucose distribution and removed observations more than 2 standard deviations away from the mean glucose level (Figure S-2C). While only a small number of observations were discarded (N = 129), this removal strongly impacted the MWSL estimates for CPMG data, and the effective to actual number of test ratio dropped to less than 60% for both models.
These results suggest that MWSL estimates are sensitive to the shape of the distribution of the continuous outcome under investigation, and are specifically affected by both the relative weight of its tail, and by the presence of outlying observations. To formally assess the sensitivity of our MWSL estimates to the parametric form of the response variable, we ran a series of sensitivity analyses where we randomly sampled for each participant (and for each permutation) the glucose levels from (i) a Gamma distribution fitted on the measured glucose levels (shape = 15.90, scale = 6.18), and (ii) several Gaussian distributions (mean = 0, sd = 1; mean = 0, sd = 10; mean = 0, sd = 100, mean = mean(glucose), sd = sd(glucose). By construction, the Gamma-distributed response did not include outlying observations, but featured an inflated right tail, which provided less stringent MWSL for both NOESY. Results from the normally distributed outcomes showed consistent MWSL estimates and did not seem to be strongly affected by the parameters choices defining the Gaussian distribution (Table S-1, Figure S-3). Since the numbers of associated variables were only marginally affected by the way the ENT was computed in our example, we took forward the MWSL estimated from the Gaussian simulated outcome (mean = 0, sd = 1), which also seemed to provide the best balance between generalizability and simplicity (Figures S-3 and S-4). We further compared the proportion of associated variables from different multiple testing correction strategies including Bonferroni and FWER control, and Benjamini-Hochberg (39) false discovery rate (BH-FDR) procedure (Figure S-5). In all scenario considered, the largest proportion of associated variables was always observed when using a BH-FDR procedure, and the smallest proportion of associated variables was always observed when using a Bonferroni procedure (Figure S-5).

Metabolome Wide Association Study of Glucose: Visualization and Prioritization

We performed the MWAS of blood levels of glucose on both CPMG and NOESY spectra using both models and setting the MWSL to that estimated using the Gaussian simulated outcome (mean = 0, sd = 1, Table S-1). As expected, irrespective of the model and of the type of spectrum, a very large number of spectral features were found associated with blood concentration of glucose (Table 3): for NOESY data, 72% and 44% of the spectral variables were found significantly associated with glucose level for models 1 and 2, respectively. These proportions were 40% and 16% for models 1 and 2, respectively, in CPMG data.
Table 3. Number and Percentage of Associated Variables for Models 1 and 2 for Both NOESY and CPMG by Class of Metabolitea
  NOESYCPMG
model number of data points per group(%)number of significant data points (%)number of significant data points after results prioritization (%)number of data points per group (%)number of significant data points (%)number of significant data points after results prioritization (%)
1allb30 590 (100)22 066 (72)18 340 (83)30 590 (100)12 303 (40)9920 (81)
amino-acidsc4786 (15.65)3653 (76.3)3222 (88.2)4524 (14.79)2780 (61.5)2078 (74.7)
carbohydratesc2394 (7.83)2204 (92.1)2192 (99.5)1826 (5.97)1763 (96.5)1734 (98.4)
drug derivativesc191 (0.62)191 (100)116 (60.7)81 (0.26)30 (37)26 (86.7)
lipidsc4085 (13.35)2815 (68.9)2569 (91.3)3906 (12.77)2772 (71)2530 (91.3)
nucleosidesc116 (0.38)41 (35.3)35 (85.4)103 (0.34)13 (12.6)1 (7.7)
organic acidsc1383 (4.52)949 (68.6)838 (88.3)864 (2.82)390 (45.1)266 (68.2)
othersc377 (1.23)229 (60.7)173 (75.5)363 (1.19)261 (71.9)237 (90.8)
proteinsc549 (1.79)542 (98.7)500 (92.3)499 (1.63)499 (100)494 (99)
unassigned15 978 (52.23)11 081 (69.4)8472 (76.5)10 315 (33.72)3414 (33.1)2252 (66)
2allb30 590 (100)13 449 (44)9610 (71)30 590 (100)4909 (16)3355 (68)
amino-acidsc4786 (15.65)3123 (65.3)2168 (69.4)4524 (14.79)1048 (23.2)602 (57.4)
carbohydratesc2394 (7.83)2079 (86.8)1906 (91.7)1826 (5.97)1695 (92.8)1653 (97.5)
drug derivativesc191 (0.62)65 (34)36 (55.4)81 (0.26)24 (29.6)17 (70.8)
lipidsc4085 (13.35)1172 (28.7)176 (15)3906 (12.77)263 (6.7)116 (44.1)
nucleosidesc116 (0.38)7 (6)0 (0)103 (0.34)0(−)0(−)
organic acidsc1383 (4.52)798 (57.7)462 (57.9)864 (2.82)114 (13.2)18 (15.8)
othersc377 (1.23)112 (29.7)58 (51.8)363 (1.19)87 (24)35 (40.2)
proteinsc549 (1.79)529 (96.4)400 (75.6)499 (1.63)407 (81.6)352 (86.5)
unassigned15 978 (52.23)5421 (33.9)4317 (79.6)10 315 (33.72)1067 (10.3)413 (38.7)
a

Results are also given after results prioritization: variables identified in the discovery set and replicated in the validation set in at least one split across the 100 splits (see Methods).

b

Figures are given for the whole spectra (N = 30 590 variables) including the unassigned regions.

c

Figures are based on the achieved NMR assignment: not all variables have been assigned in the spectra.

Analyses by classes of metabolite for model 1 revealed that nearly all variables assigned to drug derivative (100%), proteins (98.7%), and carbohydrates (92.1%) were associated with glucose in NOESY. Similarly, all variables assigned to proteins (100%), carbohydrates (96.5%), and others (71.9%, which include choline and glycerol) were associated with glucose in CPMG. The proportion of unassigned associated variables was higher for NOESY (69.4%) compared to CPMG (33.9%).
Adjusting for Framingham risk score (FRS) variables in model 2 reduced the total number of associations for both NOESY and CPMG spectra. This reduction was mostly observed for the lipids class (NOESY, 68.9% vs 28.7%; CPMG, 71% vs 6.7%) and the “others” class (NOESY, 60.7% vs 29.7%; CPMG, 71.9% vs 24%). As expected, the proportion of carbohydrate associated variables was one of the least affected classes by the FRS adjustment (NOESY, 92.1% vs 86.8%; CPMG, 96.5% vs 92.8%).
The utility of the MWAS approach in metabolic phenotyping relies on explicit visualization of the results. One primary output to help identify relevant spectral regions borrows from the field of genome-wide association studies and reports for each spectral variable the–log10 p-value multiplied by the sign of the corresponding regression coefficient. The resulting signed Manhattan plots (Figure 2 and Figure S-6 for CPMG and NOESY, respectively) offer a global view of the spectral regions associated with the outcome of interest, and can be further informed by the annotation of spectral features. Assigned metabolites found associated with conventional serum glucose measurements are reported in Tables S-2 and S-3 for CPMG and NOESY, respectively. For CPMG, 47 unique metabolites were associated with glucose for model 1, of which 34 were still associated after controlling for the FRS variable (44 and 41 for NOESY).

Figure 2

Figure 2. Metabolome wide study of glucose (model 2). This Manhattan plot shows the analysis of the 30 590 CPMG features. The signed negative log10 p-value is plotted against the chemical shift in ppm. To ease the visualization, all log p-value ≤ 10–30 were set to 1 × 10–30. The horizontal dashed line indicates the α′ per-test significance level controlling the FWER at a 5% level using the Gaussian simulated outcome. Data points are colored by class of metabolites. Components were: 1, L1; 2, L2; 3, isoleucine; 4, leucine, isoleucine; 5, leucine; 6, valine; 7, L3; 8, lactate; 9, alanine; 10, L4; 11, arginine; 12, lysine; 13, acetate; 14, L5; 15, acetylglycoproteins; 16, methionine; 17, glutamate; 18, glutamine; 19, L6; 20, 3-hydroxybutyrate; 21, pyruvate; 22, pyroglutamate; 23, citrate; 24, L7; 25, aspartate; 26, albumin; 27, creatine; 28, creatinine; 29, ornithine, tyrosine; 30, ornithine; 31, phenylalanine; 32, tyrosine; 33, choline; 34, beta-glucose; 35, proline; 36, alpha-glucose, beta-glucose; 37, alpha-glucose; 38, glycine; 39, glycerol; 40, mannose; 41, glyceryl groups of lipids; 42, APAP glucuronide; 43, L8; 44, uridine; 45, 1-Methylhistidine; 46, histidine; 47, 3-methylhistidine; 48, formate.

High Resolution Image
Download MS PowerPoint Slide
High-resolution visualization is also key to enable peak validation and subsequent annotation through the inspection of the shape and multiplicity of the spectral features in the neighborhood of the associated regions. Such visualization could be provided by regional plots as exemplified in Figure 3, which focuses on the glutamine region ([2.4535–2.5045] ppm). The upper panel represents the high resolution (unsigned) Manhattan plot where the −log10p-value (left Y axis) measuring the strength of association each between peak height and the log10 transformed blood glucose level is represented by a triangle (down pointing for negative associations). For the region under investigation, we define the reference feature as the strongest association (here 2.47175 ppm, p-value <2.10–16, represented in gray) and color-code the pairwise correlation of each spectral variable with this reference. The mean-corrected intensity (i.e., residuals removing the effect of possible confounders: phase and cohort, see Methods) is also represented (right Y axis) for the samples in the fifth (blue line) and 95th (green line) quantiles of the residuals log10 glucose levels after controlling for the FRS variables.

Figure 3

Figure 3. CPMG-model 2 regional association plots with log10 (glucose) for the glutamine. In the upper plot, the −log10 p-value for the features at two regions are shown on each plot. Features are colored based on their correlation with the gray hit that has the smallest p-value in the region. The lines show the mean corrected intensity (i.e., residuals removing the linear effect of the phase and the cohort) in the 5% of samples with high residual glucose in green and 5% of the samples with low residual glucose in blue. The bottom plot shows the mean spectral intensity in MESA phase 1 (plain line) and in MESA phase 2 (dashed line). Green circles indicate the proportion of replication after results prioritization.

High Resolution Image
Download MS PowerPoint Slide
From this plot, it is clear that there are strong, and exclusively positive correlation coefficients among 1H NMR data points throughout the region as indicated by the color-coded pairwise correlation coefficients (75% are >0.77). As expected, because of the local correlation structures in NMR spectra (caused by the finite peak width and the resonance being split by the J couplings), data points neighboring the reference signal exhibit higher correlation levels. More distant spectral variables may also show strong correlation with the reference peak (e.g., 2.5045, Spearman correlation = 0.26; 2.4535, Spearman correlation = 0.30), and irrespective of their location in the spectrum, intensities that are the most correlated to the reference peak exhibit the strongest associations with glucose levels. This arises because a given metabolite often has several NMR peaks and this procedure offers a way of finding such linked resonances as an aid to molecular annotation. For these variables (dark orange and red triangles), the difference in the mean corrected intensity in participants with the highest and lowest glucose levels are larger compared to the spectral variables less correlated to the reference peak (yellow triangles).
To get further insights into the nature (shape and multiplicity) of the variables found associated with glucose we represent the mean spectrum in the bottom panel (left Y axis). Mean spectra are plotted for each of the analytical phases to assess both potential technically induced bias across experimental batches, and possible population heterogeneity across samples assayed in each phase. While the mean spectrum from phase 2 appears to have more marked variables (with higher modes), both data sets yield consistent subspectra in terms of alignment, multiplicity and overall peak shape.
To accommodate the expected large number of associations with glucose levels, efficient signal prioritization strategies are key. One established way to identify robust and replicable associations is to seek external replication in an independent data set. However, owing to the specificity of experimental protocols and the nature of the measurements (which are relative and not absolute concentrations), external validation is challenging in metabolic phenotyping, but could however be sought for using either standardized or annotated data. As a workable alternative and complementary approach prioritizing relevant metabolic features, we propose to seek for internal validation and randomly split the study population in a discovery (80% of the full population) and a validation (the remaining 20%) set. The robustness of an association identified in the full population is then quantified by the number of times discovered and replicated over the (N = 100) independent splits. This proportion is represented on the regional plot (green dots on the bottom panel of Figure 2). Owing to the strong signals we identify in our glucose MWAS, we report very high replication proportions in glucose-associated regions. As a first approach, prioritized associations were defined as those discovered and validated in at least one split (i.e., with a proportion of replication >0). As illustrated in Table 3 and in Figures 4 and S-7 for CPMG and NOESY, this prioritization strategy reduced the number of significant associations for model 2 by almost 50 and 30%, respectively. As illustrated in Figure 4, the associated spectral variables are strongly clustered and define clear regions that are associated with the blood levels of glucose. Our prioritization strategy identifies overall the same regions along the spectra but selects preferentially the strongest associated variables within each region and regions that have been assigned. This is even more apparent when more stringent prioritization strategies, for instance, in Figure 4 we plot the signed Manhattan plot for the features discovered and replicated in a least half of the splits (in red).

Figure 4

Figure 4. Comparison of results from the analysis in MESA to those from the 80:20 split strategy. Results are presented for the CPMG (N = 30 590 data points) metabolome wide association study of glucose using model 2. To ease the visualization, all p-value ≤ 10–30 were set to 1 × 10–30. The −log10(p-value) is signed by the direction of the effect size estimate and is plotted against the chemical shift. The horizontal dashed line indicates the per-test significance level controlling the FWER at a 5% level. Variables found from the analyses in MESA are presented in black, those discovered and replicated at least once across the 100 splits are presented in green and those discovered and replicated in 50% of the split are presented in red. Components were: 1, L1; 2, L2; 3, isoleucine; 4, leucine, isoleucine; 5, leucine; 6, valine; 7, L3; 8, lactate; 9, alanine; 10, L4; 11, arginine; 12, lysine; 13, acetate; 14, L5; 15, acetylglycoproteins; 16, methionine; 17, glutamate; 18, glutamine; 19, L6; 20, 3-hydroxybutyrate; 21, pyruvate; 22, pyroglutamate; 23, citrate; 24, L7; 25, aspartate; 26, albumin; 27, creatine; 28, creatinine; 29, ornithine, tyrosine; 30, ornithine; 31, phenylalanine; 32, tyrosine; 33, choline; 34, beta-glucose; 35, proline; 36, alpha-glucose, beta-glucose; 37, alpha-glucose; 38, glycine; 39, glycerol; 40, mannose; 41, glyceryl groups of lipids; 42, APAP glucuronide; 43, L8; 44, uridine; 45, 1-methylhistidine; 46, histidine; 47, 3-methylhistidine; 48, formate.

High Resolution Image
Download MS PowerPoint Slide

Conclusion

ARTICLE SECTIONS
Jump To
The identification and interpretation of metabolic features that contribute to a physiological and/or disease-induced outcome from full resolution NMR data is challenging for many reasons related to the dimensionality and complexity of metabolic profiles. The main aim of the present work is to address some of these issues through the development of a robust multiple testing strategy, intuitive visualizations and objective ways to prioritize results. The MWAS approach requires accurate correction for the large number of tests performed, and therefore needs to appropriately account for the strong and complex correlation structures within the NMR spectra. All methods for performing multiple testing correction assume a valid statistical model that captures dependencies in the data. While approaches controlling the FDR usually provide less stringent multiple testing correction, these have been reported to misperform in cases of high correlations among predictors. (40) As an extension of the MWAS approach to accommodate continuous variables in a regression context, we proposed an estimation of the metabolome wide significance level adopting the same permutation strategy and investigated the sensitivity of the estimates to the data structure and to the model parametrization. Our results suggest that in the case of highly correlated variables (spectral variables) that are strongly associated with an outcome (here glucose levels), permutations do not succeed in destroying the predictor-outcome relationship, hence yielding ENT estimates greater than the actual number of tests performed. Sensitivity analyses removing extreme values, and/or log transforming glucose levels showed that outlying observations are driving this unexpected estimate of the ENT. In the presence of strong correlation (i) between blood glucose and most of the assayed metabolites, (ii) among the metabolites, and (iii) between metabolites and adjustment variables, assumptions (e.g., observations exchangeability under the null) on which permutation inferences are based may be violated, and especially in the presence of strong outlying observations. From our data, MWSL estimates appeared robust to model parametrization (i.e., marginally affected by the set of confounding variables considered), but clearly depended on the correlation structure in the data, which are population and platform specific. This suggests that the MWSL should be tailored and re-estimated for each data set. MWSL estimates were also found to be sensitive to the distribution of the outcome and especially in the case of heavy tailed distributions due to the presence of outlying observations. While numerical transformations and truncation could be a way forward, one more general and conservative option could be to calculate the MWSL once using a virtual predictor sampled from a Gaussian distribution (Figure S-4).
Using this MWSL approach, we propose extensions of existing visualization tools for the MWAS. This includes full resolution signed Manhattan plots included functional annotation and higher resolution regional plots displaying the per-variable strength of association as well as spectral summary features including correlation patterns across spectral variables. In our proof-of-principle example, we chose glucose as the outcome of interest which defined a challenging context in terms of results interpretability as a very large number of strong associations were identified and the assessment of their relevance went beyond the observed strong positive correlation for the expected glucose peaks along the 1H NMR spectra. This called for the definition of strong result prioritization strategy, which, in less extreme situations, is also critical to identify the most relevant associations that are worth dedicating resources for molecular assignment. Our approach relies on subsampling strategy where discovery (80%)-replication (20%) splits are used to identify associations that internally replicate. This strategy was found to be efficient to prioritize the most robust associations, which were not only those with the strongest p-values. We performed our internal validation at the metabolome-wide level, which was computationally intensive. To scale this approach in real-life studies, one alternative would consist in restricting the discovery-replication split strategy for the MWAS candidates that have been identified in the full population. The overall analytical strategy presented here provides a general framework for the analysis of cohort studies where large number of samples are profiled using untargeted technologies (e.g., high-resolution NMR, mass spectrometry). Overall, we believe the strategy and approach we present are generalizable and scalable and may therefore be relevant to aid the MWAS approach, particularly improving the interpretation of results.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jproteome.7b00344.

  • Plots of first two PCA scores for CPMG and NOESY data in MESA; distributions of three different transformations of glucose used to investigate MWSL; percentage of effective/actual number of tests and 95% confidence intervals for CPMG and NOESY; percentage of associated variables for CPMG and NOESY derived from each simulated continuous response; percentage of associated variables for CPMG and NOESY derived from different multiple testing correction strategies; metabolome wide study of glucose from analysis of 30 590 NOESY features; comparison of results from analysis in MESA to those from 80:20 split strategy from NOESY metabolome wide association study of glucose using model 2; significance threshold a′ and ENT based on Bonferonni correction; CPMG and NOESY metabolic features associated with log10 (glucose) in MESA at metabolome-wide significance level for models 1 and 2 (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To
  • Corresponding Author
    • Marc Chadeau-Hyam - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States;  Orcidhttp://orcid.org/0000-0001-8341-5436;  Email: [email protected]
  • Authors
    • Raphaële Castagné - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
    • Claire Laurence Boulangé - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Orcidhttp://orcid.org/0000-0002-9022-0159
    • Ibrahim Karaman - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States;  Orcidhttp://orcid.org/0000-0001-9341-8155
    • Gianluca Campanella - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
    • Diana L. Santos Ferreira - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
    • Manuja R. Kaluarachchi - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.
    • Benjamin Lehne - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
    • Alireza Moayyeri - Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom
    • Matthew R. Lewis - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.
    • Konstantina Spagou - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.
    • Anthony C. Dona - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.
    • Vangelis Evangelos - Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece
    • Russell Tracy - Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States
    • Philip Greenland - Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States
    • John C. Lindon - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom
    • David Herrington - Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
    • Timothy M. D. Ebbels - Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom
    • Paul Elliott - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States;  Orcidhttp://orcid.org/0000-0002-7511-5684
    • Ioanna Tzoulaki - †Department of Epidemiology and Biostatistics, School of Public Health and ‡MRC-PHE Centre for Environment and Health, Imperial College London, W2 1PG London, United Kingdom;  Bioincubator Unit, Metabometrix Ltd, Bessemer Building, Prince Consort Road, South Kensington, London SW7 2BP U.K.;  Farr Institute of Health Informatics Research, University College London Institute of Health Informatics, 222 Euston Road, NW1 2DA London, United Kingdom;  Department of Hygiene and Epidemiology, University of Ioannina Medical School, Ioannina 45110, Greece;  Department of Pathology and Laboratory Medicine, University of Vermont Larner College of Medicine, Burlington, Vermont 05405, United States;  Department of Preventive Medicine and the Institute for Public Health and Medicine, Northwestern University, Chicago, Illinois 60611, United States;  Computational and Systems Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, SW7 2AZ London, United Kingdom;  Section on Cardiovascular Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157, United States
  • Notes
    The authors declare no competing financial interest.

    A Step-by-step R tutorial used to estimate the effective number of tests and to produce Figures 2 and 3 is publicly available at https://figshare.com/articles/Tutorial_for_MWAS_and_regional_plots/5319247.

Acknowledgment

ARTICLE SECTIONS
Jump To

This work has been carried out as part of the of the FP7 project COMBI-BIO [305422 to P. E.]. MESA was supported by Contract Nos. HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by Grant Nos. UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from NCATS. P.E. is Director of the MRC-PHE Centre for Environment and Health and acknowledges support from the Medical Research Council and Public Health England (MR/L01341X/1). P.E. acknowledges support from the NIHR Biomedical Research Centre at Imperial College Healthcare NHS Trust and Imperial College London, and the NIHR Health Protection Research Unit in Health Impact of Environmental Hazards (HPRU-2012–10141). This work used the computing resources of the UK MEDical BIOinformatics partnership (UKMED-BIO) supported by the Medical Research Council (MR/L01632X/1). The authors wish to thank all the centres that took part in the study and the additional members of the COMBI-BIO consortium. The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.

References

ARTICLE SECTIONS
Jump To

This article references 40 other publications.

  1. 1
    Lenz, E. M.; Wilson, I. D. Analytical strategies in metabonomics J. Proteome Res. 2007, 6 (2) 443– 458 DOI: 10.1021/pr0605217
    Google Scholar
    1
    Analytical Strategies in Metabonomics
    Lenz, Eva Maria; Wilson, Ian D.
    Journal of Proteome Research (2007), 6 (2), 443-458CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
    A review. To perform metabonomics investigations, it is necessary to generate comprehensive metabolite profiles for complex samples such as biofluids and tissue/tissue exts. Anal. technologies that can be used to achieve this aim are constantly evolving, and new developments are changing the way in which such profiles' metabolite profiles can be generated. Here, the utility of various anal. techniques for global metabolite profiling, such as, e.g., 1H NMR, MS, HPLC-MS, and GC-MS, are explored and compared.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD28Xht12qt7%252FM&md5=b63fd8bf30a64a62ef035e0c8880608a
  2. 2
    Lindon, J. C.; Nicholson, J. K. Analytical technologies for metabonomics and metabolomics, and multi-omic information recovery TrAC, Trends Anal. Chem. 2008, 27 (3) 194– 204 DOI: 10.1016/j.trac.2007.08.009
    Google Scholar
    There is no corresponding record for this reference.
  3. 3
    Holmes, E.; Wilson, I. D.; Nicholson, J. K. Metabolic phenotyping in health and disease Cell 2008, 134 (5) 714– 717 DOI: 10.1016/j.cell.2008.08.026
    Google Scholar
    3
    Metabolic phenotyping in health and disease
    Holmes, Elaine; Wilson, Ian D.; Nicholson, Jeremy K.
    Cell (Cambridge, MA, United States) (2008), 134 (5), 714-717CODEN: CELLB5; ISSN:0092-8674. (Cell Press)
    A review. Analyzing metabolites (small mols. <1 kDa) in body fluids such as urine and plasma using various spectroscopic methods provides information on the metabotype (metabolic phenotype) of individuals or populations, information that can be applied to personalized medicine or public healthcare.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCqs7jE&md5=f82b72659258117a777e4834c7f3233f
  4. 4
    Madsen, R.; Lundstedt, T.; Trygg, J. Chemometrics in metabolomics--a review in human disease diagnosis Anal. Chim. Acta 2010, 659 (1–2) 23– 33 DOI: 10.1016/j.aca.2009.11.042
    Google Scholar
    4
    Chemometrics in metabolomics. A review in human disease diagnosis
    Madsen, Rasmus; Lundstedt, Torbjoern; Trygg, Johan
    Analytica Chimica Acta (2010), 659 (1-2), 23-33CODEN: ACACAM; ISSN:0003-2670. (Elsevier B.V.)
    A review. Metabolomics is a post genomic research field concerned with developing methods for anal. of low mol. wt. compds. in biol. systems, such as cells, organs or organisms. Analyzing metabolic differences between unperturbed and perturbed systems, such as healthy volunteers and patients with a disease, can lead to insights into the underlying pathol. In metabolomics anal., large amts. of data are routinely produced in order to characterize samples. The use of multivariate data anal. techniques and chemometrics is a commonly used strategy for obtaining reliable results. Metabolomics have been applied in different fields such as disease diagnosis, toxicol., plant science and pharmaceutical and environmental research. In this review we take a closer look at the chemometric methods used and the available results within the field of disease diagnosis. We will first present some current strategies for performing metabolomics studies, esp. regarding disease diagnosis. The main focus will be on data anal. strategies and validation of multivariate models, since there are many pitfalls in this regard. Further, we highlight the most interesting metabolomics publications and discuss these in detail; addnl. studies are mentioned as a ref. for the interested reader. A general trend is an increased focus on biol. interpretation rather than merely the ability to classify samples. In the conclusions, the general trends and some recommendations for improving metabolomics data anal. are provided.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXhs1WnsbrL&md5=7e95e755f785e361f6cb67647969e565
  5. 5
    Heather, L. C.; Wang, X.; West, J. A.; Griffin, J. L. A practical guide to metabolomic profiling as a discovery tool for human heart disease J. Mol. Cell. Cardiol. 2013, 55, 2– 11 DOI: 10.1016/j.yjmcc.2012.12.001
    Google Scholar
    5
    A practical guide to metabolomic profiling as a discovery tool for human heart disease
    Heather, Lisa C.; Wang, Xinzhu; West, James A.; Griffin, Julian L.
    Journal of Molecular and Cellular Cardiology (2013), 55 (), 2-11CODEN: JMCDAY; ISSN:0022-2828. (Elsevier B.V.)
    A review. Metabolomics has refreshed interest in metab. across biol. and medicine, particularly in the areas of functional genomics and biomarker discovery. In this review we will discuss the exptl. techniques and challenges involved in metabolomic profiling and how these technologies have been applied to cardiovascular disease. Open profiling and targeted approaches to metabolomics are compared, focusing on high resoln. NMR spectroscopy and mass spectrometry, as well as discussing how to analyze the large amts. of data generated using multivariate statistics. Finally, the current literature on metabolomic profiling in human cardiovascular disease is reviewed to illustrate the diversity of approaches, and discuss some of the key metabolites and pathways found to be perturbed in plasma, urine and tissue from patients with these diseases. This includes studies of coronary artery disease, myocardial infarction, and ischemic heart disease. These studies demonstrate the potential of the technol. for biomarker discovery and elucidating metabolic mechanisms assocd. with given pathologies, but also in some cases provide a warning of the pitfalls of poor study design. This article is part of a Special Issue entitled "Focus on Cardiac Metab.".
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXkt1SgsQ%253D%253D&md5=e8364a80fa036bb28291f64c0b7b50a3
  6. 6
    Tzoulaki, I.; Ebbels, T. M. D.; Valdes, A.; Elliott, P.; Ioannidis, J. P. A. Design and analysis of metabolomics studies in epidemiologic research: a primer on -omic technologies Am. J. Epidemiol. 2014, 180 (2) 129– 139 DOI: 10.1093/aje/kwu143
    Google Scholar
    6
    Design and analysis of metabolomics studies in epidemiologic research: a primer on -omic technologies
    Tzoulaki Ioanna; Ebbels Timothy M D; Valdes Ana; Elliott Paul; Ioannidis John P A
    American journal of epidemiology (2014), 180 (2), 129-39 ISSN:.
    Metabolomics is the field of "-omics" research concerned with the comprehensive characterization of the small low-molecular-weight metabolites in biological samples. In epidemiology, it represents an emerging technology and an unprecedented opportunity to measure environmental and other exposures with improved precision and far less measurement error than with standard epidemiologic methods. Advances in the application of metabolomics in large-scale epidemiologic research are now being realized through a combination of improved sample preparation and handling, automated laboratory and processing methods, and reduction in costs. The number of epidemiologic studies that use metabolic profiling is still limited, but it is fast gaining popularity in this area. In the present article, we present a roadmap for metabolomic analyses in epidemiologic studies and discuss the various challenges these data pose to large-scale studies. We discuss the steps of data preprocessing, univariate and multivariate data analysis, correction for multiplicity of comparisons with correlated data, and finally the steps of cross-validation and external validation. As data from metabolomic studies accumulate in epidemiology, there is a need for large-scale replication and synthesis of findings, increased availability of raw data, and a focus on good study design, all of which will highlight the potential clinical impact of metabolomics in this field.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC2cfmslWlsA%253D%253D&md5=c186d66aeef1ec835bbc1d92cf0046fa
  7. 7
    Dumas, M.-E.; Maibaum, E. C.; Teague, C.; Ueshima, H.; Zhou, B.; Lindon, J. C.; Nicholson, J. K.; Stamler, J.; Elliott, P.; Chan, Q. Assessment of analytical reproducibility of 1H NMR spectroscopy based metabonomics for large-scale epidemiological research: the INTERMAP Study Anal. Chem. 2006, 78 (7) 2199– 2208 DOI: 10.1021/ac0517085
    Google Scholar
    There is no corresponding record for this reference.
  8. 8
    Dona, A. C.; Jiménez, B.; Schäfer, H.; Humpfer, E.; Spraul, M.; Lewis, M. R.; Pearce, J. T. M.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. Precision high-throughput proton NMR spectroscopy of human urine, serum, and plasma for large-scale metabolic phenotyping Anal. Chem. 2014, 86 (19) 9887– 9894 DOI: 10.1021/ac5025039
    Google Scholar
    8
    Precision High-Throughput Proton NMR Spectroscopy of Human Urine, Serum, and Plasma for Large-Scale Metabolic Phenotyping
    Dona, Anthony C.; Jimenez, Beatriz; Schafer, Hartmut; Humpfer, Eberhard; Spraul, Manfred; Lewis, Matthew R.; Pearce, Jake T. M.; Holmes, Elaine; Lindon, John C.; Nicholson, Jeremy K.
    Analytical Chemistry (Washington, DC, United States) (2014), 86 (19), 9887-9894CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Proton NMR-based metabolic phenotyping of urine and blood plasma/serum samples provides important prognostic and diagnostic information and permits monitoring of disease progression in an objective manner. Much effort has been made in recent years to develop NMR instrumentation and technol. to allow the acquisition of data in an effective, reproducible, and high-throughput approach that allows the study of general population samples from epidemiol. collections for biomarkers of disease risk. The challenge remains to develop highly reproducible methods and standardized protocols that minimize tech. or exptl. bias, allowing realistic interlab. comparisons of subtle biomarker information. Here the authors present a detailed set of updated protocols that carefully consider major exptl. conditions, including sample prepn., spectrometer parameters, NMR pulse sequences, throughput, reproducibility, quality control, and resoln. These results provide an exptl. platform that facilitates NMR spectroscopy usage across different large cohorts of biofluid samples, enabling integration of global metabolic profiling that is a prerequisite for personalized healthcare.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhsVGhtLbE&md5=34870c3f541fd9d5397ba7252c6010fa
  9. 9
    Nicholson, J. K. Global systems biology, personalized medicine and molecular epidemiology Mol. Syst. Biol. 2006, 2, 52 DOI: 10.1038/msb4100095
    Google Scholar
    There is no corresponding record for this reference.
  10. 10
    Nicholson, J. K.; Holmes, E.; Kinross, J. M.; Darzi, A. W.; Takats, Z.; Lindon, J. C. Metabolic phenotyping in clinical and surgical environments Nature 2012, 491 (7424) 384– 392 DOI: 10.1038/nature11708
    Google Scholar
    10
    Metabolic phenotyping in clinical and surgical environments
    Nicholson, Jeremy K.; Holmes, Elaine; Kinross, James M.; Darzi, Ara W.; Takats, Zoltan; Lindon, John C.
    Nature (London, United Kingdom) (2012), 491 (7424), 384-392CODEN: NATUAS; ISSN:0028-0836. (Nature Publishing Group)
    A review. Metabolic phenotyping involves the comprehensive anal. of biol. fluids or tissue samples. This anal. allows biochem. classification of a person's physiol. or pathol. states that relate to disease diagnosis or prognosis at the individual level and to disease risk factors at the population level. These approaches are currently being implemented in hospital environments and in regional phenotyping centers worldwide. The ultimate aim of such work is to generate information on patient biol. using techniques such as patient stratification to better inform clinicians on factors that will enhance diagnosis or the choice of therapy. There have been many reports of direct applications of metabolic phenotyping in a clin. setting.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ogs7vM&md5=5fe49f7f75ab76e6c55b338021559098
  11. 11
    Sévin, D. C.; Kuehne, A.; Zamboni, N.; Sauer, U. Biological insights through nontargeted metabolomics Curr. Opin. Biotechnol. 2015, 34, 1– 8 DOI: 10.1016/j.copbio.2014.10.001
    Google Scholar
    11
    Biological insights through nontargeted metabolomics
    Sevin, Daniel C.; Kuehne, Andreas; Zamboni, Nicola; Sauer, Uwe
    Current Opinion in Biotechnology (2015), 34 (), 1-8CODEN: CUOBE3; ISSN:0958-1669. (Elsevier B.V.)
    A review. Metabolomics is increasingly employed to investigate metab. and its reciprocal crosstalk with cellular signaling and regulation. In recent years, several nontargeted metabolomics methods providing substantial metabolome coverage have been developed. Here, we review and compare the contributions of traditional targeted and nontargeted metabolomics in advancing different research areas ranging from biotechnol. to human health. Although some studies demonstrated the power of nontargeted profiling in generating unexpected and yet highly important insights, we found that most mechanistic links were still revealed by hypothesis-driven targeted methods. Novel computational approaches for formal interpretation of complex metabolic patterns and integration of complementary mol. layers are required to tap the full potential of nontargeted metabolomics for data-driven, discovery-oriented research and rapidly nucleating novel biol. insights.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhslOkt7jI&md5=4bbd0528f7b2346c2c07191ca165cdd2
  12. 12
    Bictash, M.; Ebbels, T. M.; Chan, Q.; Loo, R. L.; Yap, I. K. S.; Brown, I. J.; de Iorio, M.; Daviglus, M. L.; Holmes, E.; Stamler, J. Opening up the “Black Box”: metabolic phenotyping and metabolome-wide association studies in epidemiology J. Clin. Epidemiol. 2010, 63 (9) 970– 979 DOI: 10.1016/j.jclinepi.2009.10.001
    Google Scholar
    12
    Opening up the "Black Box": metabolic phenotyping and metabolome-wide association studies in epidemiology
    Bictash Magda; Ebbels Timothy M; Chan Queenie; Loo Ruey Leng; Yap Ivan K S; Brown Ian J; de Iorio Maria; Daviglus Martha L; Holmes Elaine; Stamler Jeremiah; Nicholson Jeremy K; Elliott Paul
    Journal of clinical epidemiology (2010), 63 (9), 970-9 ISSN:.
    BACKGROUND: Metabolic phenotyping of humans allows information to be captured on the interactions between dietary, xenobiotic, other lifestyle and environmental exposures, and genetic variation, which together influence the balance between health and disease risks at both individual and population levels. OBJECTIVES: We describe here the main procedures in large-scale metabolic phenotyping and their application to metabolome-wide association (MWA) studies. METHODS: By use of high-throughput technologies and advanced spectroscopic methods, application of metabolic profiling to large-scale epidemiologic sample collections, including metabolome-wide association (MWA) studies for biomarker discovery and identification. DISCUSSION: Metabolic profiling at epidemiologic scale requires optimization of experimental protocol to maximize reproducibility, sensitivity, and quantitative reliability, and to reduce analytical drift. Customized multivariate statistical modeling approaches are needed for effective data visualization and biomarker discovery with control for false-positive associations since 100s or 1,000s of complex metabolic spectra are being processed. CONCLUSION: Metabolic profiling is an exciting addition to the armamentarium of the epidemiologist for the discovery of new disease-risk biomarkers and diagnostics, and to provide novel insights into etiology, biological mechanisms, and pathways.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC3cjhsVaqsw%253D%253D&md5=4bb1ffa9f64917c6931a7ee45bcd3788
  13. 13
    Chadeau-Hyam, M.; Ebbels, T. M. D.; Brown, I. J.; Chan, Q.; Stamler, J.; Huang, C. C.; Daviglus, M. L.; Ueshima, H.; Zhao, L.; Holmes, E. Metabolic profiling and the metabolome-wide association study: significance level for biomarker identification J. Proteome Res. 2010, 9 (9) 4620– 4627 DOI: 10.1021/pr1003449
    Google Scholar
    13
    Metabolic Profiling and the Metabolome-Wide Association Study: Significance Level For Biomarker Identification
    Chadeau-Hyam, Marc; Ebbels, Timothy M. D.; Brown, Ian J.; Chan, Queenie; Stamler, Jeremiah; Huang, Chiang Ching; Daviglus, Martha L.; Ueshima, Hirotsugu; Zhao, Liancheng; Holmes, Elaine; Nicholson, Jeremy K.; Elliott, Paul; De Iorio, Maria
    Journal of Proteome Research (2010), 9 (9), 4620-4627CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
    High throughput metabolic profiling via the metabolome-wide assocn. study (MWAS) is a powerful new approach to identify biomarkers of disease risk, but there are methodol. challenges: high dimensionality, high level of collinearity, the existence of peak overlap within metabolic spectral data, multiple testing, and selection of a suitable significance threshold. The authors define the metabolome-wide significance level (MWSL) as the threshold required to control the family wise error rate through a permutation approach. The authors used 1H NMR spectroscopic profiles of 24 h urinary collections from the INTERMAP study. The authors' results show that the MWSL primarily depends on sample size and spectral resoln. The MWSL ests. can be used to guide selection of discriminatory biomarkers in MWA studies. In a simulation study, the authors compare statistical performance of the MWSL approach to two variants of orthogonal partial least-squares (OPLS) method with respect to statistical power, false pos. rate and correspondence of ranking of the most significant spectral variables. The authors' results show that the MWSL approach as estd. by the univariate t test is not outperformed by OPLS and offers a fast and simple method to detect disease-related discriminatory features in human NMR urinary metabolic profiles.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXhtVSju7nI&md5=c4e86345e9c4ef1f6e2b40bac05956a2
  14. 14
    De Livera, A. M.; Dias, D. A.; De Souza, D.; Rupasinghe, T.; Pyke, J.; Tull, D.; Roessner, U.; McConville, M.; Speed, T. P. Normalizing and integrating metabolomics data Anal. Chem. 2012, 84 (24) 10768– 10776 DOI: 10.1021/ac302748b
    Google Scholar
    14
    Normalizing and Integrating Metabolomics Data
    De Livera, Alysha M.; Dias, Daniel A.; De Souza, David; Rupasinghe, Thusitha; Pyke, James; Tull, Dedreia; Roessner, Ute; McConville, Malcolm; Speed, Terence P.
    Analytical Chemistry (Washington, DC, United States) (2012), 84 (24), 10768-10776CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Metabolomics research often requires the use of multiple anal. platforms, batches of samples, and labs., any of which can introduce a component of unwanted variation. In addn., every expt. is subject to within-platform and other exptl. variation, which often includes unwanted biol. variation. Such variation must be removed in order to focus on the biol. information of interest. We present a broadly applicable method for the removal of unwanted variation arising from various sources for the identification of differentially abundant metabolites and, hence, for the systematic integration of data on the same quantities from different sources. We illustrate the versatility and the performance of the approach in four applications, and we show that it has several advantages over the existing normalization methods.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ons7fF&md5=2de0f2d242c1a42dfb539d2570ca50f5
  15. 15
    De Livera, A. M.; Olshansky, M.; Speed, T. P. Statistical analysis of metabolomics data Methods Mol. Biol. 2013, 1055, 291– 307 DOI: 10.1007/978-1-62703-577-4_20
    Google Scholar
    15
    Statistical analysis of metabolomics data
    De Livera, Alysha M.; Olshansky, Moshe; Speed, Terence P.
    Methods in Molecular Biology (New York, NY, United States) (2013), 1055 (Metabolomics Tools for Natural Product Discovery), 291-307CODEN: MMBIED; ISSN:1064-3745. (Springer)
    Statistical matters form an integral part of a metabolomics expt. In this chapter we describe several important aspects in the anal. of metabolomics data such as the removal of unwanted variation and the identification of differentially abundant metabolites, along with a no. of other essential statistical considerations.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhslyhs7nN&md5=8940f4025856d8f08dbabd81679022eb
  16. 16
    Korman, A.; Oh, A.; Raskind, A.; Banks, D. Statistical methods in metabolomics Methods Mol. Biol. 2012, 856, 381– 413 DOI: 10.1007/978-1-61779-585-5_16
    Google Scholar
    16
    Statistical methods in metabolomics
    Korman Alexander; Oh Amy; Raskind Alexander; Banks David
    Methods in molecular biology (Clifton, N.J.) (2012), 856 (), 381-413 ISSN:.
    Metabolomics is the relatively new field in bioinformatics that uses measurements on metabolite abundance as a tool for disease diagnosis and other medical purposes. Although closely related to proteomics, the statistical analysis is potentially simpler since biochemists have significantly more domain knowledge about metabolites. This chapter reviews the challenges that metabolomics poses in the areas of quality control, statistical metrology, and data mining.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC383psVCltQ%253D%253D&md5=6be2769522ef030929b346dd4014a945
  17. 17
    Ebbels, T. M. D.; Lindon, J. C.; Coen, M. Processing and modeling of nuclear magnetic resonance (NMR) metabolic profiles Methods Mol. Biol. 2011, 708, 365– 388 DOI: 10.1007/978-1-61737-985-7_21
    Google Scholar
    17
    Processing and modeling of nuclear magnetic resonance (NMR) metabolic profiles
    Ebbels, Timothy M. D.; Lindon, John C.; Coen, Muireann
    Methods in Molecular Biology (New York, NY, United States) (2011), 708 (Metabolic Profiling), 365-388CODEN: MMBIED; ISSN:1064-3745. (Springer)
    Modern NMR spectroscopy generates complex and information-rich metabolic profiles. These require robust, accurate, and often sophisticated statistical techniques to yield the max. meaningful knowledge. In this chapter, we describe methods typically used to analyze such data. We begin by describing seven goals of metabolic profile anal., ranging from prodn. of a data table to multi-omic integration for systems biol. Methods for preprocessing and pretreatment are then presented, including issues such as instrument-level spectral processing, data redn. and deconvolution, normalization, scaling, and transformations of the data. We then discuss methods for exploratory modeling and exemplify three techniques: principal components anal., hierarchical clustering, and self-organizing maps. Moving to predictive modeling, we focus our discussion on partial least squares regression, orthogonal partial least squares regression, and genetic algorithm approaches. A typical set of in vitro metabolic profiles is used where possible to compare and contrast the methods. The importance of validating statistical models is highlighted, and std. techniques for doing so, such as training/test set and cross-validation are described. Finally, we discuss the contributions of statistical techniques such as statistical total correlation spectroscopy, and other correlation-based methods have made to the process of structural characterization for unknown metabolites.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXmsVSgur4%253D&md5=70b8d3811208b2588145d74fa39eabde
  18. 18
    Bild, D. E.; Bluemke, D. A.; Burke, G. L.; Detrano, R.; Diez Roux, A. V.; Folsom, A. R.; Greenland, P.; Jacob, D. R.; Kronmal, R.; Liu, K. Multi-ethnic study of atherosclerosis: objectives and design Am. J. Epidemiol. 2002, 156 (9) 871– 881 DOI: 10.1093/aje/kwf113
    Google Scholar
    18
    Multi-Ethnic Study of Atherosclerosis: objectives and design
    Bild Diane E; Bluemke David A; Burke Gregory L; Detrano Robert; Diez Roux Ana V; Folsom Aaron R; Greenland Philip; Jacob David R Jr; Kronmal Richard; Liu Kiang; Nelson Jennifer Clark; O'Leary Daniel; Saad Mohammed F; Shea Steven; Szklo Moyses; Tracy Russell P
    American journal of epidemiology (2002), 156 (9), 871-81 ISSN:0002-9262.
    The Multi-Ethnic Study of Atherosclerosis was initiated in July 2000 to investigate the prevalence, correlates, and progression of subclinical cardiovascular disease (CVD) in a population-based sample of 6,500 men and women aged 45-84 years. The cohort will be selected from six US field centers. Approximately 38% of the cohort will be White, 28% African-American, 23% Hispanic, and 11% Asian (of Chinese descent). Baseline measurements will include measurement of coronary calcium using computed tomography; measurement of ventricular mass and function using cardiac magnetic resonance imaging; measurement of flow-mediated brachial artery endothelial vasodilation, carotid intimal-medial wall thickness, and distensibility of the carotid arteries using ultrasonography; measurement of peripheral vascular disease using ankle and brachial blood pressures; electrocardiography; and assessments of microalbuminuria, standard CVD risk factors, sociodemographic factors, life habits, and psychosocial factors. Blood samples will be assayed for putative biochemical risk factors and stored for use in nested case-control studies. DNA will be extracted and lymphocytes will be immortalized for genetic studies. Measurement of selected subclinical disease indicators and risk factors will be repeated for the study of progression over 7 years. Participants will be followed through 2008 for identification and characterization of CVD events, including acute myocardial infarction and other coronary heart disease, stroke, peripheral vascular disease, and congestive heart failure; therapeutic interventions for CVD; and mortality.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BD38njslemsA%253D%253D&md5=7cc196a4c0a768cbb6d77484bc663abe
  19. 19
    Friedewald, W. T.; Levy, R. I.; Fredrickson, D. S. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge Clin. Chem. 1972, 18 (6) 499– 502
    Google Scholar
    19
    Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge
    Friedewald, William T.; Levy, Robert I.; Fredrickson, Donald S.
    Clinical Chemistry (Washington, DC, United States) (1972), 18 (6), 499-502CODEN: CLCHAU; ISSN:0009-9147.
    A method for evaluating the cholesterol content of the serum low-d. lipoprotein fraction (Sf 0-20) involves measurements of fasting plasma total cholesterol, triglyceride, and high-d. lipoprotein cholesterol concns. none of which requires the use of the preparative ultracentrifuge. Comparison of this procedure with the more direct procedure, in which the ultracentrifuge is used, yielded correlation coeffs. of 0.94-0.99, depending on the patient population compared.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaE38Xkslagsbw%253D&md5=0422d2b1673a49c8344179ec2cf4d7e4
  20. 20
    Karaman, I.; Ferreira, D. L. S.; Boulangé, C. L.; Kaluarachchi, M. R.; Herrington, D.; Dona, A. C.; Castagné, R.; Moayyeri, A.; Lehne, B.; Loh, M. Workflow for Integrated Processing of Multicohort Untargeted (1)H NMR Metabolomics Data in Large-Scale Metabolic Epidemiology J. Proteome Res. 2016, 15 (12) 4188– 4194 DOI: 10.1021/acs.jproteome.6b00125
    Google Scholar
    20
    Workflow for Integrated Processing of Multicohort Untargeted 1H NMR Metabolomics Data in Large-Scale Metabolic Epidemiology
    Karaman, Ibrahim; Ferreira, Diana L. S.; Boulange, Claire L.; Kaluarachchi, Manuja R.; Herrington, David; Dona, Anthony C.; Castagne, Raphaele; Moayyeri, Alireza; Lehne, Benjamin; Loh, Marie; de Vries, Paul S.; Dehghan, Abbas; Franco, Oscar H.; Hofman, Albert; Evangelou, Evangelos; Tzoulaki, Ioanna; Elliott, Paul; Lindon, John C.; Ebbels, Timothy M. D.
    Journal of Proteome Research (2016), 15 (12), 4188-4194CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
    Large-scale metabolomics studies involving thousands of samples present multiple challenges in data anal., particularly when an untargeted platform is used. Studies with multiple cohorts and anal. platforms exacerbate existing problems such as peak alignment and normalization. Therefore, there is a need for robust processing pipelines which can ensure reliable data for statistical anal. The COMBI-BIO project incorporates serum from ∼8000 individuals, in 3 cohorts, profiled by 6 assays in 2 phases using both 1H-NMR and UPLC-MS. Here the authors present the COMBI-BIO NMR anal. pipeline and demonstrate its fitness for purpose using representative quality control (QC) samples. NMR spectra were first aligned and normalized. After eliminating interfering signals, outliers identified using Hotelling's T2 were removed and a cohort/phase adjustment was applied, resulting in two NMR datasets (CPMG and NOESY). Alignment of the NMR data increases the correlation-based alignment quality measure from 0.319 to 0.391 for CPMG and from 0.536 to 0.586 for NOESY, showing that the improvement was present across both large and small peaks. End-to-end quality assessment of the pipeline was achieved using Hotelling's T2 distributions. For CPMG spectra, the interquartile range decreased from 1.425 in raw QC data to 0.679 in processed spectra, while the corresponding change for NOESY spectra was 0.795-0.636 indicating an improvement in precision following processing. PCA indicated that gross phase and cohort differences were no longer present. These results illustrate that the pipeline produces robust and reproducible data, successfully addressing the methodol. challenges of this large multi-faceted study.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhsFWjtbfE&md5=5c953b368c549337338ef48eec74a13c
  21. 21
    Veselkov, K. A.; Lindon, J. C.; Ebbels, T. M. D.; Crockford, D.; Volynkin, V. V.; Holmes, E.; Davies, D. B.; Nicholson, J. K. Recursive segment-wise peak alignment of biological (1)h NMR spectra for improved metabolic biomarker recovery Anal. Chem. 2009, 81 (1) 56– 66 DOI: 10.1021/ac8011544
    Google Scholar
    21
    Recursive Segment-Wise Peak Alignment of Biological 1H NMR Spectra for Improved Metabolic Biomarker Recovery
    Veselkov, Kirill A.; Lindon, John C.; Ebbels, Timothy M. D.; Crockford, Derek; Volynkin, Vladimir V.; Holmes, Elaine; Davies, David B.; Nicholson, Jeremy K.
    Analytical Chemistry (Washington, DC, United States) (2009), 81 (1), 56-66CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Chem. shift variation in small-mol. 1H NMR signals of biofluids complicates biomarker information recovery in metabonomic studies when using multivariate statistical and pattern recognition tools. Current peak realignment methods are generally time-consuming or align major peaks at the expense of minor peak shift accuracy. The authors present a novel recursive segment-wise peak alignment (RSPA) method to reduce variability in peak positions across the multiple 1H NMR spectra used in metabonomic studies. The method refines a segmentation of ref. and test spectra in a top-down fashion, sequentially subdividing the initial larger segments, as required, to improve the local spectral alignment. The authors also describe a general procedure that allows robust comparison of realignment quality of various available methods for a range of peak intensities. The RSPA method is illustrated with respect to 140 1H NMR rat urine spectra from a caloric restriction study and is compared with several other widely used peak alignment methods. The authors demonstrate the superior performance of the RSPA alignment over a wide range of peaks and its capacity to enhance interpretability and robustness of multivariate statistical tools. The approach is widely applicable for NMR-based metabolic studies and is potentially suitable for many other types of data sets such as chromatog. profiles and MS data.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhsVGnsLnJ&md5=f8386d82f13d01b0609d0d504f899787
  22. 22
    Dieterle, F.; Ross, A.; Schlotterbeck, G.; Senn, H. Probabilistic quotient normalization as robust method to account for dilution of complex biological mixtures. Application in 1H NMR metabonomics Anal. Chem. 2006, 78 (13) 4281– 4290 DOI: 10.1021/ac051632c
    Google Scholar
    22
    Probabilistic Quotient Normalization as Robust Method to Account for Dilution of Complex Biological Mixtures. Application in 1H NMR Metabonomics
    Dieterle, Frank; Ross, Alfred; Schlotterbeck, Goetz; Senn, Hans
    Analytical Chemistry (2006), 78 (13), 4281-4290CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    For the anal. of the spectra of complex biofluids, preprocessing methods play a crucial role in rendering the subsequent data analyses more robust and accurate. Normalization is a preprocessing method, which accounts for different dilns. of samples by scaling the spectra to the same virtual overall concn. In the field of 1H NMR metabonomics integral normalization, which scales spectra to the same total integral, is the de facto std. In this work, it is shown that integral normalization is a suboptimal method for normalizing spectra from metabonomic studies. Esp. strong metabonomic changes, evident as massive amts. of single metabolites in samples, significantly hamper the integral normalization resulting in incorrectly scaled spectra. The probabilistic quotient normalization is introduced in this work. This method is based on the calcn. of a most probable diln. factor by looking at the distribution of the quotients of the amplitudes of a test spectrum by those of a ref. spectrum. Simulated spectra, spectra of urine samples from a metabonomic study with cyclosporin-A as the active compd., and spectra of more than 4000 samples of control animals demonstrate that the probabilistic quotient normalization is by far more robust and more accurate than the widespread integral normalization and vector length normalization.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD28XltVCgtro%253D&md5=6eb6377326a9df2a59b6afb2a9c6e47d
  23. 23
    van Velzen, E. J. J.; Westerhuis, J. A.; van Duynhoven, J. P. M.; van Dorsten, F. A.; Hoefsloot, H. C. J.; Jacobs, D. M.; Smit, S.; Draijer, R.; Kroner, C. I.; Smilde, A. K. Multilevel data analysis of a crossover designed human nutritional intervention study J. Proteome Res. 2008, 7 (10) 4483– 4491 DOI: 10.1021/pr800145j
    Google Scholar
    23
    Multilevel data analysis of a crossover designed human nutritional intervention study
    van Velzen, Ewoud J. J.; Westerhuis, Johan A.; van Duynhoven, John P. M.; van Dorsten, Ferdi A.; Hoefsloot, Huub C. J.; Jacobs, Doris M.; Smit, Suzanne; Draijer, Richard; Kroner, Christine I.; Smilde, Age K.
    Journal of Proteome Research (2008), 7 (10), 4483-4491CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
    A new method is introduced for the anal. of '-omics' data derived from crossover designed drug or nutritional intervention studies. The method aims at finding systematic variations in metabolic profiles after a drug or nutritional challenge and takes advantage of the crossover design in the data. The method, which can be considered as a multivariate extension of paired t test, generates different multivariate submodels for the between- and the within-subject variation in the data. A major advantage of this variation splitting is that each submodel can be analyzed sep. without confounding with other variation sources. The power of the multilevel approach is demonstrated in a human nutritional intervention study which used NMR-based metabolomics to assess the metabolic impact of grape/wine ext. consumption. The variations in urine metabolic profiles were studied between and within the human subjects using the multilevel anal. After variation splitting, the multilevel PCA was used to investigate the exptl. and biol. differences between the subjects, whereas a multilevel PLS-DA model was used to reveal the net treatment effect within the subjects. The obsd. treatment effect was validated with cross model validation and permutations. The statistical significance of the multilevel classification model is a major improvement compared to ordinary PLS-DA models without variation splitting. Rank products were used to det. which NMR signals were most important in the multilevel classification model.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtVGgu77J&md5=7374579a6254bb8b23fb922336b8d3a2
  24. 24
    Cloarec, O.; Dumas, M.-E.; Craig, A.; Barton, R. H.; Trygg, J.; Hudson, J.; Blancher, C.; Gauguier, D.; Lindon, J. C.; Holmes, E. Statistical total correlation spectroscopy: an exploratory approach for latent biomarker identification from metabolic 1H NMR data sets Anal. Chem. 2005, 77 (5) 1282– 1289 DOI: 10.1021/ac048630x
    Google Scholar
    24
    Statistical Total Correlation Spectroscopy: An Exploratory Approach for Latent Biomarker Identification from Metabolic 1H NMR Data Sets
    Cloarec, Olivier; Dumas, Marc-Emmanuel; Craig, Andrew; Barton, Richard H.; Trygg, Johan; Hudson, Jane; Blancher, Christine; Gauguier, Dominique; Lindon, John C.; Holmes, Elaine; Nicholson, Jeremy
    Analytical Chemistry (2005), 77 (5), 1282-1289CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    We describe here the implementation of the statistical total correlation spectroscopy (STOCSY) anal. method for aiding the identification of potential biomarker mols. in metabonomic studies based on NMR spectroscopic data. STOCSY takes advantage of the multicollinearity of the intensity variables in a set of spectra (in this case 1H NMR spectra) to generate a pseudo-two-dimensional NMR spectrum that displays the correlation among the intensities of the various peaks across the whole sample. This method is not limited to the usual connectivities that are deducible from more std. two-dimensional NMR spectroscopic methods, such as TOCSY. Moreover, two or more mols. involved in the same pathway can also present high intermol. correlations because of biol. covariance or can even be anticorrelated. This combination of STOCSY with supervised pattern recognition and particularly orthogonal projection on latent structure-discriminant anal. (O-PLS-DA) offers a new powerful framework for anal. of metabonomic data. In a first step O-PLS-DA exts. the part of NMR spectra related to discrimination. This information is then cross-combined with the STOCSY results to help identify the mols. responsible for the metabolic variation. To illustrate the applicability of the method, it has been applied to 1H NMR spectra of urine from a metabonomic study of a model of insulin resistance based on the administration of a carbohydrate diet to three different mice strains (C57BL/6Oxjr, BALB/cOxjr, and 129S6/SvEvOxjr) in which a series of metabolites of biol. importance can be conclusively assigned and identified by use of the STOCSY approach.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXmsVSqsg%253D%253D&md5=6095939b18fae599b0090820561e894c
  25. 25
    Posma, J. M.; Garcia-Perez, I.; De Iorio, M.; Lindon, J. C.; Elliott, P.; Holmes, E.; Ebbels, T. M. D.; Nicholson, J. K. Subset optimization by reference matching (STORM): an optimized statistical approach for recovery of metabolic biomarker structural information from 1H NMR spectra of biofluids Anal. Chem. 2012, 84 (24) 10694– 10701 DOI: 10.1021/ac302360v
    Google Scholar
    25
    Subset Optimization by Reference Matching (STORM): An Optimized Statistical Approach for Recovery of Metabolic Biomarker Structural Information from 1H NMR Spectra of Biofluids
    Posma, Joram M.; Garcia-Perez, Isabel; De Iorio, Maria; Lindon, John C.; Elliott, Paul; Holmes, Elaine; Ebbels, Timothy M. D.; Nicholson, Jeremy K.
    Analytical Chemistry (Washington, DC, United States) (2012), 84 (24), 10694-10701CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    We describe a new multivariate statistical approach to recover metabolite structure information from multiple 1H NMR spectra in population sample sets. Subset optimization by ref. matching (STORM) was developed to select subsets of 1H NMR spectra that contain specific spectroscopic signatures of biomarkers differentiating between different human populations. STORM aims to improve the visualization of structural correlations in spectroscopic data by using these reduced spectral subsets contg. smaller nos. of samples than the no. of variables (n « p). We have used statistical shrinkage to limit the no. of false pos. assocns. and to simplify the overall interpretation of the autocorrelation matrix. The STORM approach has been applied to findings from an ongoing human metabolome-wide assocn. study on body mass index to identify a biomarker metabolite present in a subset of the population. Moreover, we have shown how STORM improves the visualization of more abundant NMR peaks compared to a previously published method (statistical total correlation spectroscopy, STOCSY). STORM is a useful new tool for biomarker discovery in the 'omic' sciences that has widespread applicability. It can be applied to any type of data, provided that there is interpretable correlation among variables, and can also be applied to data with more than one dimension (e.g., 2D NMR spectra).
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ons7zF&md5=0fbd04f3eac9cf7b475e4724359b9cad
  26. 26
    Navratil, V.; Pontoizeau, C.; Billoir, E.; Blaise, B. J. SRV: an open-source toolbox to accelerate the recovery of metabolic biomarkers and correlations from metabolic phenotyping datasets Bioinformatics 2013, 29 (10) 1348– 1349 DOI: 10.1093/bioinformatics/btt136
    Google Scholar
    There is no corresponding record for this reference.
  27. 27
    Nicholson, J. K.; Foxall, P. J.; Spraul, M.; Farrant, R. D.; Lindon, J. C. 750 MHz 1H and 1H-13C NMR spectroscopy of human blood plasma Anal. Chem. 1995, 67 (5) 793– 811 DOI: 10.1021/ac00101a004
    Google Scholar
    27
    750 MHz 1H and 1H-13C NMR Spectroscopy of Human Blood Plasma
    Nicholson, Jeremy K.; Foxall, Peta J. D.; Spraul, Manfred; Farrant, R. Duncan; Lindon, John C.
    Analytical Chemistry (1995), 67 (5), 793-811CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    High-resoln. 750 MHz 1H NMR spectra of control human blood plasma have been measured and assigned by the concerted use of a range of spin-echo, two-dimensional J-resolved, and homonuclear and heteronuclear (1H-13C) correlation methods. The increased spectral dispersion and sensitivity at 750 MHz enable the assignment of numerous 1H and 13C resonances from many mol. species that cannot be detected at lower frequencies. This work presents the most comprehensive assignment of the 1H NMR spectra of blood plasma yet achieved and includes the assignment of signals from 43 low Mr metabolites, including many with complex or strongly coupled spin systems. New assignments are also provided from the 1H and 13C NMR signals from several important macromol. species in whole blood plasma, i.e., very-low-d., low-d., and high-d. lipoproteins, albumin, and α1-acid glycoprotein. The temp. dependence of the one-dimensional and spin-echo 750 MHz 1H NMR spectra of plasma was investigated over the range 292-310 K. The 1H NMR signals from the fatty acyl side chains of the lipoproteins increased substantially with temp. (hence also mol. mobility), with a disproportionate increase from lipids in low-d. lipoprotein. Two-dimensional 1H-13C heteronuclear multiple quantum coherence spectroscopy at 292 and 310 K allowed both the direct detection of cholesterol and choline species bound in high-d. lipoprotein and the assignment of their signals and confirmed the assignment of most of the lipoprotein resonances.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXjsFajurs%253D&md5=026f5d082dc75abb431f04c957231002
  28. 28
    Merrifield, C. A.; Lewis, M.; Claus, S. P.; Beckonert, O. P.; Dumas, M.-E.; Duncker, S.; Kochhar, S.; Rezzi, S.; Lindon, J. C.; Bailey, M. A metabolic system-wide characterisation of the pig: a model for human physiology Mol. BioSyst. 2011, 7 (9) 2577– 2588 DOI: 10.1039/c1mb05023k
    Google Scholar
    28
    A metabolic system-wide characterisation of the pig: a model for human physiology
    Merrifield, Claire A.; Lewis, Marie; Claus, Sandrine P.; Beckonert, Olaf P.; Dumas, Marc-Emmanuel; Duncker, Swantje; Kochhar, Sunil; Rezzi, Serge; Lindon, John C.; Bailey, Mick; Holmes, Elaine; Nicholson, Jeremy K.
    Molecular BioSystems (2011), 7 (9), 2577-2588CODEN: MBOIBW; ISSN:1742-2051. (Royal Society of Chemistry)
    The pig is a single-stomached omnivorous mammal and is an important model of human disease and nutrition. As such, it is necessary to establish a metabolic framework from which pathol.-based variation can be compared. Here, a combination of one and two-dimensional 1H and 13C NMR spectroscopy (NMR) and high-resoln. magic angle spinning (HR-MAS) NMR was used to provide a systems overview of porcine metab. via characterization of the urine, serum, liver and kidney metabolomes. The metabolites obsd. in each of these biol. compartments were found to be qual. comparable to the metabolic signature of the same biol. matrixes in humans and rodents. The data were modelled using a combination of principal components anal. and Venn diagram mapping. Urine represented the most metabolically distinct biol. compartment studied, with a relatively greater no. of NMR detectable metabolites present, many of which are implicated in gut-microbial co-metabolic processes. The major inter-species differences obsd. were in the phase II conjugation of extra-genomic metabolites; the pig was obsd. to conjugate p-cresol, a gut microbial metabolite of tyrosine, with glucuronide rather than sulfate as seen in man. These observations are important to note when considering the translatability of exptl. data derived from porcine models.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhtVSmtL%252FM&md5=6784be376df517f8b75c6449cba9e9f8
  29. 29
    Wishart, D. S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A. C.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S. HMDB: the Human Metabolome Database Nucleic Acids Res. 2007, 35 (Database) D521– D526 DOI: 10.1093/nar/gkl923
    Google Scholar
    29
    HMDB: the Human Metabolome Database
    Wishart, David S.; Tzur, Dan; Knox, Craig; Eisner, Roman; Guo, An Chi; Young, Nelson; Cheng, Dean; Jewell, Kevin; Arndt, David; Sawhney, Summit; Fung, Chris; Nikolai, Lisa; Lewis, Mike; Coutouly, Marie-Aude; Forsythe, Ian; Tang, Peter; Shrivastava, Savita; Jeroncic, Kevin; Stothard, Paul; Amegbey, Godwin; Block, David; Hau, David. D.; Wagner, James; Miniaci, Jessica; Clements, Melisa; Gebremedhin, Mulu; Guo, Natalie; Zhang, Ying; Duggan, Gavin E.; MacInnis, Glen D.; Weljie, Alim M.; Dowlatabadi, Reza; Bamforth, Fiona; Clive, Derrick; Greiner, Russ; Li, Liang; Marrie, Tom; Sykes, Brian D.; Vogel, Hans J.; Querengesser, Lori
    Nucleic Acids Research (2007), 35 (Database Iss), D521-D526CODEN: NARHAD; ISSN:0305-1048. (Oxford University Press)
    The Human Metabolome Database (HMDB) is currently the most complete and comprehensive curated collection of human metabolite and human metab. data in the world. It contains records for more than 2180 endogenous metabolites with information gathered from thousands of books, journal articles and electronic databases. In addn. to its comprehensive literature-derived data, the HMDB also contains an extensive collection of exptl. metabolite concn. data compiled from hundreds of mass spectra (MS) and NMR metabolomic analyses performed on urine, blood and cerebrospinal fluid samples. This is further supplemented with thousands of NMR and MS spectra collected on purified, ref. metabolites. Each metabolite entry in the HMDB contains an av. of 90 sep. data fields including a comprehensive compd. description, names and synonyms, structural information, physico-chem. data, ref. NMR and MS spectra, biofluid concns., disease assocns., pathway information, enzyme data, gene sequence data, SNP and mutation data as well as extensive links to images, refs. and other public databases. Extensive searching, relational querying and data browsing tools are also provided. The HMDB is designed to address the broad needs of biochemists, clin. chemists, physicians, medical geneticists, nutritionists and members of the metabolomics community.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXivFKhtA%253D%253D&md5=adb0346dd43ced1de7154c4c8d2c57d4
  30. 30
    Ulrich, E. L.; Akutsu, H.; Doreleijers, J. F.; Harano, Y.; Ioannidis, Y. E.; Lin, J.; Livny, M.; Mading, S.; Maziuk, D.; Miller, Z. BioMagResBank Nucleic Acids Res. 2008, 36 (Database) D402– D408 DOI: 10.1093/nar/gkm957
    Google Scholar
    30
    BioMagResBank
    Ulrich, Eldon L.; Akutsu, Hideo; Doreleijers, Jurgen F.; Harano, Yoko; Ioannidis, Yannis E.; Lin, Jundong; Livny, Miron; Mading, Steve; Maziuk, Dimitri; Miller, Zachary; Nakatani, Eiichi; Schulte, Christopher F.; Tolmie, David E.; Kent Wenger, R.; Yao, Hongyang; Markley, John L.
    Nucleic Acids Research (2008), 36 (Database Iss), D402-D408CODEN: NARHAD; ISSN:0305-1048. (Oxford University Press)
    The BioMagResBank (BMRB: www.bmrb.wisc.edu) is a repository for exptl. and derived data gathered from NMR (NMR) spectroscopic studies of biol. mols. BMRB is a partner in the Worldwide Protein Data Bank (wwPDB). The BMRB archive consists of four main data depositories: (i) quant. NMR spectral parameters for proteins, peptides, nucleic acids, carbohydrates and ligands or cofactors (assigned chem. shifts, coupling consts. and peak lists) and derived data (relaxation parameters, residual dipolar couplings, hydrogen exchange rates, pKa values, etc.), (ii) databases for NMR restraints processed from original author depositions available from the Protein Data Bank, (iii) time-domain (raw) spectral data from NMR expts. used to assign spectral resonances and det. the structures of biol. macromols. and (iv) a database of one- and two-dimensional 1H and 13C one- and two-dimensional NMR spectra for over 250 metabolites. The BMRB website provides free access to all of these data. BMRB has tools for querying the archive and retrieving information and an ftp site (ftp.bmrb.wisc.edu) where data in the archive can be downloaded in bulk. Two BMRB mirror sites exist: one at the PDBj, Protein Research Institute, Osaka University, Osaka, Japan (bmrb.protein.osaka-u.ac.jp) and the other at CERM, University of Florence, Florence, Italy (bmrb.postgenomicnmr.net/). The site at Osaka also accepts and processes data depositions.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtVWitb0%253D&md5=520c47beb1bc57e57ef14fac08a6fd3d
  31. 31
    Sumner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W.-M.; Fiehn, O.; Goodacre, R.; Griffin, J. L. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI) Metabolomics 2007, 3 (3) 211– 221 DOI: 10.1007/s11306-007-0082-2
    Google Scholar
    31
    Proposed minimum reporting standards for chemical analysis. Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI)
    Sumner, Lloyd W.; Amberg, Alexander; Barrett, Dave; Beale, Michael H.; Beger, Richard; Daykin, Clare A.; Fan, Teresa W.-M.; Fiehn, Oliver; Goodacre, Royston; Griffin, Julian L.; Hankemeier, Thomas; Hardy, Nigel; Harnly, James; Higashi, Richard; Kopka, Joachim; Lane, Andrew N.; Lindon, John C.; Marriott, Philip; Nicholls, Andrew W.; Reily, Michael D.; Thaden, John J.; Viant, Mark R.
    Metabolomics (2007), 3 (3), 211-221CODEN: METAHQ; ISSN:1573-3882. (Springer)
    There is a general consensus that supports the need for standardized reporting of metadata or information describing large-scale metabolomics and other functional genomics data sets. Reporting of std. metadata provides a biol. and empirical context for the data, facilitates exptl. replication, and enables the reinterrogation and comparison of data by others. Accordingly, the Metabolomics Stds. Initiative is building a general consensus concerning the min. reporting stds. for metabolomics expts. of which the Chem. Anal. Working Group (CAWG) is a member of this community effort. This article proposes the min. reporting stds. related to the chem. anal. aspects of metabolomics expts. including: sample prepn., exptl. anal., quality control, metabolite identification, and data pre-processing. These min. stds. currently focus mostly upon mass spectrometry and NMR spectroscopy due to the popularity of these techniques in metabolomics. However, addnl. input concerning other techniques is welcomed and can be provided via the CAWG online discussion forum at http://msi-workgroups.sourceforge.net/ or http://[email protected]. Further, community input related to this document can also be provided via this electronic forum.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhtFOntbvL&md5=e636dad276f9f1b2e756db169368eb53
  32. 32
    Altmaier, E.; Fobo, G.; Heier, M.; Thorand, B.; Meisinger, C.; Römisch-Margl, W.; Waldenberger, M.; Gieger, C.; Illig, T.; Adamski, J. Metabolomics approach reveals effects of antihypertensives and lipid-lowering drugs on the human metabolism Eur. J. Epidemiol. 2014, 29 (5) 325– 336 DOI: 10.1007/s10654-014-9910-7
    Google Scholar
    32
    Metabolomics approach reveals effects of antihypertensives and lipid-lowering drugs on the human metabolism
    Altmaier, Elisabeth; Fobo, Gisela; Heier, Margit; Thorand, Barbara; Meisinger, Christine; Roemisch-Margl, Werner; Waldenberger, Melanie; Gieger, Christian; Illig, Thomas; Adamski, Jerzy; Suhre, Karsten; Kastenmueller, Gabi
    European Journal of Epidemiology (2014), 29 (5), 325-336CODEN: EJEPE8; ISSN:0393-2990. (Springer)
    The mechanism of antihypertensive and lipid-lowering drugs on the human organism is still not fully understood. New insights on the drugs' action can be provided by a metabolomics-driven approach, which offers a detailed view of the physiol. state of an organism. Here, we report a metabolome-wide assocn. study with 295 metabolites in human serum from 1,762 participants of the KORA F4 (Cooperative Health Research in the Region of Augsburg) study population. Our intent was to find variations of metabolite concns. related to the intake of various drug classes and-based on the assocns. found-to generate new hypotheses about on-target as well as off-target effects of these drugs. In total, we found 41 significant assocns. for the drug classes investigated: For beta-blockers (11 assocns.), angiotensin-converting enzyme (ACE) inhibitors (four assoc.), diuretics (seven assoc.), statins (ten assoc.), and fibrates (nine assoc.) the top hits were pyroglutamine, phenylalanylphenylalanine, pseudouridine, 1-arachidonoylglycerophosphocholine, and 2-hydroxyisobutyrate, resp. For beta-blockers we obsd. significant assocns. with metabolite concns. that are indicative of drug side-effects, such as increased serotonin and decreased free fatty acid levels. Intake of ACE inhibitors and statins assocd. with metabolites that provide insight into the action of the drug itself on its target, such as an assocn. of ACE inhibitors with des-Arg(9)-bradykinin and aspartylphenylalanine, a substrate and a product of the drug-inhibited ACE. The intake of statins which reduce blood cholesterol levels, resulted in changes in the concn. of metabolites of the biosynthesis as well as of the degrdn. of cholesterol. Fibrates showed the strongest assocn. with 2-hydroxyisobutyrate which might be a breakdown product of fenofibrate and, thus, a possible marker for the degrdn. of this drug in the human organism. The anal. of diuretics showed a heterogeneous picture that is difficult to interpret. Taken together, our results provide a basis for a deeper functional understanding of the action and side-effects of antihypertensive and lipid-lowering drugs in the general population.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXnslemt78%253D&md5=958e76f48eb59d5a81ffdf6a5428497d
  33. 33
    Sekula, P.; Goek, O.-N.; Quaye, L.; Barrios, C.; Levey, A. S.; Römisch-Margl, W.; Menni, C.; Yet, I.; Gieger, C.; Inker, L. A. A Metabolome-Wide Association Study of Kidney Function and Disease in the General Population J. Am. Soc. Nephrol. 2016, 27 (4) 1175– 1188 DOI: 10.1681/ASN.2014111099
    Google Scholar
    33
    A metabolome-wide association study of kidney function and disease in the general population
    Sekula, Peggy; Goek, Oemer-Necmi; Quaye, Lydia; Barrios, Clara; Levey, Andrew S.; Roemisch-Margl, Werner; Menni, Cristina; Yet, Idil; Gieger, Christian; Inker, Lesley A.; Adamski, Jerzy; Gronwald, Wolfram; Illig, Thomas; Dettmer, Katja; Krumsiek, Jan; Oefner, Peter J.; Valdes, Ana M.; Meisinger, Christa; Coresh, Josef; Spector, Tim D.; Mohney, Robert P.; Suhre, Karsten; Kastenmueller, Gabi; Koettgen, Anna
    Journal of the American Society of Nephrology (2016), 27 (4), 1175-1189CODEN: JASNEU; ISSN:1046-6673. (American Society of Nephrology)
    Small mols. are extensively metabolized and cleared by the kidney. Changes in serum metabolite concns. may result from impaired kidney function and can be used to est. filtration (e.g., the established marker creatinine) or may precede and potentially contribute to CKD development. Here, we applied a nontargeted metabolomics approach using gas and liq. chromatog. coupled to mass spectrometry to quantify 493 small mols. in human serum. The assocns. of these mols. with GFR estd. on the basis of creatinine (eGFRcr) and cystatin C levels were assessed in ≤1735 participants in the KORA F4 study, followed by replication in 1164 individuals in the TwinsUK registry. After correction for multiple testing, 54 replicated metabolites significantly assocd. with eGFRcr, and six of these showed pairwise correlation (r≥0.50) with established kidney function measures: C-mannosyltryptophan, pseudouridine, N-acetylalanine, erythronate, myo-inositol, and N-acetylcarnosine. Higher C-mannosyltryptophan, pseudouridine, and O-sulfo-L-tyrosine concns. assocd. with incident CKD (eGFRcr <60 mL/min per 1.73 m2) in the KORA F4 study. In contrast with serum creatinine, C-mannosyltryptophan and pseudouridine concns. showed little dependence on sex. Furthermore, correlation with measured GFR in 200 participants in the AASK study was 0.78 for both C-mannosyltryptophan and pseudouridine concn., and highly significant assocns. of both metabolites with incident ESRD disappeared upon adjustment for measured GFR. Thus, these mols. may be alternative or complementary markers of kidney function. In conclusion, our study provides a comprehensive list of kidney function-assocd. metabolites and highlights potential novel filtration markers that may help to improve the estn. of GFR.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XitFSjtLzM&md5=7b34021440ef1938afc8538539bd70a3
  34. 34
    Adkins, D. E.; McClay, J. L.; Vunck, S. A.; Batman, A. M.; Vann, R. E.; Clark, S. L.; Souza, R. P.; Crowley, J. J.; Sullivan, P. F.; van den Oord, E. J. C. G. Behavioral metabolomics analysis identifies novel neurochemical signatures in methamphetamine sensitization Genes Brain Behav. 2013, 12 (8) 780– 791 DOI: 10.1111/gbb.12081
    Google Scholar
    34
    Behavioral metabolomics analysis identifies novel neurochemical signatures in methamphetamine sensitization
    Adkins, D. E.; McClay, J. L.; Vunck, S. A.; Batman, A. M.; Vann, R. E.; Clark, S. L.; Souza, R. P.; Crowley, J. J.; Sullivan, P. F.; van den Oord, E. J. C. G.; Beardsley, P. M.
    Genes, Brain and Behavior (2013), 12 (8), 780-791CODEN: GBBEAO; ISSN:1601-1848. (Wiley-Blackwell)
    Behavioral sensitization has been widely studied in animal models and is theorized to reflect neural modifications assocd. with human psychostimulant addiction. While the mesolimbic dopaminergic pathway is known to play a role, the neurochem. mechanisms underlying behavioral sensitization remain incompletely understood. In this study, we conducted the first metabolomics anal. to globally characterize neurochem. differences assocd. with behavioral sensitization. Methamphetamine (MA)-induced sensitization measures were generated by statistically modeling longitudinal activity data for eight inbred strains of mice. Subsequent to behavioral testing, nontargeted liq. and gas chromatog.-mass spectrometry profiling was performed on 48 brain samples, yielding 301 metabolite levels per sample after quality control. Assocn. testing between metabolite levels and three primary dimensions of behavioral sensitization (total distance, stereotypy and margin time) showed four robust, significant assocns. at a stringent metabolome-wide significance threshold (false discovery rate, FDR <0.05). Results implicated homocarnosine, a dipeptide of GABA and histidine, in total distance sensitization, GABA metabolite 4-guanidinobutanoate and pantothenate in stereotypy sensitization, and myo-inositol in margin time sensitization. Secondary analyses indicated that these assocns. were independent of concurrent MA levels and, with the exception of the myo-inositol assocn., suggest a mechanism whereby strain-based genetic variation produces specific baseline neurochem. differences that substantially influence the magnitude of MA-induced sensitization. These findings demonstrate the utility of mouse metabolomics for identifying novel biomarkers, and developing more comprehensive neurochem. models, of psychostimulant sensitization.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhslWqsL3E&md5=d01fd2ace97b8309c5e984c83dc35133
  35. 35
    Patterson, N.; Price, A. L.; Reich, D. Population structure and eigenanalysis PLoS Genet. 2006, 2 (12) e190 DOI: 10.1371/journal.pgen.0020190
    Google Scholar
    There is no corresponding record for this reference.
  36. 36
    Schäfer, J.; Strimmer, K. A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics Stat. Appl. Genet. Mol. Biol. 2005, 4 (1) 1– 32 DOI: 10.2202/1544-6115.1175
    Google Scholar
    There is no corresponding record for this reference.
  37. 37
    Auro, K.; Joensuu, A.; Fischer, K.; Kettunen, J.; Salo, P.; Mattsson, H.; Niironen, M.; Kaprio, J.; Eriksson, J. G.; Lehtimäki, T. A metabolic view on menopause and ageing Nat. Commun. 2014, 5, 4708 DOI: 10.1038/ncomms5708
    Google Scholar
    37
    A metabolic view on menopause and ageing
    Auro, Kirsi; Joensuu, Anni; Fischer, Krista; Kettunen, Johannes; Salo, Perttu; Mattsson, Hannele; Niironen, Marjo; Kaprio, Jaakko; Eriksson, Johan G.; Lehtimaki, Terho; Raitakari, Olli; Jula, Antti; Tiitinen, Aila; Jauhiainen, Matti; Soininen, Pasi; Kangas, Antti J.; Kahonen, Mika; Havulinna, Aki S.; Ala-Korpela, Mika; Salomaa, Veikko; Metspalu, Andres; Perola, Markus
    Nature Communications (2014), 5 (), 4708CODEN: NCAOBW; ISSN:2041-1723. (Nature Publishing Group)
    The ageing of the global population calls for a better understanding of age-related metabolic consequences. Here we report the effects of age, sex and menopause on serum metabolites in 26,065 individuals of Northern European ancestry. Age-specific metabolic fingerprints differ significantly by gender and, in females, a substantial atherogenic shift overlapping the time of menopausal transition is obsd. In meta-anal. of 10,083 women, menopause status assocs. with amino acids glutamine, tyrosine and isoleucine, along with serum cholesterol measures and atherogenic lipoproteins. Among 3,204 women aged 40-55 years, menopause status assocs. addnl. with glycine and total, monounsatd., and omega-7 and -9 fatty acids. Our findings suggest that, in addn. to lipid alterations, menopause may contribute to future metabolic and cardiovascular risk via influencing amino-acid concns., adding to the growing evidence of the importance of amino acids in metabolic disease progression. These observations shed light on the metabolic consequences of ageing, gender and menopause at the population level.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXksVCjsb4%253D&md5=0ea518a01a3da198fdced0c89f3ae664
  38. 38
    R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2014; Vol. 73 (1), pp 3– 36.
    Google Scholar
    There is no corresponding record for this reference.
  39. 39
    Benjamini, Y.; Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing J. R. Stat. Soc. Ser. B Methodol. 1995, 57 (1) 289– 300
    Google Scholar
    There is no corresponding record for this reference.
  40. 40
    Zhang, J.; Coombes, K. R. Sources of variation in false discovery rate estimation include sample size, correlation, and inherent differences between groups BMC Bioinf. 2012, 13 (Suppl 13) S1 DOI: 10.1186/1471-2105-13-S13-S1
    Google Scholar
    There is no corresponding record for this reference.

Cited By

ARTICLE SECTIONS
Jump To

This article is cited by 18 publications.

  1. Felix O’Farrell, Benjamin Aleyakpo, Rima Mustafa, Xiyun Jiang, Rui Climaco Pinto, Paul Elliott, Ioanna Tzoulaki, Abbas Dehghan, Samantha H. Y. Loh, Jeff W. Barclay, L. Miguel Martins, Raha Pazoki. Evidence for involvement of the alcohol consumption WDPCP gene in lipid metabolism, and liver cirrhosis. Scientific Reports 2023, 13 (1) https://doi.org/10.1038/s41598-023-47371-7
  2. Alexis C. Wood, Mark O. Goodarzi, Mackenzie K. Senn, Meghana D. Gadgil, Goncalo Graca, Matthew A. Allison, Ioanna Tzoulaki, Michael Y. Mi, Philip Greenland, Timothy Ebbels, Paul Elliott, Russell P. Tracy, David M. Herrington, Jerome I. Rotter. Associations between Metabolomic Biomarkers of Avocado Intake and Glycemia in the Multi-Ethnic Study of Atherosclerosis. The Journal of Nutrition 2023, 153 (10) , 2797-2807. https://doi.org/10.1016/j.tjnut.2023.07.013
  3. Olga I. Kiseleva, Ilya Y. Kurbatov, Viktoriia A. Arzumanian, Ekaterina V. Ilgisonis, Svyatoslav V. Zakharov, Ekaterina V. Poverennaya. The Expectation and Reality of the HepG2 Core Metabolic Profile. Metabolites 2023, 13 (8) , 908. https://doi.org/10.3390/metabo13080908
  4. Ruoyu Yang, Yi Wang, Chunhua Yuan, Xunzhang Shen, Ming Cai, Liyan Wang, Jingyun Hu, Haihan Song, Hongbiao Wang, Lei Zhang. The combined analysis of urine and blood metabolomics profiles provides an accurate prediction of the training and competitive status of Chinese professional swimmers. Frontiers in Physiology 2023, 14 https://doi.org/10.3389/fphys.2023.1197224
  5. Monica Emili Garcia‐Segura, Brenan R. Durainayagam, Sonia Liggi, Gonçalo Graça, Beatriz Jimenez, Abbas Dehghan, Ioanna Tzoulaki, Ibrahim Karaman, Paul Elliott, Julian L. Griffin. Pathway‐based integration of multi‐omics data reveals lipidomics alterations validated in an Alzheimer's disease mouse model and risk loci carriers. Journal of Neurochemistry 2023, 164 (1) , 57-76. https://doi.org/10.1111/jnc.15719
  6. Abbas Dehghan, Rui Climaco Pinto, Ibrahim Karaman, Jian Huang, Brenan R. Durainayagam, Mohsen Ghanbari, Areesha Nazeer, Qi Zhong, Sonia Liggi, Luke Whiley, Rima Mustafa, Miia Kivipelto, Alina Solomon, Tiia Ngandu, Takahisa Kanekiyo, Tomonori Aikawa, Carola I. Radulescu, Samuel J. Barnes, Gonçalo Graça, Elena Chekmeneva, Stephane Camuzeaux, Matthew R. Lewis, Manuja R. Kaluarachchi, M. Arfan Ikram, Elaine Holmes, Ioanna Tzoulaki, Paul M. Matthews, Julian L. Griffin, Paul Elliott. Metabolome-wide association study on ABCA7 indicates a role of ceramide metabolism in Alzheimer’s disease. Proceedings of the National Academy of Sciences 2022, 119 (43) https://doi.org/10.1073/pnas.2206083119
  7. Alina Peluso, Robert Glen, Timothy M. D. Ebbels. Multiple-testing correction in metabolome-wide association studies. BMC Bioinformatics 2021, 22 (1) https://doi.org/10.1186/s12859-021-03975-2
  8. Aikaterini Iliou, Emmanuel Mikros, Ibrahim Karaman, Freya Elliott, Julian L Griffin, Ioanna Tzoulaki, Paul Elliott. Metabolic phenotyping and cardiovascular disease: an overview of evidence from epidemiological settings. Heart 2021, 107 (14) , 1123-1129. https://doi.org/10.1136/heartjnl-2019-315615
  9. Beatriz Calvo-Serra, Léa Maitre, Chung-Ho E Lau, Alexandros P Siskos, Kristine B Gützkow, Sandra Andrušaitytė, Maribel Casas, Solène Cadiou, Leda Chatzi, Juan R González, Regina Grazuleviciene, Rosemary McEachan, Rémy Slama, Marina Vafeiadi, John Wright, Murieann Coen, Martine Vrijheid, Hector C Keun, Geòrgia Escaramís, Mariona Bustamante. Urinary metabolite quantitative trait loci in children and their interaction with dietary factors. Human Molecular Genetics 2021, 29 (23) , 3830-3844. https://doi.org/10.1093/hmg/ddaa257
  10. Sonia Dagnino, Barbara Bodinier, Hasmik Grigoryan, Stephen M. Rappaport, Maryam Karimi, Florence Guida, Silvia Polidoro, WIlliam B. Edmands, Alessio Naccarati, Giovanni Fiorito, Carlotta Sacerdote, Vittorio Krogh, Roel Vermeulen, Paolo Vineis, Marc Chadeau‐Hyam. Agnostic Cys34‐albumin adductomics and DNA methylation: Implication of N‐acetylcysteine in lung carcinogenesis years before diagnosis. International Journal of Cancer 2020, 146 (12) , 3294-3303. https://doi.org/10.1002/ijc.32680
  11. Wimal Pathmasiri, Kristine Kay, Susan McRitchie, Susan Sumner. Analysis of NMR Metabolomics Data. 2020, 61-97. https://doi.org/10.1007/978-1-0716-0239-3_5
  12. Ioanna Tzoulaki, Raphaële Castagné, Claire L Boulangé, Ibrahim Karaman, Elena Chekmeneva, Evangelos Evangelou, Timothy M D Ebbels, Manuja R Kaluarachchi, Marc Chadeau-Hyam, David Mosen, Abbas Dehghan, Alireza Moayyeri, Diana L Santos Ferreira, Xiuqing Guo, Jerome I Rotter, Kent D Taylor, Maryam Kavousi, Paul S de Vries, Benjamin Lehne, Marie Loh, Albert Hofman, Jeremy K Nicholson, John Chambers, Christian Gieger, Elaine Holmes, Russell Tracy, Jaspal Kooner, Philip Greenland, Oscar H Franco, David Herrington, John C Lindon, Paul Elliott. Serum metabolic signatures of coronary and carotid atherosclerosis and subsequent cardiovascular disease. European Heart Journal 2019, 40 (34) , 2883-2896. https://doi.org/10.1093/eurheartj/ehz235
  13. Megan M. Niedzwiecki, Douglas I. Walker, Roel Vermeulen, Marc Chadeau-Hyam, Dean P. Jones, Gary W. Miller. The Exposome: Molecules to Populations. Annual Review of Pharmacology and Toxicology 2019, 59 (1) , 107-127. https://doi.org/10.1146/annurev-pharmtox-010818-021315
  14. Lorraine Brennan, Frank B. Hu. Metabolomics‐Based Dietary Biomarkers in Nutritional Epidemiology—Current Status and Future Opportunities. Molecular Nutrition & Food Research 2019, 63 (1) https://doi.org/10.1002/mnfr.201701064
  15. Timothy M. D. Ebbels, Ibrahim Karaman, Gonçalo Graça. Processing and Analysis of Untargeted Multicohort NMR Data. 2019, 453-470. https://doi.org/10.1007/978-1-4939-9690-2_25
  16. Marc Chadeau-Hyam, Roel Vermeulen. Statistical Models to Explore the Exposome: From OMICs Profiling to ‘Mechanome’ Characterization. 2019, 279-314. https://doi.org/10.1007/978-3-319-89321-1_11
  17. Raphaële Castagné, Marc Chadeau-Hyam. Univariate Statistical Modeling, Multiple Testing Correction, and Visualization in Metabolome-Wide Association Studies. 2019, 237-260. https://doi.org/10.1016/B978-0-12-812293-8.00008-6
  18. Nicholas J. W. Rattray, Nicole C. Deziel, Joshua D. Wallach, Sajid A. Khan, Vasilis Vasiliou, John P. A. Ioannidis, Caroline H. Johnson. Beyond genomics: understanding exposotypes through metabolomics. Human Genomics 2018, 12 (1) https://doi.org/10.1186/s40246-018-0134-x
Download PDF

  • Abstract

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 1

    Figure 1. Analytical workflow.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 2

    Figure 2. Metabolome wide study of glucose (model 2). This Manhattan plot shows the analysis of the 30 590 CPMG features. The signed negative log10 p-value is plotted against the chemical shift in ppm. To ease the visualization, all log p-value ≤ 10–30 were set to 1 × 10–30. The horizontal dashed line indicates the α′ per-test significance level controlling the FWER at a 5% level using the Gaussian simulated outcome. Data points are colored by class of metabolites. Components were: 1, L1; 2, L2; 3, isoleucine; 4, leucine, isoleucine; 5, leucine; 6, valine; 7, L3; 8, lactate; 9, alanine; 10, L4; 11, arginine; 12, lysine; 13, acetate; 14, L5; 15, acetylglycoproteins; 16, methionine; 17, glutamate; 18, glutamine; 19, L6; 20, 3-hydroxybutyrate; 21, pyruvate; 22, pyroglutamate; 23, citrate; 24, L7; 25, aspartate; 26, albumin; 27, creatine; 28, creatinine; 29, ornithine, tyrosine; 30, ornithine; 31, phenylalanine; 32, tyrosine; 33, choline; 34, beta-glucose; 35, proline; 36, alpha-glucose, beta-glucose; 37, alpha-glucose; 38, glycine; 39, glycerol; 40, mannose; 41, glyceryl groups of lipids; 42, APAP glucuronide; 43, L8; 44, uridine; 45, 1-Methylhistidine; 46, histidine; 47, 3-methylhistidine; 48, formate.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 3

    Figure 3. CPMG-model 2 regional association plots with log10 (glucose) for the glutamine. In the upper plot, the −log10 p-value for the features at two regions are shown on each plot. Features are colored based on their correlation with the gray hit that has the smallest p-value in the region. The lines show the mean corrected intensity (i.e., residuals removing the linear effect of the phase and the cohort) in the 5% of samples with high residual glucose in green and 5% of the samples with low residual glucose in blue. The bottom plot shows the mean spectral intensity in MESA phase 1 (plain line) and in MESA phase 2 (dashed line). Green circles indicate the proportion of replication after results prioritization.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 4

    Figure 4. Comparison of results from the analysis in MESA to those from the 80:20 split strategy. Results are presented for the CPMG (N = 30 590 data points) metabolome wide association study of glucose using model 2. To ease the visualization, all p-value ≤ 10–30 were set to 1 × 10–30. The −log10(p-value) is signed by the direction of the effect size estimate and is plotted against the chemical shift. The horizontal dashed line indicates the per-test significance level controlling the FWER at a 5% level. Variables found from the analyses in MESA are presented in black, those discovered and replicated at least once across the 100 splits are presented in green and those discovered and replicated in 50% of the split are presented in red. Components were: 1, L1; 2, L2; 3, isoleucine; 4, leucine, isoleucine; 5, leucine; 6, valine; 7, L3; 8, lactate; 9, alanine; 10, L4; 11, arginine; 12, lysine; 13, acetate; 14, L5; 15, acetylglycoproteins; 16, methionine; 17, glutamate; 18, glutamine; 19, L6; 20, 3-hydroxybutyrate; 21, pyruvate; 22, pyroglutamate; 23, citrate; 24, L7; 25, aspartate; 26, albumin; 27, creatine; 28, creatinine; 29, ornithine, tyrosine; 30, ornithine; 31, phenylalanine; 32, tyrosine; 33, choline; 34, beta-glucose; 35, proline; 36, alpha-glucose, beta-glucose; 37, alpha-glucose; 38, glycine; 39, glycerol; 40, mannose; 41, glyceryl groups of lipids; 42, APAP glucuronide; 43, L8; 44, uridine; 45, 1-methylhistidine; 46, histidine; 47, 3-methylhistidine; 48, formate.

    High Resolution Image
    Download MS PowerPoint Slide

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 40 other publications.

    1. 1
      Lenz, E. M.; Wilson, I. D. Analytical strategies in metabonomics J. Proteome Res. 2007, 6 (2) 443– 458 DOI: 10.1021/pr0605217
      Google Scholar
      1
      Analytical Strategies in Metabonomics
      Lenz, Eva Maria; Wilson, Ian D.
      Journal of Proteome Research (2007), 6 (2), 443-458CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
      A review. To perform metabonomics investigations, it is necessary to generate comprehensive metabolite profiles for complex samples such as biofluids and tissue/tissue exts. Anal. technologies that can be used to achieve this aim are constantly evolving, and new developments are changing the way in which such profiles' metabolite profiles can be generated. Here, the utility of various anal. techniques for global metabolite profiling, such as, e.g., 1H NMR, MS, HPLC-MS, and GC-MS, are explored and compared.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD28Xht12qt7%252FM&md5=b63fd8bf30a64a62ef035e0c8880608a
    2. 2
      Lindon, J. C.; Nicholson, J. K. Analytical technologies for metabonomics and metabolomics, and multi-omic information recovery TrAC, Trends Anal. Chem. 2008, 27 (3) 194– 204 DOI: 10.1016/j.trac.2007.08.009
      Google Scholar
      There is no corresponding record for this reference.
    3. 3
      Holmes, E.; Wilson, I. D.; Nicholson, J. K. Metabolic phenotyping in health and disease Cell 2008, 134 (5) 714– 717 DOI: 10.1016/j.cell.2008.08.026
      Google Scholar
      3
      Metabolic phenotyping in health and disease
      Holmes, Elaine; Wilson, Ian D.; Nicholson, Jeremy K.
      Cell (Cambridge, MA, United States) (2008), 134 (5), 714-717CODEN: CELLB5; ISSN:0092-8674. (Cell Press)
      A review. Analyzing metabolites (small mols. <1 kDa) in body fluids such as urine and plasma using various spectroscopic methods provides information on the metabotype (metabolic phenotype) of individuals or populations, information that can be applied to personalized medicine or public healthcare.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtFCqs7jE&md5=f82b72659258117a777e4834c7f3233f
    4. 4
      Madsen, R.; Lundstedt, T.; Trygg, J. Chemometrics in metabolomics--a review in human disease diagnosis Anal. Chim. Acta 2010, 659 (1–2) 23– 33 DOI: 10.1016/j.aca.2009.11.042
      Google Scholar
      4
      Chemometrics in metabolomics. A review in human disease diagnosis
      Madsen, Rasmus; Lundstedt, Torbjoern; Trygg, Johan
      Analytica Chimica Acta (2010), 659 (1-2), 23-33CODEN: ACACAM; ISSN:0003-2670. (Elsevier B.V.)
      A review. Metabolomics is a post genomic research field concerned with developing methods for anal. of low mol. wt. compds. in biol. systems, such as cells, organs or organisms. Analyzing metabolic differences between unperturbed and perturbed systems, such as healthy volunteers and patients with a disease, can lead to insights into the underlying pathol. In metabolomics anal., large amts. of data are routinely produced in order to characterize samples. The use of multivariate data anal. techniques and chemometrics is a commonly used strategy for obtaining reliable results. Metabolomics have been applied in different fields such as disease diagnosis, toxicol., plant science and pharmaceutical and environmental research. In this review we take a closer look at the chemometric methods used and the available results within the field of disease diagnosis. We will first present some current strategies for performing metabolomics studies, esp. regarding disease diagnosis. The main focus will be on data anal. strategies and validation of multivariate models, since there are many pitfalls in this regard. Further, we highlight the most interesting metabolomics publications and discuss these in detail; addnl. studies are mentioned as a ref. for the interested reader. A general trend is an increased focus on biol. interpretation rather than merely the ability to classify samples. In the conclusions, the general trends and some recommendations for improving metabolomics data anal. are provided.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXhs1WnsbrL&md5=7e95e755f785e361f6cb67647969e565
    5. 5
      Heather, L. C.; Wang, X.; West, J. A.; Griffin, J. L. A practical guide to metabolomic profiling as a discovery tool for human heart disease J. Mol. Cell. Cardiol. 2013, 55, 2– 11 DOI: 10.1016/j.yjmcc.2012.12.001
      Google Scholar
      5
      A practical guide to metabolomic profiling as a discovery tool for human heart disease
      Heather, Lisa C.; Wang, Xinzhu; West, James A.; Griffin, Julian L.
      Journal of Molecular and Cellular Cardiology (2013), 55 (), 2-11CODEN: JMCDAY; ISSN:0022-2828. (Elsevier B.V.)
      A review. Metabolomics has refreshed interest in metab. across biol. and medicine, particularly in the areas of functional genomics and biomarker discovery. In this review we will discuss the exptl. techniques and challenges involved in metabolomic profiling and how these technologies have been applied to cardiovascular disease. Open profiling and targeted approaches to metabolomics are compared, focusing on high resoln. NMR spectroscopy and mass spectrometry, as well as discussing how to analyze the large amts. of data generated using multivariate statistics. Finally, the current literature on metabolomic profiling in human cardiovascular disease is reviewed to illustrate the diversity of approaches, and discuss some of the key metabolites and pathways found to be perturbed in plasma, urine and tissue from patients with these diseases. This includes studies of coronary artery disease, myocardial infarction, and ischemic heart disease. These studies demonstrate the potential of the technol. for biomarker discovery and elucidating metabolic mechanisms assocd. with given pathologies, but also in some cases provide a warning of the pitfalls of poor study design. This article is part of a Special Issue entitled "Focus on Cardiac Metab.".
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXkt1SgsQ%253D%253D&md5=e8364a80fa036bb28291f64c0b7b50a3
    6. 6
      Tzoulaki, I.; Ebbels, T. M. D.; Valdes, A.; Elliott, P.; Ioannidis, J. P. A. Design and analysis of metabolomics studies in epidemiologic research: a primer on -omic technologies Am. J. Epidemiol. 2014, 180 (2) 129– 139 DOI: 10.1093/aje/kwu143
      Google Scholar
      6
      Design and analysis of metabolomics studies in epidemiologic research: a primer on -omic technologies
      Tzoulaki Ioanna; Ebbels Timothy M D; Valdes Ana; Elliott Paul; Ioannidis John P A
      American journal of epidemiology (2014), 180 (2), 129-39 ISSN:.
      Metabolomics is the field of "-omics" research concerned with the comprehensive characterization of the small low-molecular-weight metabolites in biological samples. In epidemiology, it represents an emerging technology and an unprecedented opportunity to measure environmental and other exposures with improved precision and far less measurement error than with standard epidemiologic methods. Advances in the application of metabolomics in large-scale epidemiologic research are now being realized through a combination of improved sample preparation and handling, automated laboratory and processing methods, and reduction in costs. The number of epidemiologic studies that use metabolic profiling is still limited, but it is fast gaining popularity in this area. In the present article, we present a roadmap for metabolomic analyses in epidemiologic studies and discuss the various challenges these data pose to large-scale studies. We discuss the steps of data preprocessing, univariate and multivariate data analysis, correction for multiplicity of comparisons with correlated data, and finally the steps of cross-validation and external validation. As data from metabolomic studies accumulate in epidemiology, there is a need for large-scale replication and synthesis of findings, increased availability of raw data, and a focus on good study design, all of which will highlight the potential clinical impact of metabolomics in this field.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC2cfmslWlsA%253D%253D&md5=c186d66aeef1ec835bbc1d92cf0046fa
    7. 7
      Dumas, M.-E.; Maibaum, E. C.; Teague, C.; Ueshima, H.; Zhou, B.; Lindon, J. C.; Nicholson, J. K.; Stamler, J.; Elliott, P.; Chan, Q. Assessment of analytical reproducibility of 1H NMR spectroscopy based metabonomics for large-scale epidemiological research: the INTERMAP Study Anal. Chem. 2006, 78 (7) 2199– 2208 DOI: 10.1021/ac0517085
      Google Scholar
      There is no corresponding record for this reference.
    8. 8
      Dona, A. C.; Jiménez, B.; Schäfer, H.; Humpfer, E.; Spraul, M.; Lewis, M. R.; Pearce, J. T. M.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. Precision high-throughput proton NMR spectroscopy of human urine, serum, and plasma for large-scale metabolic phenotyping Anal. Chem. 2014, 86 (19) 9887– 9894 DOI: 10.1021/ac5025039
      Google Scholar
      8
      Precision High-Throughput Proton NMR Spectroscopy of Human Urine, Serum, and Plasma for Large-Scale Metabolic Phenotyping
      Dona, Anthony C.; Jimenez, Beatriz; Schafer, Hartmut; Humpfer, Eberhard; Spraul, Manfred; Lewis, Matthew R.; Pearce, Jake T. M.; Holmes, Elaine; Lindon, John C.; Nicholson, Jeremy K.
      Analytical Chemistry (Washington, DC, United States) (2014), 86 (19), 9887-9894CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Proton NMR-based metabolic phenotyping of urine and blood plasma/serum samples provides important prognostic and diagnostic information and permits monitoring of disease progression in an objective manner. Much effort has been made in recent years to develop NMR instrumentation and technol. to allow the acquisition of data in an effective, reproducible, and high-throughput approach that allows the study of general population samples from epidemiol. collections for biomarkers of disease risk. The challenge remains to develop highly reproducible methods and standardized protocols that minimize tech. or exptl. bias, allowing realistic interlab. comparisons of subtle biomarker information. Here the authors present a detailed set of updated protocols that carefully consider major exptl. conditions, including sample prepn., spectrometer parameters, NMR pulse sequences, throughput, reproducibility, quality control, and resoln. These results provide an exptl. platform that facilitates NMR spectroscopy usage across different large cohorts of biofluid samples, enabling integration of global metabolic profiling that is a prerequisite for personalized healthcare.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhsVGhtLbE&md5=34870c3f541fd9d5397ba7252c6010fa
    9. 9
      Nicholson, J. K. Global systems biology, personalized medicine and molecular epidemiology Mol. Syst. Biol. 2006, 2, 52 DOI: 10.1038/msb4100095
      Google Scholar
      There is no corresponding record for this reference.
    10. 10
      Nicholson, J. K.; Holmes, E.; Kinross, J. M.; Darzi, A. W.; Takats, Z.; Lindon, J. C. Metabolic phenotyping in clinical and surgical environments Nature 2012, 491 (7424) 384– 392 DOI: 10.1038/nature11708
      Google Scholar
      10
      Metabolic phenotyping in clinical and surgical environments
      Nicholson, Jeremy K.; Holmes, Elaine; Kinross, James M.; Darzi, Ara W.; Takats, Zoltan; Lindon, John C.
      Nature (London, United Kingdom) (2012), 491 (7424), 384-392CODEN: NATUAS; ISSN:0028-0836. (Nature Publishing Group)
      A review. Metabolic phenotyping involves the comprehensive anal. of biol. fluids or tissue samples. This anal. allows biochem. classification of a person's physiol. or pathol. states that relate to disease diagnosis or prognosis at the individual level and to disease risk factors at the population level. These approaches are currently being implemented in hospital environments and in regional phenotyping centers worldwide. The ultimate aim of such work is to generate information on patient biol. using techniques such as patient stratification to better inform clinicians on factors that will enhance diagnosis or the choice of therapy. There have been many reports of direct applications of metabolic phenotyping in a clin. setting.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ogs7vM&md5=5fe49f7f75ab76e6c55b338021559098
    11. 11
      Sévin, D. C.; Kuehne, A.; Zamboni, N.; Sauer, U. Biological insights through nontargeted metabolomics Curr. Opin. Biotechnol. 2015, 34, 1– 8 DOI: 10.1016/j.copbio.2014.10.001
      Google Scholar
      11
      Biological insights through nontargeted metabolomics
      Sevin, Daniel C.; Kuehne, Andreas; Zamboni, Nicola; Sauer, Uwe
      Current Opinion in Biotechnology (2015), 34 (), 1-8CODEN: CUOBE3; ISSN:0958-1669. (Elsevier B.V.)
      A review. Metabolomics is increasingly employed to investigate metab. and its reciprocal crosstalk with cellular signaling and regulation. In recent years, several nontargeted metabolomics methods providing substantial metabolome coverage have been developed. Here, we review and compare the contributions of traditional targeted and nontargeted metabolomics in advancing different research areas ranging from biotechnol. to human health. Although some studies demonstrated the power of nontargeted profiling in generating unexpected and yet highly important insights, we found that most mechanistic links were still revealed by hypothesis-driven targeted methods. Novel computational approaches for formal interpretation of complex metabolic patterns and integration of complementary mol. layers are required to tap the full potential of nontargeted metabolomics for data-driven, discovery-oriented research and rapidly nucleating novel biol. insights.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhslOkt7jI&md5=4bbd0528f7b2346c2c07191ca165cdd2
    12. 12
      Bictash, M.; Ebbels, T. M.; Chan, Q.; Loo, R. L.; Yap, I. K. S.; Brown, I. J.; de Iorio, M.; Daviglus, M. L.; Holmes, E.; Stamler, J. Opening up the “Black Box”: metabolic phenotyping and metabolome-wide association studies in epidemiology J. Clin. Epidemiol. 2010, 63 (9) 970– 979 DOI: 10.1016/j.jclinepi.2009.10.001
      Google Scholar
      12
      Opening up the "Black Box": metabolic phenotyping and metabolome-wide association studies in epidemiology
      Bictash Magda; Ebbels Timothy M; Chan Queenie; Loo Ruey Leng; Yap Ivan K S; Brown Ian J; de Iorio Maria; Daviglus Martha L; Holmes Elaine; Stamler Jeremiah; Nicholson Jeremy K; Elliott Paul
      Journal of clinical epidemiology (2010), 63 (9), 970-9 ISSN:.
      BACKGROUND: Metabolic phenotyping of humans allows information to be captured on the interactions between dietary, xenobiotic, other lifestyle and environmental exposures, and genetic variation, which together influence the balance between health and disease risks at both individual and population levels. OBJECTIVES: We describe here the main procedures in large-scale metabolic phenotyping and their application to metabolome-wide association (MWA) studies. METHODS: By use of high-throughput technologies and advanced spectroscopic methods, application of metabolic profiling to large-scale epidemiologic sample collections, including metabolome-wide association (MWA) studies for biomarker discovery and identification. DISCUSSION: Metabolic profiling at epidemiologic scale requires optimization of experimental protocol to maximize reproducibility, sensitivity, and quantitative reliability, and to reduce analytical drift. Customized multivariate statistical modeling approaches are needed for effective data visualization and biomarker discovery with control for false-positive associations since 100s or 1,000s of complex metabolic spectra are being processed. CONCLUSION: Metabolic profiling is an exciting addition to the armamentarium of the epidemiologist for the discovery of new disease-risk biomarkers and diagnostics, and to provide novel insights into etiology, biological mechanisms, and pathways.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC3cjhsVaqsw%253D%253D&md5=4bb1ffa9f64917c6931a7ee45bcd3788
    13. 13
      Chadeau-Hyam, M.; Ebbels, T. M. D.; Brown, I. J.; Chan, Q.; Stamler, J.; Huang, C. C.; Daviglus, M. L.; Ueshima, H.; Zhao, L.; Holmes, E. Metabolic profiling and the metabolome-wide association study: significance level for biomarker identification J. Proteome Res. 2010, 9 (9) 4620– 4627 DOI: 10.1021/pr1003449
      Google Scholar
      13
      Metabolic Profiling and the Metabolome-Wide Association Study: Significance Level For Biomarker Identification
      Chadeau-Hyam, Marc; Ebbels, Timothy M. D.; Brown, Ian J.; Chan, Queenie; Stamler, Jeremiah; Huang, Chiang Ching; Daviglus, Martha L.; Ueshima, Hirotsugu; Zhao, Liancheng; Holmes, Elaine; Nicholson, Jeremy K.; Elliott, Paul; De Iorio, Maria
      Journal of Proteome Research (2010), 9 (9), 4620-4627CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
      High throughput metabolic profiling via the metabolome-wide assocn. study (MWAS) is a powerful new approach to identify biomarkers of disease risk, but there are methodol. challenges: high dimensionality, high level of collinearity, the existence of peak overlap within metabolic spectral data, multiple testing, and selection of a suitable significance threshold. The authors define the metabolome-wide significance level (MWSL) as the threshold required to control the family wise error rate through a permutation approach. The authors used 1H NMR spectroscopic profiles of 24 h urinary collections from the INTERMAP study. The authors' results show that the MWSL primarily depends on sample size and spectral resoln. The MWSL ests. can be used to guide selection of discriminatory biomarkers in MWA studies. In a simulation study, the authors compare statistical performance of the MWSL approach to two variants of orthogonal partial least-squares (OPLS) method with respect to statistical power, false pos. rate and correspondence of ranking of the most significant spectral variables. The authors' results show that the MWSL approach as estd. by the univariate t test is not outperformed by OPLS and offers a fast and simple method to detect disease-related discriminatory features in human NMR urinary metabolic profiles.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXhtVSju7nI&md5=c4e86345e9c4ef1f6e2b40bac05956a2
    14. 14
      De Livera, A. M.; Dias, D. A.; De Souza, D.; Rupasinghe, T.; Pyke, J.; Tull, D.; Roessner, U.; McConville, M.; Speed, T. P. Normalizing and integrating metabolomics data Anal. Chem. 2012, 84 (24) 10768– 10776 DOI: 10.1021/ac302748b
      Google Scholar
      14
      Normalizing and Integrating Metabolomics Data
      De Livera, Alysha M.; Dias, Daniel A.; De Souza, David; Rupasinghe, Thusitha; Pyke, James; Tull, Dedreia; Roessner, Ute; McConville, Malcolm; Speed, Terence P.
      Analytical Chemistry (Washington, DC, United States) (2012), 84 (24), 10768-10776CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Metabolomics research often requires the use of multiple anal. platforms, batches of samples, and labs., any of which can introduce a component of unwanted variation. In addn., every expt. is subject to within-platform and other exptl. variation, which often includes unwanted biol. variation. Such variation must be removed in order to focus on the biol. information of interest. We present a broadly applicable method for the removal of unwanted variation arising from various sources for the identification of differentially abundant metabolites and, hence, for the systematic integration of data on the same quantities from different sources. We illustrate the versatility and the performance of the approach in four applications, and we show that it has several advantages over the existing normalization methods.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ons7fF&md5=2de0f2d242c1a42dfb539d2570ca50f5
    15. 15
      De Livera, A. M.; Olshansky, M.; Speed, T. P. Statistical analysis of metabolomics data Methods Mol. Biol. 2013, 1055, 291– 307 DOI: 10.1007/978-1-62703-577-4_20
      Google Scholar
      15
      Statistical analysis of metabolomics data
      De Livera, Alysha M.; Olshansky, Moshe; Speed, Terence P.
      Methods in Molecular Biology (New York, NY, United States) (2013), 1055 (Metabolomics Tools for Natural Product Discovery), 291-307CODEN: MMBIED; ISSN:1064-3745. (Springer)
      Statistical matters form an integral part of a metabolomics expt. In this chapter we describe several important aspects in the anal. of metabolomics data such as the removal of unwanted variation and the identification of differentially abundant metabolites, along with a no. of other essential statistical considerations.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhslyhs7nN&md5=8940f4025856d8f08dbabd81679022eb
    16. 16
      Korman, A.; Oh, A.; Raskind, A.; Banks, D. Statistical methods in metabolomics Methods Mol. Biol. 2012, 856, 381– 413 DOI: 10.1007/978-1-61779-585-5_16
      Google Scholar
      16
      Statistical methods in metabolomics
      Korman Alexander; Oh Amy; Raskind Alexander; Banks David
      Methods in molecular biology (Clifton, N.J.) (2012), 856 (), 381-413 ISSN:.
      Metabolomics is the relatively new field in bioinformatics that uses measurements on metabolite abundance as a tool for disease diagnosis and other medical purposes. Although closely related to proteomics, the statistical analysis is potentially simpler since biochemists have significantly more domain knowledge about metabolites. This chapter reviews the challenges that metabolomics poses in the areas of quality control, statistical metrology, and data mining.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BC383psVCltQ%253D%253D&md5=6be2769522ef030929b346dd4014a945
    17. 17
      Ebbels, T. M. D.; Lindon, J. C.; Coen, M. Processing and modeling of nuclear magnetic resonance (NMR) metabolic profiles Methods Mol. Biol. 2011, 708, 365– 388 DOI: 10.1007/978-1-61737-985-7_21
      Google Scholar
      17
      Processing and modeling of nuclear magnetic resonance (NMR) metabolic profiles
      Ebbels, Timothy M. D.; Lindon, John C.; Coen, Muireann
      Methods in Molecular Biology (New York, NY, United States) (2011), 708 (Metabolic Profiling), 365-388CODEN: MMBIED; ISSN:1064-3745. (Springer)
      Modern NMR spectroscopy generates complex and information-rich metabolic profiles. These require robust, accurate, and often sophisticated statistical techniques to yield the max. meaningful knowledge. In this chapter, we describe methods typically used to analyze such data. We begin by describing seven goals of metabolic profile anal., ranging from prodn. of a data table to multi-omic integration for systems biol. Methods for preprocessing and pretreatment are then presented, including issues such as instrument-level spectral processing, data redn. and deconvolution, normalization, scaling, and transformations of the data. We then discuss methods for exploratory modeling and exemplify three techniques: principal components anal., hierarchical clustering, and self-organizing maps. Moving to predictive modeling, we focus our discussion on partial least squares regression, orthogonal partial least squares regression, and genetic algorithm approaches. A typical set of in vitro metabolic profiles is used where possible to compare and contrast the methods. The importance of validating statistical models is highlighted, and std. techniques for doing so, such as training/test set and cross-validation are described. Finally, we discuss the contributions of statistical techniques such as statistical total correlation spectroscopy, and other correlation-based methods have made to the process of structural characterization for unknown metabolites.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXmsVSgur4%253D&md5=70b8d3811208b2588145d74fa39eabde
    18. 18
      Bild, D. E.; Bluemke, D. A.; Burke, G. L.; Detrano, R.; Diez Roux, A. V.; Folsom, A. R.; Greenland, P.; Jacob, D. R.; Kronmal, R.; Liu, K. Multi-ethnic study of atherosclerosis: objectives and design Am. J. Epidemiol. 2002, 156 (9) 871– 881 DOI: 10.1093/aje/kwf113
      Google Scholar
      18
      Multi-Ethnic Study of Atherosclerosis: objectives and design
      Bild Diane E; Bluemke David A; Burke Gregory L; Detrano Robert; Diez Roux Ana V; Folsom Aaron R; Greenland Philip; Jacob David R Jr; Kronmal Richard; Liu Kiang; Nelson Jennifer Clark; O'Leary Daniel; Saad Mohammed F; Shea Steven; Szklo Moyses; Tracy Russell P
      American journal of epidemiology (2002), 156 (9), 871-81 ISSN:0002-9262.
      The Multi-Ethnic Study of Atherosclerosis was initiated in July 2000 to investigate the prevalence, correlates, and progression of subclinical cardiovascular disease (CVD) in a population-based sample of 6,500 men and women aged 45-84 years. The cohort will be selected from six US field centers. Approximately 38% of the cohort will be White, 28% African-American, 23% Hispanic, and 11% Asian (of Chinese descent). Baseline measurements will include measurement of coronary calcium using computed tomography; measurement of ventricular mass and function using cardiac magnetic resonance imaging; measurement of flow-mediated brachial artery endothelial vasodilation, carotid intimal-medial wall thickness, and distensibility of the carotid arteries using ultrasonography; measurement of peripheral vascular disease using ankle and brachial blood pressures; electrocardiography; and assessments of microalbuminuria, standard CVD risk factors, sociodemographic factors, life habits, and psychosocial factors. Blood samples will be assayed for putative biochemical risk factors and stored for use in nested case-control studies. DNA will be extracted and lymphocytes will be immortalized for genetic studies. Measurement of selected subclinical disease indicators and risk factors will be repeated for the study of progression over 7 years. Participants will be followed through 2008 for identification and characterization of CVD events, including acute myocardial infarction and other coronary heart disease, stroke, peripheral vascular disease, and congestive heart failure; therapeutic interventions for CVD; and mortality.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BD38njslemsA%253D%253D&md5=7cc196a4c0a768cbb6d77484bc663abe
    19. 19
      Friedewald, W. T.; Levy, R. I.; Fredrickson, D. S. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge Clin. Chem. 1972, 18 (6) 499– 502
      Google Scholar
      19
      Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge
      Friedewald, William T.; Levy, Robert I.; Fredrickson, Donald S.
      Clinical Chemistry (Washington, DC, United States) (1972), 18 (6), 499-502CODEN: CLCHAU; ISSN:0009-9147.
      A method for evaluating the cholesterol content of the serum low-d. lipoprotein fraction (Sf 0-20) involves measurements of fasting plasma total cholesterol, triglyceride, and high-d. lipoprotein cholesterol concns. none of which requires the use of the preparative ultracentrifuge. Comparison of this procedure with the more direct procedure, in which the ultracentrifuge is used, yielded correlation coeffs. of 0.94-0.99, depending on the patient population compared.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaE38Xkslagsbw%253D&md5=0422d2b1673a49c8344179ec2cf4d7e4
    20. 20
      Karaman, I.; Ferreira, D. L. S.; Boulangé, C. L.; Kaluarachchi, M. R.; Herrington, D.; Dona, A. C.; Castagné, R.; Moayyeri, A.; Lehne, B.; Loh, M. Workflow for Integrated Processing of Multicohort Untargeted (1)H NMR Metabolomics Data in Large-Scale Metabolic Epidemiology J. Proteome Res. 2016, 15 (12) 4188– 4194 DOI: 10.1021/acs.jproteome.6b00125
      Google Scholar
      20
      Workflow for Integrated Processing of Multicohort Untargeted 1H NMR Metabolomics Data in Large-Scale Metabolic Epidemiology
      Karaman, Ibrahim; Ferreira, Diana L. S.; Boulange, Claire L.; Kaluarachchi, Manuja R.; Herrington, David; Dona, Anthony C.; Castagne, Raphaele; Moayyeri, Alireza; Lehne, Benjamin; Loh, Marie; de Vries, Paul S.; Dehghan, Abbas; Franco, Oscar H.; Hofman, Albert; Evangelou, Evangelos; Tzoulaki, Ioanna; Elliott, Paul; Lindon, John C.; Ebbels, Timothy M. D.
      Journal of Proteome Research (2016), 15 (12), 4188-4194CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
      Large-scale metabolomics studies involving thousands of samples present multiple challenges in data anal., particularly when an untargeted platform is used. Studies with multiple cohorts and anal. platforms exacerbate existing problems such as peak alignment and normalization. Therefore, there is a need for robust processing pipelines which can ensure reliable data for statistical anal. The COMBI-BIO project incorporates serum from ∼8000 individuals, in 3 cohorts, profiled by 6 assays in 2 phases using both 1H-NMR and UPLC-MS. Here the authors present the COMBI-BIO NMR anal. pipeline and demonstrate its fitness for purpose using representative quality control (QC) samples. NMR spectra were first aligned and normalized. After eliminating interfering signals, outliers identified using Hotelling's T2 were removed and a cohort/phase adjustment was applied, resulting in two NMR datasets (CPMG and NOESY). Alignment of the NMR data increases the correlation-based alignment quality measure from 0.319 to 0.391 for CPMG and from 0.536 to 0.586 for NOESY, showing that the improvement was present across both large and small peaks. End-to-end quality assessment of the pipeline was achieved using Hotelling's T2 distributions. For CPMG spectra, the interquartile range decreased from 1.425 in raw QC data to 0.679 in processed spectra, while the corresponding change for NOESY spectra was 0.795-0.636 indicating an improvement in precision following processing. PCA indicated that gross phase and cohort differences were no longer present. These results illustrate that the pipeline produces robust and reproducible data, successfully addressing the methodol. challenges of this large multi-faceted study.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhsFWjtbfE&md5=5c953b368c549337338ef48eec74a13c
    21. 21
      Veselkov, K. A.; Lindon, J. C.; Ebbels, T. M. D.; Crockford, D.; Volynkin, V. V.; Holmes, E.; Davies, D. B.; Nicholson, J. K. Recursive segment-wise peak alignment of biological (1)h NMR spectra for improved metabolic biomarker recovery Anal. Chem. 2009, 81 (1) 56– 66 DOI: 10.1021/ac8011544
      Google Scholar
      21
      Recursive Segment-Wise Peak Alignment of Biological 1H NMR Spectra for Improved Metabolic Biomarker Recovery
      Veselkov, Kirill A.; Lindon, John C.; Ebbels, Timothy M. D.; Crockford, Derek; Volynkin, Vladimir V.; Holmes, Elaine; Davies, David B.; Nicholson, Jeremy K.
      Analytical Chemistry (Washington, DC, United States) (2009), 81 (1), 56-66CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Chem. shift variation in small-mol. 1H NMR signals of biofluids complicates biomarker information recovery in metabonomic studies when using multivariate statistical and pattern recognition tools. Current peak realignment methods are generally time-consuming or align major peaks at the expense of minor peak shift accuracy. The authors present a novel recursive segment-wise peak alignment (RSPA) method to reduce variability in peak positions across the multiple 1H NMR spectra used in metabonomic studies. The method refines a segmentation of ref. and test spectra in a top-down fashion, sequentially subdividing the initial larger segments, as required, to improve the local spectral alignment. The authors also describe a general procedure that allows robust comparison of realignment quality of various available methods for a range of peak intensities. The RSPA method is illustrated with respect to 140 1H NMR rat urine spectra from a caloric restriction study and is compared with several other widely used peak alignment methods. The authors demonstrate the superior performance of the RSPA alignment over a wide range of peaks and its capacity to enhance interpretability and robustness of multivariate statistical tools. The approach is widely applicable for NMR-based metabolic studies and is potentially suitable for many other types of data sets such as chromatog. profiles and MS data.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhsVGnsLnJ&md5=f8386d82f13d01b0609d0d504f899787
    22. 22
      Dieterle, F.; Ross, A.; Schlotterbeck, G.; Senn, H. Probabilistic quotient normalization as robust method to account for dilution of complex biological mixtures. Application in 1H NMR metabonomics Anal. Chem. 2006, 78 (13) 4281– 4290 DOI: 10.1021/ac051632c
      Google Scholar
      22
      Probabilistic Quotient Normalization as Robust Method to Account for Dilution of Complex Biological Mixtures. Application in 1H NMR Metabonomics
      Dieterle, Frank; Ross, Alfred; Schlotterbeck, Goetz; Senn, Hans
      Analytical Chemistry (2006), 78 (13), 4281-4290CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      For the anal. of the spectra of complex biofluids, preprocessing methods play a crucial role in rendering the subsequent data analyses more robust and accurate. Normalization is a preprocessing method, which accounts for different dilns. of samples by scaling the spectra to the same virtual overall concn. In the field of 1H NMR metabonomics integral normalization, which scales spectra to the same total integral, is the de facto std. In this work, it is shown that integral normalization is a suboptimal method for normalizing spectra from metabonomic studies. Esp. strong metabonomic changes, evident as massive amts. of single metabolites in samples, significantly hamper the integral normalization resulting in incorrectly scaled spectra. The probabilistic quotient normalization is introduced in this work. This method is based on the calcn. of a most probable diln. factor by looking at the distribution of the quotients of the amplitudes of a test spectrum by those of a ref. spectrum. Simulated spectra, spectra of urine samples from a metabonomic study with cyclosporin-A as the active compd., and spectra of more than 4000 samples of control animals demonstrate that the probabilistic quotient normalization is by far more robust and more accurate than the widespread integral normalization and vector length normalization.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD28XltVCgtro%253D&md5=6eb6377326a9df2a59b6afb2a9c6e47d
    23. 23
      van Velzen, E. J. J.; Westerhuis, J. A.; van Duynhoven, J. P. M.; van Dorsten, F. A.; Hoefsloot, H. C. J.; Jacobs, D. M.; Smit, S.; Draijer, R.; Kroner, C. I.; Smilde, A. K. Multilevel data analysis of a crossover designed human nutritional intervention study J. Proteome Res. 2008, 7 (10) 4483– 4491 DOI: 10.1021/pr800145j
      Google Scholar
      23
      Multilevel data analysis of a crossover designed human nutritional intervention study
      van Velzen, Ewoud J. J.; Westerhuis, Johan A.; van Duynhoven, John P. M.; van Dorsten, Ferdi A.; Hoefsloot, Huub C. J.; Jacobs, Doris M.; Smit, Suzanne; Draijer, Richard; Kroner, Christine I.; Smilde, Age K.
      Journal of Proteome Research (2008), 7 (10), 4483-4491CODEN: JPROBS; ISSN:1535-3893. (American Chemical Society)
      A new method is introduced for the anal. of '-omics' data derived from crossover designed drug or nutritional intervention studies. The method aims at finding systematic variations in metabolic profiles after a drug or nutritional challenge and takes advantage of the crossover design in the data. The method, which can be considered as a multivariate extension of paired t test, generates different multivariate submodels for the between- and the within-subject variation in the data. A major advantage of this variation splitting is that each submodel can be analyzed sep. without confounding with other variation sources. The power of the multilevel approach is demonstrated in a human nutritional intervention study which used NMR-based metabolomics to assess the metabolic impact of grape/wine ext. consumption. The variations in urine metabolic profiles were studied between and within the human subjects using the multilevel anal. After variation splitting, the multilevel PCA was used to investigate the exptl. and biol. differences between the subjects, whereas a multilevel PLS-DA model was used to reveal the net treatment effect within the subjects. The obsd. treatment effect was validated with cross model validation and permutations. The statistical significance of the multilevel classification model is a major improvement compared to ordinary PLS-DA models without variation splitting. Rank products were used to det. which NMR signals were most important in the multilevel classification model.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtVGgu77J&md5=7374579a6254bb8b23fb922336b8d3a2
    24. 24
      Cloarec, O.; Dumas, M.-E.; Craig, A.; Barton, R. H.; Trygg, J.; Hudson, J.; Blancher, C.; Gauguier, D.; Lindon, J. C.; Holmes, E. Statistical total correlation spectroscopy: an exploratory approach for latent biomarker identification from metabolic 1H NMR data sets Anal. Chem. 2005, 77 (5) 1282– 1289 DOI: 10.1021/ac048630x
      Google Scholar
      24
      Statistical Total Correlation Spectroscopy: An Exploratory Approach for Latent Biomarker Identification from Metabolic 1H NMR Data Sets
      Cloarec, Olivier; Dumas, Marc-Emmanuel; Craig, Andrew; Barton, Richard H.; Trygg, Johan; Hudson, Jane; Blancher, Christine; Gauguier, Dominique; Lindon, John C.; Holmes, Elaine; Nicholson, Jeremy
      Analytical Chemistry (2005), 77 (5), 1282-1289CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      We describe here the implementation of the statistical total correlation spectroscopy (STOCSY) anal. method for aiding the identification of potential biomarker mols. in metabonomic studies based on NMR spectroscopic data. STOCSY takes advantage of the multicollinearity of the intensity variables in a set of spectra (in this case 1H NMR spectra) to generate a pseudo-two-dimensional NMR spectrum that displays the correlation among the intensities of the various peaks across the whole sample. This method is not limited to the usual connectivities that are deducible from more std. two-dimensional NMR spectroscopic methods, such as TOCSY. Moreover, two or more mols. involved in the same pathway can also present high intermol. correlations because of biol. covariance or can even be anticorrelated. This combination of STOCSY with supervised pattern recognition and particularly orthogonal projection on latent structure-discriminant anal. (O-PLS-DA) offers a new powerful framework for anal. of metabonomic data. In a first step O-PLS-DA exts. the part of NMR spectra related to discrimination. This information is then cross-combined with the STOCSY results to help identify the mols. responsible for the metabolic variation. To illustrate the applicability of the method, it has been applied to 1H NMR spectra of urine from a metabonomic study of a model of insulin resistance based on the administration of a carbohydrate diet to three different mice strains (C57BL/6Oxjr, BALB/cOxjr, and 129S6/SvEvOxjr) in which a series of metabolites of biol. importance can be conclusively assigned and identified by use of the STOCSY approach.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXmsVSqsg%253D%253D&md5=6095939b18fae599b0090820561e894c
    25. 25
      Posma, J. M.; Garcia-Perez, I.; De Iorio, M.; Lindon, J. C.; Elliott, P.; Holmes, E.; Ebbels, T. M. D.; Nicholson, J. K. Subset optimization by reference matching (STORM): an optimized statistical approach for recovery of metabolic biomarker structural information from 1H NMR spectra of biofluids Anal. Chem. 2012, 84 (24) 10694– 10701 DOI: 10.1021/ac302360v
      Google Scholar
      25
      Subset Optimization by Reference Matching (STORM): An Optimized Statistical Approach for Recovery of Metabolic Biomarker Structural Information from 1H NMR Spectra of Biofluids
      Posma, Joram M.; Garcia-Perez, Isabel; De Iorio, Maria; Lindon, John C.; Elliott, Paul; Holmes, Elaine; Ebbels, Timothy M. D.; Nicholson, Jeremy K.
      Analytical Chemistry (Washington, DC, United States) (2012), 84 (24), 10694-10701CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      We describe a new multivariate statistical approach to recover metabolite structure information from multiple 1H NMR spectra in population sample sets. Subset optimization by ref. matching (STORM) was developed to select subsets of 1H NMR spectra that contain specific spectroscopic signatures of biomarkers differentiating between different human populations. STORM aims to improve the visualization of structural correlations in spectroscopic data by using these reduced spectral subsets contg. smaller nos. of samples than the no. of variables (n « p). We have used statistical shrinkage to limit the no. of false pos. assocns. and to simplify the overall interpretation of the autocorrelation matrix. The STORM approach has been applied to findings from an ongoing human metabolome-wide assocn. study on body mass index to identify a biomarker metabolite present in a subset of the population. Moreover, we have shown how STORM improves the visualization of more abundant NMR peaks compared to a previously published method (statistical total correlation spectroscopy, STOCSY). STORM is a useful new tool for biomarker discovery in the 'omic' sciences that has widespread applicability. It can be applied to any type of data, provided that there is interpretable correlation among variables, and can also be applied to data with more than one dimension (e.g., 2D NMR spectra).
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38Xhs1Ons7zF&md5=0fbd04f3eac9cf7b475e4724359b9cad
    26. 26
      Navratil, V.; Pontoizeau, C.; Billoir, E.; Blaise, B. J. SRV: an open-source toolbox to accelerate the recovery of metabolic biomarkers and correlations from metabolic phenotyping datasets Bioinformatics 2013, 29 (10) 1348– 1349 DOI: 10.1093/bioinformatics/btt136
      Google Scholar
      There is no corresponding record for this reference.
    27. 27
      Nicholson, J. K.; Foxall, P. J.; Spraul, M.; Farrant, R. D.; Lindon, J. C. 750 MHz 1H and 1H-13C NMR spectroscopy of human blood plasma Anal. Chem. 1995, 67 (5) 793– 811 DOI: 10.1021/ac00101a004
      Google Scholar
      27
      750 MHz 1H and 1H-13C NMR Spectroscopy of Human Blood Plasma
      Nicholson, Jeremy K.; Foxall, Peta J. D.; Spraul, Manfred; Farrant, R. Duncan; Lindon, John C.
      Analytical Chemistry (1995), 67 (5), 793-811CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      High-resoln. 750 MHz 1H NMR spectra of control human blood plasma have been measured and assigned by the concerted use of a range of spin-echo, two-dimensional J-resolved, and homonuclear and heteronuclear (1H-13C) correlation methods. The increased spectral dispersion and sensitivity at 750 MHz enable the assignment of numerous 1H and 13C resonances from many mol. species that cannot be detected at lower frequencies. This work presents the most comprehensive assignment of the 1H NMR spectra of blood plasma yet achieved and includes the assignment of signals from 43 low Mr metabolites, including many with complex or strongly coupled spin systems. New assignments are also provided from the 1H and 13C NMR signals from several important macromol. species in whole blood plasma, i.e., very-low-d., low-d., and high-d. lipoproteins, albumin, and α1-acid glycoprotein. The temp. dependence of the one-dimensional and spin-echo 750 MHz 1H NMR spectra of plasma was investigated over the range 292-310 K. The 1H NMR signals from the fatty acyl side chains of the lipoproteins increased substantially with temp. (hence also mol. mobility), with a disproportionate increase from lipids in low-d. lipoprotein. Two-dimensional 1H-13C heteronuclear multiple quantum coherence spectroscopy at 292 and 310 K allowed both the direct detection of cholesterol and choline species bound in high-d. lipoprotein and the assignment of their signals and confirmed the assignment of most of the lipoprotein resonances.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXjsFajurs%253D&md5=026f5d082dc75abb431f04c957231002
    28. 28
      Merrifield, C. A.; Lewis, M.; Claus, S. P.; Beckonert, O. P.; Dumas, M.-E.; Duncker, S.; Kochhar, S.; Rezzi, S.; Lindon, J. C.; Bailey, M. A metabolic system-wide characterisation of the pig: a model for human physiology Mol. BioSyst. 2011, 7 (9) 2577– 2588 DOI: 10.1039/c1mb05023k
      Google Scholar
      28
      A metabolic system-wide characterisation of the pig: a model for human physiology
      Merrifield, Claire A.; Lewis, Marie; Claus, Sandrine P.; Beckonert, Olaf P.; Dumas, Marc-Emmanuel; Duncker, Swantje; Kochhar, Sunil; Rezzi, Serge; Lindon, John C.; Bailey, Mick; Holmes, Elaine; Nicholson, Jeremy K.
      Molecular BioSystems (2011), 7 (9), 2577-2588CODEN: MBOIBW; ISSN:1742-2051. (Royal Society of Chemistry)
      The pig is a single-stomached omnivorous mammal and is an important model of human disease and nutrition. As such, it is necessary to establish a metabolic framework from which pathol.-based variation can be compared. Here, a combination of one and two-dimensional 1H and 13C NMR spectroscopy (NMR) and high-resoln. magic angle spinning (HR-MAS) NMR was used to provide a systems overview of porcine metab. via characterization of the urine, serum, liver and kidney metabolomes. The metabolites obsd. in each of these biol. compartments were found to be qual. comparable to the metabolic signature of the same biol. matrixes in humans and rodents. The data were modelled using a combination of principal components anal. and Venn diagram mapping. Urine represented the most metabolically distinct biol. compartment studied, with a relatively greater no. of NMR detectable metabolites present, many of which are implicated in gut-microbial co-metabolic processes. The major inter-species differences obsd. were in the phase II conjugation of extra-genomic metabolites; the pig was obsd. to conjugate p-cresol, a gut microbial metabolite of tyrosine, with glucuronide rather than sulfate as seen in man. These observations are important to note when considering the translatability of exptl. data derived from porcine models.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhtVSmtL%252FM&md5=6784be376df517f8b75c6449cba9e9f8
    29. 29
      Wishart, D. S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A. C.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S. HMDB: the Human Metabolome Database Nucleic Acids Res. 2007, 35 (Database) D521– D526 DOI: 10.1093/nar/gkl923
      Google Scholar
      29
      HMDB: the Human Metabolome Database
      Wishart, David S.; Tzur, Dan; Knox, Craig; Eisner, Roman; Guo, An Chi; Young, Nelson; Cheng, Dean; Jewell, Kevin; Arndt, David; Sawhney, Summit; Fung, Chris; Nikolai, Lisa; Lewis, Mike; Coutouly, Marie-Aude; Forsythe, Ian; Tang, Peter; Shrivastava, Savita; Jeroncic, Kevin; Stothard, Paul; Amegbey, Godwin; Block, David; Hau, David. D.; Wagner, James; Miniaci, Jessica; Clements, Melisa; Gebremedhin, Mulu; Guo, Natalie; Zhang, Ying; Duggan, Gavin E.; MacInnis, Glen D.; Weljie, Alim M.; Dowlatabadi, Reza; Bamforth, Fiona; Clive, Derrick; Greiner, Russ; Li, Liang; Marrie, Tom; Sykes, Brian D.; Vogel, Hans J.; Querengesser, Lori
      Nucleic Acids Research (2007), 35 (Database Iss), D521-D526CODEN: NARHAD; ISSN:0305-1048. (Oxford University Press)
      The Human Metabolome Database (HMDB) is currently the most complete and comprehensive curated collection of human metabolite and human metab. data in the world. It contains records for more than 2180 endogenous metabolites with information gathered from thousands of books, journal articles and electronic databases. In addn. to its comprehensive literature-derived data, the HMDB also contains an extensive collection of exptl. metabolite concn. data compiled from hundreds of mass spectra (MS) and NMR metabolomic analyses performed on urine, blood and cerebrospinal fluid samples. This is further supplemented with thousands of NMR and MS spectra collected on purified, ref. metabolites. Each metabolite entry in the HMDB contains an av. of 90 sep. data fields including a comprehensive compd. description, names and synonyms, structural information, physico-chem. data, ref. NMR and MS spectra, biofluid concns., disease assocns., pathway information, enzyme data, gene sequence data, SNP and mutation data as well as extensive links to images, refs. and other public databases. Extensive searching, relational querying and data browsing tools are also provided. The HMDB is designed to address the broad needs of biochemists, clin. chemists, physicians, medical geneticists, nutritionists and members of the metabolomics community.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXivFKhtA%253D%253D&md5=adb0346dd43ced1de7154c4c8d2c57d4
    30. 30
      Ulrich, E. L.; Akutsu, H.; Doreleijers, J. F.; Harano, Y.; Ioannidis, Y. E.; Lin, J.; Livny, M.; Mading, S.; Maziuk, D.; Miller, Z. BioMagResBank Nucleic Acids Res. 2008, 36 (Database) D402– D408 DOI: 10.1093/nar/gkm957
      Google Scholar
      30
      BioMagResBank
      Ulrich, Eldon L.; Akutsu, Hideo; Doreleijers, Jurgen F.; Harano, Yoko; Ioannidis, Yannis E.; Lin, Jundong; Livny, Miron; Mading, Steve; Maziuk, Dimitri; Miller, Zachary; Nakatani, Eiichi; Schulte, Christopher F.; Tolmie, David E.; Kent Wenger, R.; Yao, Hongyang; Markley, John L.
      Nucleic Acids Research (2008), 36 (Database Iss), D402-D408CODEN: NARHAD; ISSN:0305-1048. (Oxford University Press)
      The BioMagResBank (BMRB: www.bmrb.wisc.edu) is a repository for exptl. and derived data gathered from NMR (NMR) spectroscopic studies of biol. mols. BMRB is a partner in the Worldwide Protein Data Bank (wwPDB). The BMRB archive consists of four main data depositories: (i) quant. NMR spectral parameters for proteins, peptides, nucleic acids, carbohydrates and ligands or cofactors (assigned chem. shifts, coupling consts. and peak lists) and derived data (relaxation parameters, residual dipolar couplings, hydrogen exchange rates, pKa values, etc.), (ii) databases for NMR restraints processed from original author depositions available from the Protein Data Bank, (iii) time-domain (raw) spectral data from NMR expts. used to assign spectral resonances and det. the structures of biol. macromols. and (iv) a database of one- and two-dimensional 1H and 13C one- and two-dimensional NMR spectra for over 250 metabolites. The BMRB website provides free access to all of these data. BMRB has tools for querying the archive and retrieving information and an ftp site (ftp.bmrb.wisc.edu) where data in the archive can be downloaded in bulk. Two BMRB mirror sites exist: one at the PDBj, Protein Research Institute, Osaka University, Osaka, Japan (bmrb.protein.osaka-u.ac.jp) and the other at CERM, University of Florence, Florence, Italy (bmrb.postgenomicnmr.net/). The site at Osaka also accepts and processes data depositions.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtVWitb0%253D&md5=520c47beb1bc57e57ef14fac08a6fd3d
    31. 31
      Sumner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W.-M.; Fiehn, O.; Goodacre, R.; Griffin, J. L. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI) Metabolomics 2007, 3 (3) 211– 221 DOI: 10.1007/s11306-007-0082-2
      Google Scholar
      31
      Proposed minimum reporting standards for chemical analysis. Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI)
      Sumner, Lloyd W.; Amberg, Alexander; Barrett, Dave; Beale, Michael H.; Beger, Richard; Daykin, Clare A.; Fan, Teresa W.-M.; Fiehn, Oliver; Goodacre, Royston; Griffin, Julian L.; Hankemeier, Thomas; Hardy, Nigel; Harnly, James; Higashi, Richard; Kopka, Joachim; Lane, Andrew N.; Lindon, John C.; Marriott, Philip; Nicholls, Andrew W.; Reily, Michael D.; Thaden, John J.; Viant, Mark R.
      Metabolomics (2007), 3 (3), 211-221CODEN: METAHQ; ISSN:1573-3882. (Springer)
      There is a general consensus that supports the need for standardized reporting of metadata or information describing large-scale metabolomics and other functional genomics data sets. Reporting of std. metadata provides a biol. and empirical context for the data, facilitates exptl. replication, and enables the reinterrogation and comparison of data by others. Accordingly, the Metabolomics Stds. Initiative is building a general consensus concerning the min. reporting stds. for metabolomics expts. of which the Chem. Anal. Working Group (CAWG) is a member of this community effort. This article proposes the min. reporting stds. related to the chem. anal. aspects of metabolomics expts. including: sample prepn., exptl. anal., quality control, metabolite identification, and data pre-processing. These min. stds. currently focus mostly upon mass spectrometry and NMR spectroscopy due to the popularity of these techniques in metabolomics. However, addnl. input concerning other techniques is welcomed and can be provided via the CAWG online discussion forum at http://msi-workgroups.sourceforge.net/ or http://[email protected]. Further, community input related to this document can also be provided via this electronic forum.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhtFOntbvL&md5=e636dad276f9f1b2e756db169368eb53
    32. 32
      Altmaier, E.; Fobo, G.; Heier, M.; Thorand, B.; Meisinger, C.; Römisch-Margl, W.; Waldenberger, M.; Gieger, C.; Illig, T.; Adamski, J. Metabolomics approach reveals effects of antihypertensives and lipid-lowering drugs on the human metabolism Eur. J. Epidemiol. 2014, 29 (5) 325– 336 DOI: 10.1007/s10654-014-9910-7
      Google Scholar
      32
      Metabolomics approach reveals effects of antihypertensives and lipid-lowering drugs on the human metabolism
      Altmaier, Elisabeth; Fobo, Gisela; Heier, Margit; Thorand, Barbara; Meisinger, Christine; Roemisch-Margl, Werner; Waldenberger, Melanie; Gieger, Christian; Illig, Thomas; Adamski, Jerzy; Suhre, Karsten; Kastenmueller, Gabi
      European Journal of Epidemiology (2014), 29 (5), 325-336CODEN: EJEPE8; ISSN:0393-2990. (Springer)
      The mechanism of antihypertensive and lipid-lowering drugs on the human organism is still not fully understood. New insights on the drugs' action can be provided by a metabolomics-driven approach, which offers a detailed view of the physiol. state of an organism. Here, we report a metabolome-wide assocn. study with 295 metabolites in human serum from 1,762 participants of the KORA F4 (Cooperative Health Research in the Region of Augsburg) study population. Our intent was to find variations of metabolite concns. related to the intake of various drug classes and-based on the assocns. found-to generate new hypotheses about on-target as well as off-target effects of these drugs. In total, we found 41 significant assocns. for the drug classes investigated: For beta-blockers (11 assocns.), angiotensin-converting enzyme (ACE) inhibitors (four assoc.), diuretics (seven assoc.), statins (ten assoc.), and fibrates (nine assoc.) the top hits were pyroglutamine, phenylalanylphenylalanine, pseudouridine, 1-arachidonoylglycerophosphocholine, and 2-hydroxyisobutyrate, resp. For beta-blockers we obsd. significant assocns. with metabolite concns. that are indicative of drug side-effects, such as increased serotonin and decreased free fatty acid levels. Intake of ACE inhibitors and statins assocd. with metabolites that provide insight into the action of the drug itself on its target, such as an assocn. of ACE inhibitors with des-Arg(9)-bradykinin and aspartylphenylalanine, a substrate and a product of the drug-inhibited ACE. The intake of statins which reduce blood cholesterol levels, resulted in changes in the concn. of metabolites of the biosynthesis as well as of the degrdn. of cholesterol. Fibrates showed the strongest assocn. with 2-hydroxyisobutyrate which might be a breakdown product of fenofibrate and, thus, a possible marker for the degrdn. of this drug in the human organism. The anal. of diuretics showed a heterogeneous picture that is difficult to interpret. Taken together, our results provide a basis for a deeper functional understanding of the action and side-effects of antihypertensive and lipid-lowering drugs in the general population.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXnslemt78%253D&md5=958e76f48eb59d5a81ffdf6a5428497d
    33. 33
      Sekula, P.; Goek, O.-N.; Quaye, L.; Barrios, C.; Levey, A. S.; Römisch-Margl, W.; Menni, C.; Yet, I.; Gieger, C.; Inker, L. A. A Metabolome-Wide Association Study of Kidney Function and Disease in the General Population J. Am. Soc. Nephrol. 2016, 27 (4) 1175– 1188 DOI: 10.1681/ASN.2014111099
      Google Scholar
      33
      A metabolome-wide association study of kidney function and disease in the general population
      Sekula, Peggy; Goek, Oemer-Necmi; Quaye, Lydia; Barrios, Clara; Levey, Andrew S.; Roemisch-Margl, Werner; Menni, Cristina; Yet, Idil; Gieger, Christian; Inker, Lesley A.; Adamski, Jerzy; Gronwald, Wolfram; Illig, Thomas; Dettmer, Katja; Krumsiek, Jan; Oefner, Peter J.; Valdes, Ana M.; Meisinger, Christa; Coresh, Josef; Spector, Tim D.; Mohney, Robert P.; Suhre, Karsten; Kastenmueller, Gabi; Koettgen, Anna
      Journal of the American Society of Nephrology (2016), 27 (4), 1175-1189CODEN: JASNEU; ISSN:1046-6673. (American Society of Nephrology)
      Small mols. are extensively metabolized and cleared by the kidney. Changes in serum metabolite concns. may result from impaired kidney function and can be used to est. filtration (e.g., the established marker creatinine) or may precede and potentially contribute to CKD development. Here, we applied a nontargeted metabolomics approach using gas and liq. chromatog. coupled to mass spectrometry to quantify 493 small mols. in human serum. The assocns. of these mols. with GFR estd. on the basis of creatinine (eGFRcr) and cystatin C levels were assessed in ≤1735 participants in the KORA F4 study, followed by replication in 1164 individuals in the TwinsUK registry. After correction for multiple testing, 54 replicated metabolites significantly assocd. with eGFRcr, and six of these showed pairwise correlation (r≥0.50) with established kidney function measures: C-mannosyltryptophan, pseudouridine, N-acetylalanine, erythronate, myo-inositol, and N-acetylcarnosine. Higher C-mannosyltryptophan, pseudouridine, and O-sulfo-L-tyrosine concns. assocd. with incident CKD (eGFRcr <60 mL/min per 1.73 m2) in the KORA F4 study. In contrast with serum creatinine, C-mannosyltryptophan and pseudouridine concns. showed little dependence on sex. Furthermore, correlation with measured GFR in 200 participants in the AASK study was 0.78 for both C-mannosyltryptophan and pseudouridine concn., and highly significant assocns. of both metabolites with incident ESRD disappeared upon adjustment for measured GFR. Thus, these mols. may be alternative or complementary markers of kidney function. In conclusion, our study provides a comprehensive list of kidney function-assocd. metabolites and highlights potential novel filtration markers that may help to improve the estn. of GFR.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XitFSjtLzM&md5=7b34021440ef1938afc8538539bd70a3
    34. 34
      Adkins, D. E.; McClay, J. L.; Vunck, S. A.; Batman, A. M.; Vann, R. E.; Clark, S. L.; Souza, R. P.; Crowley, J. J.; Sullivan, P. F.; van den Oord, E. J. C. G. Behavioral metabolomics analysis identifies novel neurochemical signatures in methamphetamine sensitization Genes Brain Behav. 2013, 12 (8) 780– 791 DOI: 10.1111/gbb.12081
      Google Scholar
      34
      Behavioral metabolomics analysis identifies novel neurochemical signatures in methamphetamine sensitization
      Adkins, D. E.; McClay, J. L.; Vunck, S. A.; Batman, A. M.; Vann, R. E.; Clark, S. L.; Souza, R. P.; Crowley, J. J.; Sullivan, P. F.; van den Oord, E. J. C. G.; Beardsley, P. M.
      Genes, Brain and Behavior (2013), 12 (8), 780-791CODEN: GBBEAO; ISSN:1601-1848. (Wiley-Blackwell)
      Behavioral sensitization has been widely studied in animal models and is theorized to reflect neural modifications assocd. with human psychostimulant addiction. While the mesolimbic dopaminergic pathway is known to play a role, the neurochem. mechanisms underlying behavioral sensitization remain incompletely understood. In this study, we conducted the first metabolomics anal. to globally characterize neurochem. differences assocd. with behavioral sensitization. Methamphetamine (MA)-induced sensitization measures were generated by statistically modeling longitudinal activity data for eight inbred strains of mice. Subsequent to behavioral testing, nontargeted liq. and gas chromatog.-mass spectrometry profiling was performed on 48 brain samples, yielding 301 metabolite levels per sample after quality control. Assocn. testing between metabolite levels and three primary dimensions of behavioral sensitization (total distance, stereotypy and margin time) showed four robust, significant assocns. at a stringent metabolome-wide significance threshold (false discovery rate, FDR <0.05). Results implicated homocarnosine, a dipeptide of GABA and histidine, in total distance sensitization, GABA metabolite 4-guanidinobutanoate and pantothenate in stereotypy sensitization, and myo-inositol in margin time sensitization. Secondary analyses indicated that these assocns. were independent of concurrent MA levels and, with the exception of the myo-inositol assocn., suggest a mechanism whereby strain-based genetic variation produces specific baseline neurochem. differences that substantially influence the magnitude of MA-induced sensitization. These findings demonstrate the utility of mouse metabolomics for identifying novel biomarkers, and developing more comprehensive neurochem. models, of psychostimulant sensitization.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXhslWqsL3E&md5=d01fd2ace97b8309c5e984c83dc35133
    35. 35
      Patterson, N.; Price, A. L.; Reich, D. Population structure and eigenanalysis PLoS Genet. 2006, 2 (12) e190 DOI: 10.1371/journal.pgen.0020190
      Google Scholar
      There is no corresponding record for this reference.
    36. 36
      Schäfer, J.; Strimmer, K. A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics Stat. Appl. Genet. Mol. Biol. 2005, 4 (1) 1– 32 DOI: 10.2202/1544-6115.1175
      Google Scholar
      There is no corresponding record for this reference.
    37. 37
      Auro, K.; Joensuu, A.; Fischer, K.; Kettunen, J.; Salo, P.; Mattsson, H.; Niironen, M.; Kaprio, J.; Eriksson, J. G.; Lehtimäki, T. A metabolic view on menopause and ageing Nat. Commun. 2014, 5, 4708 DOI: 10.1038/ncomms5708
      Google Scholar
      37
      A metabolic view on menopause and ageing
      Auro, Kirsi; Joensuu, Anni; Fischer, Krista; Kettunen, Johannes; Salo, Perttu; Mattsson, Hannele; Niironen, Marjo; Kaprio, Jaakko; Eriksson, Johan G.; Lehtimaki, Terho; Raitakari, Olli; Jula, Antti; Tiitinen, Aila; Jauhiainen, Matti; Soininen, Pasi; Kangas, Antti J.; Kahonen, Mika; Havulinna, Aki S.; Ala-Korpela, Mika; Salomaa, Veikko; Metspalu, Andres; Perola, Markus
      Nature Communications (2014), 5 (), 4708CODEN: NCAOBW; ISSN:2041-1723. (Nature Publishing Group)
      The ageing of the global population calls for a better understanding of age-related metabolic consequences. Here we report the effects of age, sex and menopause on serum metabolites in 26,065 individuals of Northern European ancestry. Age-specific metabolic fingerprints differ significantly by gender and, in females, a substantial atherogenic shift overlapping the time of menopausal transition is obsd. In meta-anal. of 10,083 women, menopause status assocs. with amino acids glutamine, tyrosine and isoleucine, along with serum cholesterol measures and atherogenic lipoproteins. Among 3,204 women aged 40-55 years, menopause status assocs. addnl. with glycine and total, monounsatd., and omega-7 and -9 fatty acids. Our findings suggest that, in addn. to lipid alterations, menopause may contribute to future metabolic and cardiovascular risk via influencing amino-acid concns., adding to the growing evidence of the importance of amino acids in metabolic disease progression. These observations shed light on the metabolic consequences of ageing, gender and menopause at the population level.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXksVCjsb4%253D&md5=0ea518a01a3da198fdced0c89f3ae664
    38. 38
      R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2014; Vol. 73 (1), pp 3– 36.
      Google Scholar
      There is no corresponding record for this reference.
    39. 39
      Benjamini, Y.; Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing J. R. Stat. Soc. Ser. B Methodol. 1995, 57 (1) 289– 300
      Google Scholar
      There is no corresponding record for this reference.
    40. 40
      Zhang, J.; Coombes, K. R. Sources of variation in false discovery rate estimation include sample size, correlation, and inherent differences between groups BMC Bioinf. 2012, 13 (Suppl 13) S1 DOI: 10.1186/1471-2105-13-S13-S1
      Google Scholar
      There is no corresponding record for this reference.

  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jproteome.7b00344.

    • Plots of first two PCA scores for CPMG and NOESY data in MESA; distributions of three different transformations of glucose used to investigate MWSL; percentage of effective/actual number of tests and 95% confidence intervals for CPMG and NOESY; percentage of associated variables for CPMG and NOESY derived from each simulated continuous response; percentage of associated variables for CPMG and NOESY derived from different multiple testing correction strategies; metabolome wide study of glucose from analysis of 30 590 NOESY features; comparison of results from analysis in MESA to those from 80:20 split strategy from NOESY metabolome wide association study of glucose using model 2; significance threshold a′ and ENT based on Bonferonni correction; CPMG and NOESY metabolic features associated with log10 (glucose) in MESA at metabolome-wide significance level for models 1 and 2 (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

PDF [ 2MB]

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect