QComics: Recommendations and Guidelines for Robust, Easily Implementable and Reportable Quality Control of Metabolomics Data.

The implementation of quality control strategies is crucial to ensure the reproducibility, accuracy, and meaningfulness of metabolomics data. However, this pivotal step is often overlooked within the metabolomics workflow and frequently relies on the use of nonstandardized and poorly reported protocols. To address current limitations in this respect, we have developed QComics, a robust, easily implementable and reportable method for monitoring and controlling data quality. The protocol operates in various sequential steps aimed to (i) correct for background noise and carryover, (ii) detect signal drifts and "out-of-control" observations, (iii) deal with missing data, (iv) remove outliers, (v) monitor quality markers to identify samples affected by improper collection, preprocessing, or storage, and (vi) assess overall data quality in terms of precision and accuracy. Notably, this tool considers important issues often neglected along quality control, such as the need of separately handling missing values and truly absent data to avoid losing relevant biological information, as well as the large impact that preanalytical factors may elicit on metabolomics results. Altogether, the guidelines compiled in QComics might contribute to establishing gold standard recommendations and best practices for quality control within the metabolomics community.

Exploration of differential responses to FODMAPs and gluten in people with irritable bowel syndrome- a double-blind randomized cross-over challenge study.

INTRODUCTION: There is large variation in response to diet in irritable bowel syndrome (IBS) and determinants for differential response are poorly understood. OBJECTIVES: Our aim was to investigate differential clinical and molecular responses to provocation with fermentable oligo-, di-, monosaccharides, and polyols (FODMAPs) and gluten in individuals with IBS. METHODS: Data were used from a crossover study with week-long interventions with either FODMAPs, gluten or placebo. The study also included a rapid provocation test. Molecular data consisted of fecal microbiota, short chain fatty acids, and untargeted plasma metabolomics. IBS symptoms were evaluated with the IBS severity scoring system. IBS symptoms were modelled against molecular and baseline questionnaire data, using Random Forest (RF; regression and clustering), Parallel Factor Analysis (PARAFAC), and univariate methods. RESULTS: Regression and classification RF models were in general of low predictive power (Q2 ≤ 0.22, classification rate < 0.73). Out of 864 clustering models, only 2 had significant associations to clusters (0.69 < CR < 0.73, p < 0.05), but with no associations to baseline clinical measures. Similarly, PARAFAC revealed no clear association between metabolome data and IBS symptoms. CONCLUSION: Differential IBS responses to FODMAPs or gluten exposures could not be explained from clinical and molecular data despite extensive exploration with different data analytical approaches. The trial is registered at www. CLINICALTRIALS: gov as NCT03653689 31/08/2018.

Associations of the placental metabolome with immune maturation up to one year of age in the Swedish NICE-cohort.

INTRODUCTION: Allergies and other immune-mediated diseases are thought to result from incomplete maturation of the immune system early in life. We previously showed that infants' metabolites at birth were associated with immune cell subtypes during infancy. The placenta supplies the fetus with nutrients, but may also provide immune maturation signals. OBJECTIVES: To examine the relationship between metabolites in placental villous tissue and immune maturation during the first year of life and infant and maternal characteristics (gestational length, birth weight, sex, parity, maternal age, and BMI). METHODS: Untargeted metabolomics was measured using Liquid Chromatography-Mass Spectrometry. Subpopulations of T and B cells were measured using flow cytometry at birth, 48 h, one, four, and 12 months. Random forest analysis was used to link the metabolomics data with the T and B cell sub populations as well as infant and maternal characteristics. RESULTS: Modest associations (Q2 = 0.2-0.3) were found between the placental metabolome and kappa-deleting recombination excision circles (KREC) at birth and naïve B cells and memory T cells at 12 months. Weak associations were observed between the placental metabolome and sex and parity. Still, most metabolite features of interest were of low intensity compared to associations previously found in cord blood, suggesting that underlying metabolites were not of placental origin. CONCLUSION: Our results indicate that metabolomic measurements of the placenta may not effectively recognize metabolites important for immune maturation.

OMICs Signatures Linking Persistent Organic Pollutants to Cardiovascular Disease in the Swedish Mammography Cohort.

Cardiovascular disease (CVD) development may be linked to persistent organic pollutants (POPs), including organochlorine compounds (OCs) and perfluoroalkyl and polyfluoroalkyl substances (PFAS). To explore underlying mechanisms, we investigated metabolites, proteins, and genes linking POPs with CVD risk. We used data from a nested case-control study on myocardial infarction (MI) and stroke from the Swedish Mammography Cohort - Clinical (n = 657 subjects). OCs, PFAS, and multiomics (9511 liquid chromatography-mass spectrometry (LC-MS) metabolite features; 248 proteins; 8110 gene variants) were measured in baseline plasma. POP-related omics features were selected using random forest followed by Spearman correlation adjusted for confounders. From these, CVD-related omics features were selected using conditional logistic regression. Finally, 29 (for OCs) and 12 (for PFAS) unique features associated with POPs and CVD. One omics subpattern, driven by lipids and inflammatory proteins, associated with MI (OR = 2.03; 95% CI = 1.47; 2.79), OCs, age, and BMI, and correlated negatively with PFAS. Another subpattern, driven by carnitines, associated with stroke (OR = 1.55; 95% CI = 1.16; 2.09), OCs, and age, but not with PFAS. This may imply that OCs and PFAS associate with different omics patterns with opposite effects on CVD risk, but more research is needed to disentangle potential modifications by other factors.

Comprehensive analyses of circulating cardiometabolic proteins and objective measures of fat mass.

BACKGROUND: The underlying molecular pathways for the effect of excess fat mass on cardiometabolic diseases is not well understood. Since body mass index is a suboptimal measure of body fat content, we investigated the relationship of fat mass measured by dual-energy X-ray absorptiometry with circulating cardiometabolic proteins. METHODS: We used data from a population-based cohort of 4950 Swedish women (55-85 years), divided into discovery and replication samples; 276 proteins were assessed with three Olink Proseek Multiplex panels. We used random forest to identify the most relevant biomarker candidates related to fat mass index (FMI), multivariable linear regression to further investigate the associations between FMI characteristics and circulating proteins adjusted for potential confounders, and principal component analysis (PCA) for the detection of common covariance patterns among the proteins. RESULTS: Total FMI was associated with 66 proteins following adjustment for multiple testing in discovery and replication multivariable analyses. Five proteins not previously associated with body size were associated with either lower FMI (calsyntenin-2 (CLSTN2), kallikrein-10 (KLK10)), or higher FMI (scavenger receptor cysteine-rich domain-containing group B protein (SSC4D), trem-like transcript 2 protein (TLT-2), and interleukin-6 receptor subunit alpha (IL-6RA)). PCA provided an efficient summary of the main variation in FMI-related circulating proteins involved in glucose and lipid metabolism, appetite regulation, adipocyte differentiation, immune response and inflammation. Similar patterns were observed for regional fat mass measures. CONCLUSIONS: This is the first large study showing associations between fat mass and circulating cardiometabolic proteins. Proteins not previously linked to body size are implicated in modulation of postsynaptic signals, inflammation, and carcinogenesis.

IBS randomized study: FODMAPs alter bile acids, phenolic- and tryptophan metabolites, while gluten modifies lipids.

Diet is considered a culprit for symptoms in irritable bowel syndrome (IBS), although the mechanistic understanding of underlying causes is lacking. Metabolomics, i.e., the analysis of metabolites in biological samples may offer a diet-responsive fingerprint for IBS. Our aim was to explore alterations in the plasma metabolome after interventions with fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (FODMAPs) or gluten versus control in IBS, and to relate such alterations to symptoms. People with IBS (n = 110) were included in a double-blind, randomized, crossover study with 1-wk provocations of FODMAPs, gluten, or placebo. Symptoms were evaluated with the IBS severity scoring system (IBS-SSS). Untargeted metabolomics was performed on plasma samples using LC-qTOF-MS. Discovery of metabolite alterations by treatment was performed using random forest followed by linear mixed modeling. Associations were studied using Spearman correlation. The metabolome was affected by FODMAP [classification rate (CR) 0.88, P < 0.0001], but less by gluten intake CR 0.72, P = 0.01). FODMAP lowered bile acids, whereas phenolic-derived metabolites and 3-indolepropionic acid (IPA) were higher compared with placebo. IPA and some unidentified metabolites correlated weakly to abdominal pain and quality of life. Gluten affected lipid metabolism weakly, but with no interpretable relationship to IBS. FODMAP affected gut microbial-derived metabolites relating to positive health outcomes. IPA and unknown metabolites correlated weakly to IBS severity. Minor symptom worsening by FODMAP intake must be weighed against general positive health aspects of FODMAP. The gluten intervention affected lipid metabolism weakly with no interpretable association to IBS severity. Registration: www.clinicaltrials.gov as NCT03653689.NEW & NOTEWORTHY In irritable bowel syndrome (IBS), fermentable oligo-, di-, monosaccharides, and polyols (FODMAPs) affected microbial-derived metabolites relating to positive health outcomes such as reduced risk of colon cancer, inflammation, and type 2 diabetes, as shown in previous studies. The minor IBS symptom induction by FODMAP intake must be weighed against the positive health aspects of FODMAP consumption. Gluten affected lipids weakly with no association to IBS severity.

Serum metabolites associated with wholegrain consumption using nontargeted metabolic profiling: a discovery and reproducibility study.

PURPOSE: To identify fasting serum metabolites associated with WG intake in a free-living population adjusted for potential confounders. METHODS: We selected fasting serum samples at baseline from a subset (n = 364) of the prospective population-based Kuopio Ischaemic Heart Disease Risk Factor Study (KIHD) cohort. The samples were analyzed using nontargeted metabolomics with liquid chromatography coupled with mass spectrometry (LC-MS). Association with WG intake was investigated using both random forest followed by linear regression adjusted for age, BMI, smoking, physical activity, energy and alcohol consumption, and partial Spearman correlation adjusted for the same covariates. Features selected by any of these models were shortlisted for annotation. We then checked if we could replicate the findings in an independent subset from the same cohort (n = 200). RESULTS: Direct associations were observed between WG intake and pipecolic acid betaine, tetradecanedioic acid, four glucuronidated alkylresorcinols (ARs), and an unknown metabolite both in discovery and replication cohorts. The associations remained significant (FDR<0.05) even after adjustment for the confounders in both cohorts. Sinapyl alcohol was positively correlated with WG intake in both cohorts after adjustment for the confounders but not in linear models in the replication cohort. Some microbial metabolites, such as indolepropionic acid, were positively correlated with WG intake in the discovery cohort, but the correlations were not replicated in the replication cohort. CONCLUSIONS: The identified associations between WG intake and the seven metabolites after adjusting for confounders in both discovery and replication cohorts suggest the potential of these metabolites as robust biomarkers of WG consumption.

Modest Conformity Between Self-Reporting of Bristol Stool Form and Fecal Consistency Measured by Stool Water Content in Irritable Bowel Syndrome and a FODMAP and Gluten Trial.

INTRODUCTION: Altered bowel habits constitute a criterion of irritable bowel syndrome (IBS), with the Bristol Stool Form Scale (BSFS) as the recommended tool for assessment of fecal consistency. However, BSFS is devoid of a comprehensive objective evaluation in subjects with IBS. Therefore, we aimed to evaluate the concordance between subjective reporting of BSFS and objective stool water content in subjects with IBS. Furthermore, we evaluated whether intake of fermentable oligosaccharides, disaccharides, monosaccharides, and polyols (FODMAPs) or gluten would affect stool water content. METHODS: Data from a previous crossover trial in IBS with 1-week provocations of FODMAPs, gluten, or placebo were subanalyzed. After each intervention, fecal consistency was subjectively assessed using the BSFS and stool samples were collected. The stool water content was analyzed, where ≤68.5% water content was classified as hard stool, while ≥78% was classified as diarrhea. RESULTS: BSFS correlated to stool water content ( r = 0.36, P < 0.0001). The BSFS score increased in parallel with increasing water content, but with considerable overlap between BSFS scores. Stool water content differed between the BSFS categories 1-2, 3-5, and 6-7 (hard, normal, and loose, respectively) ( P < 0.0001). For BSFS categories 1-2, 77% had water content ≤68.5%, whereas for BSFS categories 6-7, 52% had water content ≥78%. There was no difference in stool water content after consumption of FODMAPs, gluten, or placebo ( P = 0.94). DISCUSSION: Subjective reporting of BSFS conforms only modestly with stool water content in IBS, warranting caution when subtyping IBS according to the BSFS. High intake of FODMAPs and gluten does not affect stool water content.

Metabolic Disturbances in Urinary and Plasma Samples from Seven Different Systemic Autoimmune Diseases Detected by HPLC-ESI-QTOF-MS.

Systemic autoimmune diseases (SADs) are characterized by dysfunctioning of the immune system, which causes damage in several tissues and organs. Among these pathologies are systemic lupus erythematosus (SLE), systemic sclerosis or scleroderma, Sjögren's syndrome, rheumatoid arthritis, primary antiphospholipid syndrome (PAPS), mixed connective tissue disease (MCTD), and undifferentiated connective tissue disease (UCTD). Early diagnosis is difficult due to similarity in symptoms, signs, and clinical test results. Hence, our aim was to search for differentiating metabolites of these diseases in plasma and urine samples. We performed metabolomic profiling by liquid chromatography-mass spectrometry (LC-MS) of samples from 228 SADs patients and 55 healthy volunteers. Multivariate PLS models were applied to investigate classification accuracies and identify metabolites differentiating SADs and healthy controls. Furthermore, we specifically investigated UCTD against the other SADs. PLS models were able to classify most SADs vs healthy controls (area under the roc curve (AUC) > 0.7), with the exception of MCTD and PAPS. Differentiating metabolites consisted predominantly of unsaturated fatty acids, acylglycines, acylcarnitines, and amino acids. In accordance with the difficulties in defining UCTD, the UCTD metabolome did not differentiate well from the other SADs. However, most UCTD cases were classified as SLE, suggesting that metabolomics may provide a tool to reassess UCTD diagnosis into other conditions for more well-informed therapeutic strategies.

Variable selection and validation in multivariate modelling.

MOTIVATION: Validation of variable selection and predictive performance is crucial in construction of robust multivariate models that generalize well, minimize overfitting and facilitate interpretation of results. Inappropriate variable selection leads instead to selection bias, thereby increasing the risk of model overfitting and false positive discoveries. Although several algorithms exist to identify a minimal set of most informative variables (i.e. the minimal-optimal problem), few can select all variables related to the research question (i.e. the all-relevant problem). Robust algorithms combining identification of both minimal-optimal and all-relevant variables with proper cross-validation are urgently needed. RESULTS: We developed the MUVR algorithm to improve predictive performance and minimize overfitting and false positives in multivariate analysis. In the MUVR algorithm, minimal variable selection is achieved by performing recursive variable elimination in a repeated double cross-validation (rdCV) procedure. The algorithm supports partial least squares and random forest modelling, and simultaneously identifies minimal-optimal and all-relevant variable sets for regression, classification and multilevel analyses. Using three authentic omics datasets, MUVR yielded parsimonious models with minimal overfitting and improved model performance compared with state-of-the-art rdCV. Moreover, MUVR showed advantages over other variable selection algorithms, i.e. Boruta and VSURF, including simultaneous variable selection and validation scheme and wider applicability. AVAILABILITY AND IMPLEMENTATION: Algorithms, data, scripts and tutorial are open source and available as an R package ('MUVR') at https://gitlab.com/CarlBrunius/MUVR.git. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

An LC-QToF MS based method for untargeted metabolomics of human fecal samples.

INTRODUCTION: Consensus in sample preparation for untargeted human fecal metabolomics is lacking. OBJECTIVES: To obtain sample preparation with broad metabolite coverage for high-throughput LC-MS. METHODS: Extraction solvent, solvent ratio and fresh frozen-vs-lyophilized samples were evaluated by metabolite feature quality. RESULTS: Methanol at 5 mL per g wet feces provided a wide metabolite coverage with optimal balance between signal intensity and saturation for both fresh frozen and lyophilized samples. Lyophilization did not affect SCFA and is recommended because of convenience in normalizing to dry matter. CONCLUSION: The suggested sample preparation is simple, efficient and suitable for large-scale human fecal metabolomics.

Dental home care in dogs - a questionnaire study among Swedish dog owners, veterinarians and veterinary nurses.

BACKGROUND: Periodontal disease remains one of the dog's most common health issues, even though it is largely preventable by tooth brushing. Implementation of daily tooth brushing would not only improve animal welfare, but also reduce veterinary costs for the owner. There is a paucity of studies investigating attitudes, opinions and practices of dog owners, veterinarians, and veterinary nurses regarding preventative dental home care in dogs. The objective of this study was to investigate these parameters in Sweden, thereby providing a basis for improved prophylactic strategies. METHODS: Validated questionnaire surveys were distributed to all Swedish dog owners (n = 209,263), veterinarians (n = 3657) and veterinary nurses (n = 1650) with e-mail addresses in the national registry. The response rates were 32% for dog owners and veterinarians, and 38% for veterinary nurses. The survey questions concerned attitudes, opinions and practices regarding dental home care, including whether dog owners received information concerning dental home care or not, and if this information resulted in implementation. RESULTS: Attitudes, opinions and practices regarding dental home care are presented for Swedish dog owners, veterinarians, and veterinary nurses. A fundamental finding was that the absolute majority of Swedish dog owners do not perform adequate prophylactic dental home care. Considerable discrepancies were identified in the opinions of veterinary health practitioners and dog owners regarding attitudes towards dental home care and conveying of information. Several areas for improvement in the communication between dog owners and veterinary health practitioners concerning dental home care were identified. CONCLUSION: Our results illustrates the need for validated methods to increase dog owner compliance with dental home care recommendations. We also see a need of further education, regarding canine dental home care, among veterinarians, veterinary nurses, and dog owners. The results from this unique study constitute an important foundation for future development of prophylactic strategies, with the ultimate goal to improve dental health, and thereby animal welfare, in dogs.

Joint Analysis of Metabolite Markers of Fish Intake and Persistent Organic Pollutants in Relation to Type 2 Diabetes Risk in Swedish Adults.

BACKGROUND: There is conflicting evidence regarding the association between fish intake and type 2 diabetes (T2D) incidence, possibly owing to measurement errors in self-reported intake and coexposure to persistent organic pollutants (POPs) present in fish. OBJECTIVE: The aim of this study was to identify plasma metabolites associated with fish intake and to assess their association with T2D risk, independently of POPs, in Swedish adults. METHODS: In a case-control study nested in the Swedish Västerbotten Intervention Programme, fasting plasma samples from 421 matched T2D case-control pairs of men and women aged 30-60 y at baseline and 10-y follow-up samples from a subset of 149 pairs were analyzed using untargeted metabolomics. Moreover, 16 plasma POPs were analyzed for the 149 pairs who had repeated samples available. Fish-related plasma metabolites were identified using multivariate modelling and partial correlation analysis. Reproducibility of metabolites and metabolite patterns, derived via principal component analysis (PCA), was assessed by intraclass correlation. A unique component of metabolites unrelated to POPs was dissected by integrating metabolites and POPs using 2-way orthogonal partial least squares regression. ORs of T2D were estimated using conditional logistic regression. RESULTS: We identified 31 metabolites associated with fish intake that had poor to good reproducibility. A PCA-derived metabolite pattern strongly correlated with fish intake (ρ = 0.37, P < 0.001) but showed no association with T2D risk. Integrating fish-related metabolites and POPs led to a unique metabolite component independent of POPs, which tended to be inversely associated with T2D risk (OR: 0.75; 95% CI: 0.54, 1.02, P = 0.07). This component mainly consisted of metabolites reflecting fatty fish intake. CONCLUSIONS: Our results suggest that fatty fish intake may be beneficial for T2D prevention, after removing the counteractive effects of coexposure to POPs in Swedish adults. Integrating metabolite markers and POP exposures appears a promising approach to advance the understanding of associations between fish intake and T2D incidence.

The 1H NMR serum metabolomics response to a two meal challenge: a cross-over dietary intervention study in healthy human volunteers.

BACKGROUND: Metabolomics represents a powerful tool for exploring modulation of the human metabolome in response to food intake. However, the choice of multivariate statistical approach is not always evident, especially for complex experimental designs with repeated measurements per individual. Here we have investigated the serum metabolic responses to two breakfast meals: an egg and ham based breakfast and a cereal based breakfast using three different multivariate approaches based on the Projections to Latent Structures framework. METHODS: In a cross over design, 24 healthy volunteers ate the egg and ham breakfast and cereal breakfast on four occasions each. Postprandial serum samples were subjected to metabolite profiling using 1H nuclear magnetic resonance spectroscopy and metabolites were identified using 2D nuclear magnetic resonance spectroscopy. Metabolic profiles were analyzed using Orthogonal Projections to Latent Structures with Discriminant Analysis and Effect Projections and ANOVA-decomposed Projections to Latent Structures. RESULTS: The Orthogonal Projections to Latent Structures with Discriminant Analysis model correctly classified 92 and 90% of the samples from the cereal breakfast and egg and ham breakfast, respectively, but confounded dietary effects with inter-personal variability. Orthogonal Projections to Latent Structures with Effect Projections removed inter-personal variability and performed perfect classification between breakfasts, however at the expense of comparing means of respective breakfasts instead of all samples. ANOVA-decomposed Projections to Latent Structures managed to remove inter-personal variability and predicted 99% of all individual samples correctly. Proline, tyrosine, and N-acetylated amino acids were found in higher concentration after consumption of the cereal breakfast while creatine, methanol, and isoleucine were found in higher concentration after the egg and ham breakfast. CONCLUSIONS: Our results demonstrate that the choice of statistical method will influence the results and adequate methods need to be employed to manage sample dependency and repeated measurements in cross-over studies. In addition, 1H nuclear magnetic resonance serum metabolomics could reproducibly characterize postprandial metabolic profiles and identify discriminatory metabolites largely reflecting dietary composition. TRIAL REGISTRATION: Registered with ClinicalTrials.gov, identifier: NCT02039596 . Date of registration: January 17, 2014.

Impact in Plasma Metabolome as Effect of Lifestyle Intervention for Weight-Loss Reveals Metabolic Benefits in Metabolically Healthy Obese Women.

Little is known regarding metabolic benefits of weight loss (WL) on the metabolically healthy obese (MHO) patients. We aimed to examine the impact of a lifestyle weight loss (LWL) treatment on the plasma metabolomic profile in MHO individuals. Plasma samples from 57 MHO women allocated to an intensive LWL treatment group (TG, hypocaloric Mediterranean diet and regular physical activity, n = 30) or to a control group (CG, general recommendations of a healthy diet and physical activity, n = 27) were analyzed using an untargeted 1H NMR metabolomics approach at baseline, after 3 months (intervention), and 12 months (follow-up). The impact of the LWL intervention on plasma metabolome was statistically significant at 3 months but not at follow-up and included higher levels of formate and phosphocreatine and lower levels of LDL/VLDL (signals) and trimethylamine in the TG. These metabolites were also correlated with WL. Higher myo-inositol, methylguanidine, and 3-hydroxybutyrate, and lower proline, were also found in the TG; higher levels of hippurate and asparagine, and lower levels of 2-hydroxybutyrate and creatine, were associated with WL. The current findings suggest that an intensive LWL treatment, and the consequent WL, leads to an improved plasma metabolic profile in MHO women through its impact on energy, amino acid, lipoprotein, and microbial metabolism.

Plasma metabolites associated with type 2 diabetes in a Swedish population: a case-control study nested in a prospective cohort.

AIMS/HYPOTHESIS: The aims of the present work were to identify plasma metabolites that predict future type 2 diabetes, to investigate the changes in identified metabolites among individuals who later did or did not develop type 2 diabetes over time, and to assess the extent to which inclusion of predictive metabolites could improve risk prediction. METHODS: We established a nested case-control study within the Swedish prospective population-based Västerbotten Intervention Programme cohort. Using untargeted liquid chromatography-MS metabolomics, we analysed plasma samples from 503 case-control pairs at baseline (a median time of 7 years prior to diagnosis) and samples from a subset of 187 case-control pairs at 10 years of follow-up. Discriminative metabolites between cases and controls at baseline were optimally selected using a multivariate data analysis pipeline adapted for large-scale metabolomics. Conditional logistic regression was used to assess associations between discriminative metabolites and future type 2 diabetes, adjusting for several known risk factors. Reproducibility of identified metabolites was estimated by intra-class correlation over the 10 year period among the subset of healthy participants; their systematic changes over time in relation to diagnosis among those who developed type 2 diabetes were investigated using mixed models. Risk prediction performance of models made from different predictors was evaluated using area under the receiver operating characteristic curve, discrimination improvement index and net reclassification index. RESULTS: We identified 46 predictive plasma metabolites of type 2 diabetes. Among novel findings, phosphatidylcholines (PCs) containing odd-chain fatty acids (C19:1 and C17:0) and 2-hydroxyethanesulfonate were associated with the likelihood of developing type 2 diabetes; we also confirmed previously identified predictive biomarkers. Identified metabolites strongly correlated with insulin resistance and/or beta cell dysfunction. Of 46 identified metabolites, 26 showed intermediate to high reproducibility among healthy individuals. Moreover, PCs with odd-chain fatty acids, branched-chain amino acids, 3-methyl-2-oxovaleric acid and glutamate changed over time along with disease progression among diabetes cases. Importantly, we found that a combination of five of the most robustly predictive metabolites significantly improved risk prediction if added to models with an a priori defined set of traditional risk factors, but only a marginal improvement was achieved when using models based on optimally selected traditional risk factors. CONCLUSIONS/INTERPRETATION: Predictive metabolites may improve understanding of the pathophysiology of type 2 diabetes and reflect disease progression, but they provide limited incremental value in risk prediction beyond optimal use of traditional risk factors.

Prediction and modeling of pre-analytical sampling errors as a strategy to improve plasma NMR metabolomics data.

MOTIVATION: Biobanks are important infrastructures for life science research. Optimal sample handling regarding e.g. collection and processing of biological samples is highly complex, with many variables that could alter sample integrity and even more complex when considering multiple study centers or using legacy samples with limited documentation on sample management. Novel means to understand and take into account such variability would enable high-quality research on archived samples. RESULTS: This study investigated whether pre-analytical sample variability could be predicted and reduced by modeling alterations in the plasma metabolome, measured by NMR, as a function of pre-centrifugation conditions (1-36 h pre-centrifugation delay time at 4 °C and 22 °C) in 16 individuals. Pre-centrifugation temperature and delay times were predicted using random forest modeling and performance was validated on independent samples. Alterations in the metabolome were modeled at each temperature using a cluster-based approach, revealing reproducible effects of delay time on energy metabolism intermediates at both temperatures, but more pronounced at 22 °C. Moreover, pre-centrifugation delay at 4 °C resulted in large, specific variability at 3 h, predominantly of lipids. Pre-analytical sample handling error correction resulted in significant improvement of data quality, particularly at 22 °C. This approach offers the possibility to predict pre-centrifugation delay temperature and time in biobanked samples before use in costly downstream applications. Moreover, the results suggest potential to decrease the impact of undesired, delay-induced variability. However, these findings need to be validated in multiple, large sample sets and with analytical techniques covering a wider range of the metabolome, such as LC-MS. AVAILABILITY AND IMPLEMENTATION: The sampleDrift R package is available at https://gitlab.com/CarlBrunius/sampleDrift. CONTACT: carl.brunius@chalmers.se. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Adjusting for covariates and assessing modeling fitness in machine learning using MUVR2.

Motivation: Machine learning (ML) methods are frequently used in Omics research to examine associations between molecular data and for example exposures and health conditions. ML is also used for feature selection to facilitate biological interpretation. Our previous MUVR algorithm was shown to generate predictions and variable selections at state-of-the-art performance. However, a general framework for assessing modeling fitness is still lacking. In addition, enabling to adjust for covariates is a highly desired, but largely lacking trait in ML. We aimed to address these issues in the new MUVR2 framework. Results: The MUVR2 algorithm was developed to include the regularized regression framework elastic net in addition to partial least squares and random forest modeling. Compared with other cross-validation strategies, MUVR2 consistently showed state-of-the-art performance, including variable selection, while minimizing overfitting. Testing on simulated and real-world data, we also showed that MUVR2 allows for the adjustment for covariates using elastic net modeling, but not using partial least squares or random forest. Availability and implementation: Algorithms, data, scripts, and a tutorial are open source under GPL-3 license and available in the MUVR2 R package at https://github.com/MetaboComp/MUVR2.

Fasting plasma metabolites reflecting meat consumption and their associations with incident type 2 diabetes in two Swedish cohorts.

BACKGROUND: Consumption of processed red meat has been associated with increased risk of developing type 2 diabetes (T2D), but challenges in dietary assessment call for objective intake biomarkers. OBJECTIVES: This study aimed to investigate metabolite biomarkers of meat intake and their associations with T2D risk. METHODS: Fasting plasma samples were collected from a case-control study nested within Västerbotten Intervention Program (VIP) (214 females and 189 males) who developed T2D after a median follow-up of 7 years. Panels of biomarker candidates reflecting the consumption of total, processed, and unprocessed red meat and poultry were selected from the untargeted metabolomics data collected on the controls. Observed associations were then replicated in Swedish Mammography clinical subcohort in Uppsala (SMCC) (n = 4457 females). Replicated metabolites were assessed for potential association with T2D risk using multivariable conditional logistic regression in the discovery and Cox regression in the replication cohorts. RESULTS: In total, 15 metabolites were associated with ≥1 meat group in both cohorts. Acylcarnitines 8:1, 8:2, 10:3, reflecting higher processed meat intake [r > 0.22, false discovery rate (FDR) < 0.001 for VIP and r > 0.05; FDR < 0.001 for SMCC) were consistently associated with higher T2D risk in both data sets. Conversely, lysophosphatidylcholine 17:1 and phosphatidylcholine (PC) 15:0/18:2 were associated with lower processed meat intake (r < -0.12; FDR < 0.023, for VIP and r < -0.05; FDR < 0.001, for SMCC) and with lower T2D risk in both data sets, except for PC 15:0/18:2, which was significant only in the VIP cohort. All associations were attenuated after adjustment for BMI (kg/m2). CONCLUSIONS: Consistent associations of biomarker candidates involved in lipid metabolism between higher processed red meat intake with higher T2D risk and between those reflecting lower intake with the lower risk may suggest a relationship between processed meat intake and higher T2D risk. However, attenuated associations after adjusting for BMI indicates that such a relationship may at least partly be mediated or confounded by BMI.

Dietary patterns, untargeted metabolite profiles and their association with colorectal cancer risk.

We investigated data-driven and hypothesis-driven dietary patterns and their association to plasma metabolite profiles and subsequent colorectal cancer (CRC) risk in 680 CRC cases and individually matched controls. Dietary patterns were identified from combined exploratory/confirmatory factor analysis. We assessed association to LC-MS metabolic profiles by random forest regression and to CRC risk by multivariable conditional logistic regression. Principal component analysis was used on metabolite features selected to reflect dietary exposures. Component scores were associated to CRC risk and dietary exposures using partial Spearman correlation. We identified 12 data-driven dietary patterns, of which a breakfast food pattern showed an inverse association with CRC risk (OR per standard deviation increase 0.89, 95% CI 0.80-1.00, p = 0.04). This pattern was also inversely associated with risk of distal colon cancer (0.75, 0.61-0.96, p = 0.01) and was more pronounced in women (0.69, 0.49-0.96, p = 0.03). Associations between meat, fast-food, fruit soup/rice patterns and CRC risk were modified by tumor location in women. Alcohol as well as fruit and vegetables associated with metabolite profiles (Q2 0.22 and 0.26, respectively). One metabolite reflecting alcohol intake associated with increased CRC risk, whereas three metabolites reflecting fiber, wholegrain, and fruit and vegetables associated with decreased CRC risk.