peakPantheR, an R package for large-scale targeted extraction and integration of annotated metabolic features in LC-MS profiling datasets

Adipose tissue-derived metabolite risk scores and risk for type 2 diabetes in South Asians

BACKGROUND

South Asians are at higher risk for type 2 diabetes (T2D) than many other race/ethnic groups. Ectopic adiposity, specifically hepatic steatosis and visceral fat may partially explain this. Our objective was to derive metabolite risk scores for ectopic adiposity and assess associations with incident T2D in South Asians.

METHODS

We examined 550 participants in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) cohort study aged 40-84 years without known cardiovascular disease or T2D and with metabolomic data. Computed tomography scans at baseline assessed hepatic attenuation and visceral fat area, and fasting serum specimens at baseline and after 5 years assessed T2D. LC-MS-based untargeted metabolomic analysis was performed followed by targeted integration and reporting of known signals. Elastic net regularized linear regression analyses was used to derive risk scores for hepatic steatosis and visceral fat using weighted coefficients. Logistic regression models associated metabolite risk score and incident T2D, adjusting for age, gender, study site, BMI, physical activity, diet quality, energy intake and use of cholesterol-lowering medication.

RESULTS

Average age of participants was 55 years, 36% women with an average body mass index (BMI) of 25 kg/m2 and 6% prevalence of hepatic steatosis, with 47 cases of incident T2D at 5 years. There were 445 metabolites of known identity. Of these, 313 metabolites were included in the MET-Visc score and 267 in the MET-Liver score. In most fully adjusted models, MET-Liver (OR 2.04 [95% CI 1.38, 3.03]) and MET-Visc (OR 2.80 [1.75, 4.46]) were associated with higher odds of T2D. These associations remained significant after adjustment for measured adiposity.

CONCLUSIONS

Metabolite risk scores for intrahepatic fat and visceral fat were strongly related to incident T2D independent of measured adiposity. Use of these biomarkers to target risk stratification may help capture pre-clinical metabolic abnormalities.

Uncoupling of behavioral and metabolic 24-h rhythms in reindeer

Reindeer in the Arctic seasonally suppress daily circadian patterns of behavior present in most animals.1 In humans and mice, even when all daily behavioral and environmental influences are artificially suppressed, robust endogenous rhythms of metabolism governed by the circadian clock persist and are essential to health.2,3 Disrupted rhythms foster metabolic disorders and weight gain.4 To understand circadian metabolic organization in reindeer, we performed behavioral measurements and untargeted metabolomics from blood plasma samples taken from Eurasian tundra reindeer (Rangifer tarandus tarandus) across 24 h at 2-h intervals in four seasons. Our study confirmed the absence of circadian rhythms of behavior under constant darkness in the Arctic winter and constant daylight in the Arctic summer, as reported by others.1 We detected and measured the intensity of 893 metabolic features in all plasma samples using untargeted ultra-high-performance liquid chromatography-mass spectrometry (UPLC-MS). A core group of metabolites (66/893 metabolic features) consistently displayed 24-h rhythmicity. Most metabolites displayed a robust 24-h rhythm in winter and spring but were arrhythmic in summer and fall. Half of all measured metabolites displayed ultradian sleep-wake dependence in summer. Irrespective of the arrhythmic behavior, metabolism is rhythmic (24 h) in seasons of low food availability, potentially favoring energy efficiency. In seasons of food abundance, 24-h rhythmicity in metabolism is drastically reduced, again irrespective of behavioral rhythms, potentially fostering weight gain.

Synthesis and Characterization of Organic Peroxides from Monoterpene-Derived Criegee Intermediates in Secondary Organic Aerosol

Ozonolysis of alkenes is known to produce reactive intermediates─stabilized Criegee intermediates (SCIs), and their subsequent bimolecular reactions with various carboxylic acids can form α-acyloxyalkyl hydroperoxides (AAHPs), which is considered a major class of organic peroxides in secondary organic aerosol (SOA). Despite their atmospheric and health importance, the molecular-level identification of organic peroxides in atmospheric aerosols is highly challenging, preventing further assessment of their environmental fate. Here, we synthesize 20 atmospherically relevant AAHPs through liquid-phase ozonolysis, in which two types of monoterpene-derived SCIs from either α-pinene or 3-carene are scavenged by 10 different carboxylic acids to form AAHPs with diverse structures. These AAHPs are identified individually by liquid chromatography coupled with high-resolution mass spectrometry. AAHPs were previously thought to decompose quickly in an aqueous environment such as cloud droplets, but we demonstrate here that AAHPs hydrolysis rates are highly compound-dependent with rate constants differing by 2 orders of magnitude. In contrast, the aqueous-phase formation rate constants between SCI and various carboxylic acids vary only within a factor of 2-3. Finally, we identified two of the 20 synthesized AAHPs in α-pinene SOA and two in 3-carene SOA, contributing ∼0.3% to the total SOA mass. Our results improve the current molecular-level understanding of organic peroxides and are useful for a more accurate assessment of their environmental fate and health impact.

Fecal microbiota transplantation plus anti-PD-1 immunotherapy in advanced melanoma: a phase I trial

Fecal microbiota transplantation (FMT) represents a potential strategy to overcome resistance to immune checkpoint inhibitors in patients with refractory melanoma; however, the role of FMT in first-line treatment settings has not been evaluated. We conducted a multicenter phase I trial combining healthy donor FMT with the PD-1 inhibitors nivolumab or pembrolizumab in 20 previously untreated patients with advanced melanoma. The primary end point was safety. No grade 3 adverse events were reported from FMT alone. Five patients (25%) experienced grade 3 immune-related adverse events from combination therapy. Key secondary end points were objective response rate, changes in gut microbiome composition and systemic immune and metabolomics analyses. The objective response rate was 65% (13 of 20), including four (20%) complete responses. Longitudinal microbiome profiling revealed that all patients engrafted strains from their respective donors; however, the acquired similarity between donor and patient microbiomes only increased over time in responders. Responders experienced an enrichment of immunogenic and a loss of deleterious bacteria following FMT. Avatar mouse models confirmed the role of healthy donor feces in increasing anti-PD-1 efficacy. Our results show that FMT from healthy donors is safe in the first-line setting and warrants further investigation in combination with immune checkpoint inhibitors. ClinicalTrials.gov identifier NCT03772899 .

Pathobionts in the tumour microbiota predict survival following resection for colorectal cancer

BACKGROUND AND AIMS

The gut microbiota is implicated in the pathogenesis of colorectal cancer (CRC). We aimed to map the CRC mucosal microbiota and metabolome and define the influence of the tumoral microbiota on oncological outcomes.

METHODS

A multicentre, prospective observational study was conducted of CRC patients undergoing primary surgical resection in the UK (n = 74) and Czech Republic (n = 61). Analysis was performed using metataxonomics, ultra-performance liquid chromatography-mass spectrometry (UPLC-MS), targeted bacterial qPCR and tumour exome sequencing. Hierarchical clustering accounting for clinical and oncological covariates was performed to identify clusters of bacteria and metabolites linked to CRC. Cox proportional hazards regression was used to ascertain clusters associated with disease-free survival over median follow-up of 50 months.

RESULTS

Thirteen mucosal microbiota clusters were identified, of which five were significantly different between tumour and paired normal mucosa. Cluster 7, containing the pathobionts Fusobacterium nucleatum and Granulicatella adiacens, was strongly associated with CRC (P_FDR = 0.0002). Additionally, tumoral dominance of cluster 7 independently predicted favourable disease-free survival (adjusted p = 0.031). Cluster 1, containing Faecalibacterium prausnitzii and Ruminococcus gnavus, was negatively associated with cancer (P_FDR = 0.0009), and abundance was independently predictive of worse disease-free survival (adjusted p = 0.0009). UPLC-MS analysis revealed two major metabolic (Met) clusters. Met 1, composed of medium chain (MCFA), long-chain (LCFA) and very long-chain (VLCFA) fatty acid species, ceramides and lysophospholipids, was negatively associated with CRC (P_FDR = 2.61 × 10^-11); Met 2, composed of phosphatidylcholine species, nucleosides and amino acids, was strongly associated with CRC (P_FDR = 1.30 × 10^-12), but metabolite clusters were not associated with disease-free survival (p = 0.358). An association was identified between Met 1 and DNA mismatch-repair deficiency (p = 0.005). FBXW7 mutations were only found in cancers predominant in microbiota cluster 7.

CONCLUSIONS

Networks of pathobionts in the tumour mucosal niche are associated with tumour mutation and metabolic subtypes and predict favourable outcome following CRC resection. Video Abstract.

The gut microbiota and metabolome are associated with diminished COVID-19 vaccine-induced antibody responses in immunosuppressed inflammatory bowel disease patients

BACKGROUND

Patients with inflammatory bowel disease (IBD) treated with anti-TNF therapy exhibit attenuated humoral immune responses to vaccination against SARS-CoV-2. The gut microbiota and its functional metabolic output, which are perturbed in IBD, play an important role in shaping host immune responses. We explored whether the gut microbiota and metabolome could explain variation in anti-SARS-CoV-2 vaccination responses in immunosuppressed IBD patients.

METHODS

Faecal and serum samples were prospectively collected from infliximab-treated patients with IBD in the CLARITY-IBD study undergoing vaccination against SARS-CoV-2. Antibody responses were measured following two doses of either ChAdOx1 nCoV-19 or BNT162b2 vaccine. Patients were classified as having responses above or below the geometric mean of the wider CLARITY-IBD cohort. 16S rRNA gene amplicon sequencing, nuclear magnetic resonance (NMR) spectroscopy and bile acid profiling with ultra-high-performance liquid chromatography mass spectrometry (UHPLC-MS) were performed on faecal samples. Univariate, multivariable and correlation analyses were performed to determine gut microbial and metabolomic predictors of response to vaccination.

FINDINGS

Forty-three infliximab-treated patients with IBD were recruited (30 Crohn's disease, 12 ulcerative colitis, 1 IBD-unclassified; 26 with concomitant thiopurine therapy). Eight patients had evidence of prior SARS-CoV-2 infection. Seventeen patients (39.5%) had a serological response below the geometric mean. Gut microbiota diversity was lower in below average responders (p = 0.037). Bilophila abundance was associated with better serological response, while Streptococcus was associated with poorer response. The faecal metabolome was distinct between above and below average responders (OPLS-DA R²X 0.25, R²Y 0.26, Q² 0.15; CV-ANOVA p = 0.038). Trimethylamine, isobutyrate and omega-muricholic acid were associated with better response, while succinate, phenylalanine, taurolithocholate and taurodeoxycholate were associated with poorer response.

INTERPRETATION

Our data suggest that there is an association between the gut microbiota and variable serological response to vaccination against SARS-CoV-2 in immunocompromised patients. Microbial metabolites including trimethylamine may be important in mitigating anti-TNF-induced attenuation of the immune response.

FUNDING

JLA is the recipient of an NIHR Academic Clinical Lectureship (CL-2019-21-502), funded by Imperial College London and The Joyce and Norman Freed Charitable Trust. BHM is the recipient of an NIHR Academic Clinical Lectureship (CL-2019-21-002). The Division of Digestive Diseases at Imperial College London receives financial and infrastructure support from the NIHR Imperial Biomedical Research Centre (BRC) based at Imperial College Healthcare NHS Trust and Imperial College London. Metabolomics studies were performed at the MRC-NIHR National Phenome Centre at Imperial College London; this work was supported by the Medical Research Council (MRC), the National Institute of Health Research (NIHR) (grant number MC_PC_12025) and infrastructure support was provided by the NIHR Imperial Biomedical Research Centre (BRC). The NIHR Exeter Clinical Research Facility is a partnership between the University of Exeter Medical School College of Medicine and Health, and Royal Devon and Exeter NHS Foundation Trust. This project is supported by the National Institute for Health Research (NIHR) Exeter Clinical Research Facility. The views expressed are those of the authors and not necessarily those of the NIHR or the UK Department of Health and Social Care.

Bacterial and metabolic phenotypes associated with inadequate response to ursodeoxycholic acid treatment in primary biliary cholangitis

Primary biliary cholangitis (PBC) is a chronic cholestatic liver disease with ursodeoxycholic acid (UDCA) as first-line treatment. Poor response to UDCA is associated with a higher risk of progressing to cirrhosis, but the underlying mechanisms are unclear. UDCA modulates the composition of primary and bacterial-derived bile acids (BAs). We characterized the phenotypic response to UDCA based on BA and bacterial profiles of PBC patients treated with UDCA. Patients from the UK-PBC cohort (n = 419) treated with UDCA for a minimum of 12-months were assessed using the Barcelona dynamic response criteria. BAs from serum, urine, and feces were analyzed using Ultra-High-Performance Liquid Chromatography-Mass Spectrometry and fecal bacterial composition measured using 16S rRNA gene sequencing. We identified 191 non-responders, 212 responders, and a subgroup of responders with persistently elevated liver biomarkers (n = 16). Responders had higher fecal secondary and tertiary BAs than non-responders and lower urinary bile acid abundances, with the exception of 12-dehydrocholic acid, which was higher in responders. The sub-group of responders with poor liver function showed lower alpha-diversity evenness, lower abundance of fecal secondary and tertiary BAs than the other groups and lower levels of phyla with BA-deconjugation capacity (Actinobacteriota/Actinomycetota, Desulfobacterota, Verrucomicrobiota) compared to responders. UDCA dynamic response was associated with an increased capacity to generate oxo-/epimerized secondary BAs. 12-dehydrocholic acid is a potential biomarker of treatment response. Lower alpha-diversity and lower abundance of bacteria with BA deconjugation capacity might be associated with an incomplete response to treatment in some patients.

Assessing the clinical value of faecal bile acid profiling to predict recurrence in primary Clostridioides difficile infection

BACKGROUND

Factors influencing recurrence risk in primary Clostridioides difficile infection (CDI) are poorly understood, and tools predicting recurrence are lacking. Perturbations in bile acids (BAs) contribute to CDI pathogenesis and may be relevant to primary disease prognosis.

AIMS

To define stool BA dynamics in patients with primary CDI and to explore signatures predicting recurrence METHODS: Weekly stool samples were collected from patients with primary CDI from the last day of anti-CDI therapy until recurrence or, otherwise, through 8 weeks post-completion. Ultra-high performance liquid chromatography-mass spectrometry was used to profile BAs. Stool bile salt hydrolase (BSH) activity was measured to determine primary BA bacterial deconjugation capacity. Multivariate and univariate models were used to define differential BA trajectories in patients with recurrence versus those without, and to assess faecal BAs as predictive markers for recurrence.

RESULTS

Twenty (36%) of 56 patients (median age: 57, 64% male) had recurrence; 80% of recurrences occurred within the first 9 days post-antibiotic treatment. Principal component analysis of stool BA profiles demonstrated clustering by recurrence status and post-treatment timepoint. Longitudinal faecal BA trajectories showed recovery of secondary BAs and their derivatives only in patients without recurrence. BSH activity increased over time only among non-relapsing patients (β = 0.056; likelihood ratio test p = 0.018). A joint longitudinal-survival model identified five stool BAs with area under the receiver operating characteristic curve >0.73 for predicting recurrence within 9 days post-CDI treatment.

CONCLUSIONS

Gut BA metabolism dynamics differ in primary CDI patients between those developing recurrence and those who do not. Individual BAs show promise as potential novel biomarkers to predict CDI recurrence.

Diet Patterns Are Associated with Circulating Metabolites and Lipid Profiles of South Asians in the United States

BACKGROUND

South Asians are at higher risk for cardiometabolic disease than many other racial/ethnic minority groups. Diet patterns in US South Asians have unique components associated with cardiometabolic disease.

OBJECTIVES

We aimed to characterize the metabolites associated with 3 representative diet patterns.

METHODS

We included 722 participants in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) cohort study aged 40-84 y without known cardiovascular disease. Fasting serum specimens and diet and demographic questionnaires were collected at baseline and diet patterns previously generated through principal components analysis. LC-MS-based untargeted metabolomic and lipidomic analysis was conducted with targeted integration of known metabolite and lipid signals. Linear regression models of diet pattern factor score and log-transformed metabolites adjusted for age, sex, caloric intake, and BMI and adjusted for multiple comparisons were performed, followed by elastic net linear regression of significant metabolites.

RESULTS

There were 443 metabolites of known identity extracted from the profiling data. The "animal protein" diet pattern was associated with 61 metabolites and lipids, including glycerophospholipids phosphatidylethanolamine PE(O-16:1/20:4) and/or PE(P-16:0/20:4) (β: 0.13; 95% CI: 0.11, 0.14) and N-acyl phosphatidylethanolamines (NAPEs) NAPE(O-18:1/20:4/18:0) and/or NAPE(P-18:0/20:4/18:0) (β: 0.13; 95% CI: 0.11, 0.14), lysophosphatidylinositol (LPI) (22:6/0:0) (β: 0.14; 95% CI: 0.12, 0.17), and fatty acid (FA) (22:6) (β: 0.15; 95% CI: 0.13, 0.17). The "fried snacks, sweets, high-fat dairy" pattern was associated with 12 lipids, including PC(16:0/22:6) (β: -0.08; 95% CI: -0.09, -0.06) and FA (22:6) (β: 0.14; 95% CI: -0.17, -0.10). The "fruits, vegetables, nuts, and legumes" pattern was associated with 5 metabolites including proline betaine (β: 0.17; 95% CI: 0.09, 0.25) (P < 0.0002).

CONCLUSIONS

Three predominant dietary patterns in US South Asians are associated with circulating metabolites differentiated by lipids including glycerophospholipids and PUFAs and the amino acid proline betaine.

Metabolomic profiling in small vessel disease identifies multiple associations with disease severity

Cerebral small vessel disease is a major cause of vascular cognitive impairment and dementia. There are few treatments, largely reflecting limited understanding of the underlying pathophysiology. Metabolomics can be used to identify novel risk factors to better understand pathogenesis and to predict disease progression and severity. We analysed data from 624 patients with symptomatic cerebral small vessel disease from two prospective cohort studies. Serum samples were collected at baseline and patients underwent MRI scans and cognitive testing at regular intervals with up to 14 years of follow-up. Using ultra-performance liquid chromatography-mass spectrometry and nuclear magnetic resonance spectroscopy, we obtained metabolic and lipidomic profiles from 369 annotated metabolites and 54 764 unannotated features and examined their association with respect to disease severity, assessed using MRI small vessel disease markers, cognition and future risk of all-cause dementia. Our analysis identified 28 metabolites that were significantly associated with small vessel disease imaging markers and cognition. Decreased levels of multiple glycerophospholipids and sphingolipids were associated with increased small vessel disease load as evidenced by higher white matter hyperintensity volume, lower mean diffusivity normalized peak height, greater brain atrophy and impaired cognition. Higher levels of creatine, FA(18:2(OH)) and SM(d18:2/24:1) were associated with increased lacune count, higher white matter hyperintensity volume and impaired cognition. Lower baseline levels of carnitines and creatinine were associated with higher annualized change in peak width of skeletonized mean diffusivity, and 25 metabolites, including lipoprotein subclasses, amino acids and xenobiotics, were associated with future dementia incidence. Our results show multiple distinct metabolic signatures that are associated with imaging markers of small vessel disease, cognition and conversion to dementia. Further research should assess causality and the use of metabolomic screening to improve the ability to predict future disease severity and dementia risk in small vessel disease. The metabolomic profiles may also provide novel insights into disease pathogenesis and help identify novel treatment approaches.

MetEx: A Targeted Extraction Strategy for Improving the Coverage and Accuracy of Metabolite Annotation in Liquid Chromatography-High-Resolution Mass Spectrometry Data

Liquid chromatography-high-resolution mass spectrometry (LC-HRMS) is the most popular platform for untargeted metabolomics studies, but compound annotation is a challenge. In this work, we developed a new LC-HRMS data-targeted extraction method called MetEx for metabolite annotation. MetEx contains the retention time (t_R), MS1, and MS2 information of 30 620 metabolites from freely available spectral databases, including MoNA and KEGG. The t_R values of 95.4% of the compounds in our database were calculated by the GNN-RT model. The MS2 spectra of 39.4% compounds were also predicted using CFM-ID. MetEx was initially examined on a mixture of 634 standards, considering chemical coverage and accurate metabolite assignment, and later applied to human plasma (NIST SRM 1950), human urine, HepG2 cells, mouse liver tissue, and mouse feces. MetEx correctly assigned 252 out of 253 standards detected in our instruments. The platform also provided 8.0-44.2% more compounds in the biological samples compared to XCMS, MS-DIAL, and MZmine 2. MetEx is implemented and visualized in R and freely available at http://www.metaboex.cn/MetEx.

Automated Cancer Diagnostics via Analysis of Optical and Chemical Images by Deep and Shallow Learning

Optical microscopy has long been the gold standard to analyse tissue samples for the diagnostics of various diseases, such as cancer. The current diagnostic workflow is time-consuming and labour-intensive, and manual annotation by a qualified pathologist is needed. With the ever-increasing number of tissue blocks and the complexity of molecular diagnostics, new approaches have been developed as complimentary or alternative solutions for the current workflow, such as digital pathology and mass spectrometry imaging (MSI). This study compares the performance of a digital pathology workflow using deep learning for tissue recognition and an MSI approach utilising shallow learning to annotate formalin-fixed and paraffin-embedded (FFPE) breast cancer tissue microarrays (TMAs). Results show that both deep learning algorithms based on conventional optical images and MSI-based shallow learning can provide automated diagnostics with F1-scores higher than 90%, with the latter intrinsically built on biochemical information that can be used for further analysis.

Associations of NAFLD with circulating ceramides and impaired glycemia

AIM

Determine the association of circulating ceramides with NAFLD and glycemic impairment.

METHODS

Sample: 669 participants in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) cohort aged 40-84 years without cardiovascular disease, cirrhosis, or significant alcohol intake.

CLINICAL MEASURES

Computed tomography scans at baseline for hepatic attenuation. Fasting serum specimens at baseline and after 5 years. Lipidomics: LC-MS-based analysis of 19 known ceramide signals.

STATISTICAL ANALYSIS

Linear and logistic regression models of log-transformed ceramides, hepatic attenuation and glucose adjusted for age, sex, calories, study site, BMI, exercise, diet quality, alcohol, saturated fat, lipid-lowering medications and fasting glucose.

RESULTS

Average age was 55 years, 44% were women, mean BMI was 25.9 kg/m2, and 8% had NAFLD. In adjusted models, Cer(d16:1/20:0) and Cer(d18:1/18:0) were associated with lower mean hepatic attenuation (increased liver fat) (β -4.29; 95% CI [-5.98, -2.59]) and (β -3.40; 95% CI [-5.11, -1.70]), and LacCer(d18:1/16:0) with higher attenuation (β 4.44; 95% CI [2.15, 6.73]). All three ceramides partially mediated the relationship between hepatic attenuation and fasting glucose by 16%, 11% and 5%, respectively, after 5-years.

CONCLUSIONS

Three circulating ceramides were strongly associated with NAFLD and fasting glucose after 5 years, and partially mediated this association.

Antiviral metabolite 3'-deoxy-3',4'-didehydro-cytidine is detectable in serum and identifies acute viral infections including COVID-19

BACKGROUND

There is a critical need for rapid viral infection diagnostics to enable prompt case identification in pandemic settings and support targeted antimicrobial prescribing.

METHODS

Using untargeted high-resolution liquid chromatography coupled with mass spectrometry, we compared the admission serum metabolome of emergency department patients with viral infections (including COVID-19), bacterial infections, inflammatory conditions, and healthy controls. Sera from an independent cohort of emergency department patients admitted with viral or bacterial infections underwent profiling to validate findings. Associations between whole-blood gene expression and the identified metabolite of interest were examined.

FINDINGS

3'-Deoxy-3',4'-didehydro-cytidine (ddhC), a free base of the only known human antiviral small molecule ddhC-triphosphate (ddhCTP), was detected for the first time in serum. When comparing 60 viral with 101 non-viral cases in the discovery cohort, ddhC was the most significantly differentially abundant metabolite, generating an area under the receiver operating characteristic curve (AUC) of 0.954 (95% CI: 0.923-0.986). In the validation cohort, ddhC was again the most significantly differentially abundant metabolite when comparing 40 viral with 40 bacterial cases, generating an AUC of 0.81 (95% CI 0.708-0.915). Transcripts of viperin and CMPK2, enzymes responsible for ddhCTP synthesis, were among the five genes most highly correlated with ddhC abundance.

CONCLUSIONS

The antiviral precursor molecule ddhC is detectable in serum and an accurate marker for acute viral infection. Interferon-inducible genes viperin and CMPK2 are implicated in ddhC production in vivo. These findings highlight a future diagnostic role for ddhC in viral diagnosis, pandemic preparedness, and acute infection management.

FUNDING

NIHR Imperial BRC; UKRI.

JPA: Joint Metabolic Feature Extraction Increases the Depth of Chemical Coverage for LC-MS-Based Metabolomics and Exposomics

Extracting metabolic features from liquid chromatography-mass spectrometry (LC-MS) data has been a long-standing bioinformatic challenge in untargeted metabolomics. Conventional feature extraction algorithms fail to recognize features with low signal intensities, poor chromatographic peak shapes, or those that do not fit the parameter settings. This problem also poses a challenge for MS-based exposome studies, as low-abundant metabolic or exposomic features cannot be automatically recognized from raw data. To address this data processing challenge, we developed an R package, JPA (short for Joint Metabolomic Data Processing and Annotation), to comprehensively extract metabolic features from raw LC-MS data. JPA performs feature extraction by combining a conventional peak picking algorithm and strategies for (1) recognizing features with bad peak shapes but that have tandem mass spectra (MS²) and (2) picking up features from a user-defined targeted list. The performance of JPA in global metabolomics was demonstrated using serial diluted urine samples, in which JPA was able to rescue an average of 25% of metabolic features that were missed by the conventional peak picking algorithm due to dilution. More importantly, the chromatographic peak shapes, analytical accuracy, and precision of the rescued metabolic features were all evaluated. Furthermore, owing to its sensitive feature extraction, JPA was able to achieve a limit of detection (LOD) that was up to thousands of folds lower when automatically processing metabolomics data of a serial diluted metabolite standard mixture analyzed in HILIC(−) and RP(+) modes. Finally, the performance of JPA in exposome research was validated using a mixture of 250 drugs and 255 pesticides at environmentally relevant levels. JPA detected an average of 2.3-fold more exposure compounds than conventional peak picking only.