1
|
Dlugas H, Kim S. A Comparative Study of Network-Based Machine Learning Approaches for Binary Classification in Metabolomics. Metabolites 2025; 15:174. [PMID: 40137139 PMCID: PMC11944042 DOI: 10.3390/metabo15030174] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/29/2025] [Revised: 02/21/2025] [Accepted: 02/27/2025] [Indexed: 03/27/2025] Open
Abstract
Background/Objectives: Metabolomics has recently emerged as a key tool in the biological sciences, offering insights into metabolic pathways and processes. Over the last decade, network-based machine learning approaches have gained significant popularity and application across various fields. While several studies have utilized metabolomics profiles for sample classification, many network-based machine learning approaches remain unexplored for metabolomic-based classification tasks. This study aims to compare the performance of various network-based machine learning approaches, including recently developed methods, in metabolomics-based classification. Methods: A standard data preprocessing procedure was applied to 17 metabolomic datasets, and Bayesian neural network (BNN), convolutional neural network (CNN), feedforward neural network (FNN), Kolmogorov-Arnold network (KAN), and spiking neural network (SNN) were evaluated on each dataset. The datasets varied widely in size, mass spectrometry method, and response variable. Results: With respect to AUC on test data, BNN, CNN, FNN, KAN, and SNN were the top-performing models in 4, 1, 5, 3, and 4 of the 17 datasets, respectively. Regarding F1-score, the top-performing models were BNN (3 datasets), CNN (3 datasets), FNN (4 datasets), KAN (4 datasets), and SNN (3 datasets). For accuracy, BNN, CNN, FNN, KAN, and SNN performed best in 4, 1, 4, 4, and 4 datasets, respectively. Conclusions: No network-based modeling approach consistently outperformed others across the metrics of AUC, F1-score, or accuracy. Our results indicate that while no single network-based modeling approach is superior for metabolomics-based classification tasks, BNN, KAN, and SNN may be underappreciated and underutilized relative to the more commonly used CNN and FNN.
Collapse
Affiliation(s)
- Hunter Dlugas
- Biostatistics and Bioinformatics Core, Karmanos Cancer Institute, Detroit, MI 48201, USA
- Department of Oncology, Wayne State University School of Medicine, Detroit, MI 48201, USA
| | - Seongho Kim
- Biostatistics and Bioinformatics Core, Karmanos Cancer Institute, Detroit, MI 48201, USA
- Department of Oncology, Wayne State University School of Medicine, Detroit, MI 48201, USA
| |
Collapse
|
2
|
Yurekten O, Payne T, Tejera N, Amaladoss FX, Martin C, Williams M, O’Donovan C. MetaboLights: open data repository for metabolomics. Nucleic Acids Res 2024; 52:D640-D646. [PMID: 37971328 PMCID: PMC10767962 DOI: 10.1093/nar/gkad1045] [Citation(s) in RCA: 95] [Impact Index Per Article: 95.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/15/2023] [Revised: 10/16/2023] [Accepted: 10/26/2023] [Indexed: 11/19/2023] Open
Abstract
MetaboLights is a global database for metabolomics studies including the raw experimental data and the associated metadata. The database is cross-species and cross-technique and covers metabolite structures and their reference spectra as well as their biological roles and locations where available. MetaboLights is the recommended metabolomics repository for a number of leading journals and ELIXIR, the European infrastructure for life science information. In this article, we describe the continued growth and diversity of submissions and the significant developments in recent years. In particular, we highlight MetaboLights Labs, our new Galaxy Project instance with repository-scale standardized workflows, and how data public on MetaboLights are being reused by the community. Metabolomics resources and data are available under the EMBL-EBI's Terms of Use at https://www.ebi.ac.uk/metabolights and under Apache 2.0 at https://github.com/EBI-Metabolights.
Collapse
Affiliation(s)
- Ozgur Yurekten
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Thomas Payne
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Noemi Tejera
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Felix Xavier Amaladoss
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Callum Martin
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Mark Williams
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| | - Claire O’Donovan
- European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK
| |
Collapse
|
3
|
Mrowiec K, Kurczyk A, Jelonek K, Debik J, Giskeødegård GF, Bathen TF, Widłak P. Association of serum metabolome profile with the risk of breast cancer in participants of the HUNT2 study. Front Oncol 2023; 13:1116806. [PMID: 37007110 PMCID: PMC10061137 DOI: 10.3389/fonc.2023.1116806] [Citation(s) in RCA: 12] [Impact Index Per Article: 6.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/05/2022] [Accepted: 03/06/2023] [Indexed: 03/18/2023] Open
Abstract
Background The serum metabolome is a potential source of molecular biomarkers associated with the risk of breast cancer. Here we aimed to analyze metabolites present in pre-diagnostic serum samples collected from healthy women participating in the Norwegian Trøndelag Health Study (HUNT2 study) for whom long-term information about developing breast cancer was available. Methods Women participating in the HUNT2 study who developed breast cancer within a 15-year follow-up period (BC cases) and age-matched women who stayed breast cancer-free were selected (n=453 case-control pairs). Using a high-resolution mass spectrometry approach 284 compounds were quantitatively analyzed, including 30 amino acids and biogenic amines, hexoses, and 253 lipids (acylcarnitines, glycerides, phosphatidylcholines, sphingolipids, and cholesteryl esters). Results Age was a major confounding factor responsible for a large heterogeneity in the dataset, hence age-defined subgroups were analyzed separately. The largest number of metabolites whose serum levels differentiated BC cases and controls (82 compounds) were observed in the subgroup of younger women (<45 years old). Noteworthy, increased levels of glycerides, phosphatidylcholines, and sphingolipids were associated with reduced risk of cancer in younger and middle-aged women (≤64 years old). On the other hand, increased levels of serum lipids were associated with an enhanced risk of breast cancer in older women (>64 years old). Moreover, several metabolites could be detected whose serum levels were different between BC cases diagnosed earlier (<5 years) and later (>10 years) after sample collecting, yet these compounds were also correlated with the age of participants. Current results were coherent with the results of the NMR-based metabolomics study performed in the cohort of HUNT2 participants, where increased serum levels of VLDL subfractions were associated with reduced risk of breast cancer in premenopausal women. Conclusions Changes in metabolite levels detected in pre-diagnostic serum samples, which reflected an impaired lipid and amino acid metabolism, were associated with long-term risk of breast cancer in an age-dependent manner.
Collapse
Affiliation(s)
- Katarzyna Mrowiec
- Center for Translational Research and Molecular Biology of Cancer, Maria Skłodowska-Curie National Research Institute of Oncology, Gliwice Branch, Gliwice, Poland
| | - Agata Kurczyk
- Center for Translational Research and Molecular Biology of Cancer, Maria Skłodowska-Curie National Research Institute of Oncology, Gliwice Branch, Gliwice, Poland
| | - Karol Jelonek
- Center for Translational Research and Molecular Biology of Cancer, Maria Skłodowska-Curie National Research Institute of Oncology, Gliwice Branch, Gliwice, Poland
| | - Julia Debik
- Department of Circulation and Medical Imaging, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
- K.G. Jebsen Center for Genetic Epidemiology, Department of Public Health and Nursing, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
| | - Guro F. Giskeødegård
- K.G. Jebsen Center for Genetic Epidemiology, Department of Public Health and Nursing, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
- Clinic of Surgery, St. Olavs University Hospital, Trondheim, Norway
| | - Tone F. Bathen
- Department of Circulation and Medical Imaging, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
- Department of Medical Imaging and Nuclear Medicine, St. Olavs University Hospital, Trondheim, Norway
| | - Piotr Widłak
- Clinical Research Support Centre, Medical University of Gdańsk, Gdańsk, Poland
| |
Collapse
|
4
|
Zhu J, Zhang L, Ji M, Jin B, Shu J. Elevated adipose differentiation-related protein level in ovariectomized mice correlates with tissue-specific regulation of estrogen. J Obstet Gynaecol Res 2023; 49:1173-1179. [PMID: 36772863 DOI: 10.1111/jog.15565] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/16/2022] [Accepted: 01/12/2023] [Indexed: 02/12/2023]
Abstract
OBJECTIVE Redistribution of adipose tissue in the abdomen during the menopausal transition is attributable mostly to estrogen drop with aging. Adipose differentiation-related protein (ADRP), a major component of lipid droplets, is closely related to the onset of lipid accumulation. We hypothesized that estrogen exerted its tissue-specific effect in reducing abdominal fat accumulation by regulation of ADRP. METHODS Twenty-four female C57/BL6 mice aged 8 weeks were randomly divided into 3 groups: sham operation (Sham), bilateral ovariectomy (OVX), and OVX plus 17β-estradiol (OVX + E2). After being fed 8 weeks of a high-fat diet, plasma lipid profiles including total cholesterol (TC), total triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) levels, body weight gain, visceral, and subcutaneous adipose tissue, adipocyte size, and ADRP expression were measured. RESULTS In comparison to sham-operated mice, OVX mice presented a weight gain with higher plasma TC, TG, LDL-C levels, and lower HDL-C levels. E2 supplement ameliorated the increase in weight and lipid profiles. Elevated ADRP expression was observed in visceral adipose tissue of OVX mice, whereas treatment of estrogen suppressed the ADPR expression and reversed the fat accumulation in the abdomen. However, no significant difference of ADRP expression in subcutaneous adipose tissue was detected between sham, OVX, and OVX + E2 mice. CONCLUSIONS Our findings suggested that enhanced ADRP expression in ovariectomized mice correlates with the tissue-specific regulation of estrogen, which may provide useful clues for further exploring the regulatory mechanism and corresponding anti-abdominal obesity treatment in postmenopausal women.
Collapse
Affiliation(s)
- Jing Zhu
- Department of Reproductive Endocrinology, Reproductive Medicine Center, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, China
| | - Ling Zhang
- Department of Reproductive Endocrinology, Reproductive Medicine Center, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, China
| | - Mengxia Ji
- Department of Reproductive Endocrinology, Reproductive Medicine Center, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, China
| | - Bihui Jin
- Department of Reproductive Endocrinology, Reproductive Medicine Center, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, China
| | - Jing Shu
- Department of Reproductive Endocrinology, Reproductive Medicine Center, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, China
| |
Collapse
|
5
|
Barupal DK, Mahajan P, Fakouri-Baygi S, Wright RO, Arora M, Teitelbaum SL. CCDB: A database for exploring inter-chemical correlations in metabolomics and exposomics datasets. ENVIRONMENT INTERNATIONAL 2022; 164:107240. [PMID: 35461097 PMCID: PMC9195052 DOI: 10.1016/j.envint.2022.107240] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 02/01/2022] [Revised: 04/01/2022] [Accepted: 04/08/2022] [Indexed: 05/18/2023]
Abstract
Inter-chemical correlations in metabolomics and exposomics datasets provide valuable information for studying relationships among chemicals reported for human specimens. With an increase in the number of compounds for these datasets, a network graph analysis and visualization of the correlation structure is difficult to interpret. We have developed the Chemical Correlation Database (CCDB), as a systematic catalogue of inter-chemical correlation in publicly available metabolomics and exposomics studies. The database has been provided via an online interface to create single compound-centric views. We have demonstrated various applications of the database to explore: 1) the chemicals from a chemical class such as Per- and Polyfluoroalkyl Substances (PFAS), polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), phthalates and tobacco smoke related metabolites; 2) xenobiotic metabolites such as caffeine and acetaminophen; 3) endogenous metabolites (acyl-carnitines); and 4) unannotated peaks for PFAS. The database has a rich collection of 35 human studies, including the National Health and Nutrition Examination Survey (NHANES) and high-quality untargeted metabolomics datasets. CCDB is supported by a simple, interactive and user-friendly web-interface to retrieve and visualize the inter-chemical correlation data. The CCDB has the potential to be a key computational resource in metabolomics and exposomics facilitating the expansion of our understanding about biological and chemical relationships among metabolites and chemical exposures in the human body. The database is available at www.ccdb.idsl.me site.
Collapse
Affiliation(s)
- Dinesh Kumar Barupal
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA.
| | - Priyanka Mahajan
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA
| | - Sadjad Fakouri-Baygi
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA
| | - Robert O Wright
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA
| | - Manish Arora
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA
| | - Susan L Teitelbaum
- Department of Environmental Medicine and Public Health, Institute for Exposomic Research, Icahn School of Medicine at Mount Sinai, 17 E 102nd St, CAM Building, New York 10029, USA
| |
Collapse
|
6
|
Fakouri Baygi S, Kumar Y, Barupal DK. IDSL.IPA Characterizes the Organic Chemical Space in Untargeted LC/HRMS Data Sets. J Proteome Res 2022; 21:1485-1494. [PMID: 35579321 DOI: 10.1021/acs.jproteome.2c00120] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/26/2022]
Abstract
Generating comprehensive and high-fidelity metabolomics data matrices from LC/HRMS data remains to be extremely challenging for population-scale large studies (n > 200). Here, we present a new data processing pipeline, the Intrinsic Peak Analysis (IDSL.IPA) R package (https://ipa.idsl.me), to generate such data matrices specifically for organic compounds. The IDSL.IPA pipeline incorporates (1) identifying potential 12C and 13C ion pairs in individual mass spectra; (2) detecting and characterizing chromatographic peaks using a new sensitive and versatile approach to perform mass correction, peak smoothing, baseline development for local noise measurement, and peak quality determination; (3) correcting retention time and cross-referencing peaks from multiple samples by a dynamic retention index marker approach; (4) annotating peaks using a reference database of m/z and retention time; and (5) accelerating data processing using a parallel computation of the peak detection and alignment steps for larger studies. This pipeline has been successfully evaluated for studies ranging from 200 to 1600 samples. By specifically isolating high quality and reliable signals pertaining to carbon-containing compounds in untargeted LC/HRMS data sets from larger studies, IDSL.IPA opens new opportunities for discovering new biological insights in the population-scale metabolomics and exposomics projects. The package is available in the R CRAN repository at https://cran.r-project.org/package=IDSL.IPA.
Collapse
Affiliation(s)
- Sadjad Fakouri Baygi
- Department of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai, New York, New York 10029, United States
| | - Yashwant Kumar
- Non-communicable Diseases Division, Translational Health Science and Technology Institute, Faridabad, Haryana 121001, India
| | - Dinesh Kumar Barupal
- Department of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai, New York, New York 10029, United States
| |
Collapse
|
7
|
Integrative metabolomic characterization identifies plasma metabolomic signature in the diagnosis of papillary thyroid cancer. Oncogene 2022; 41:2422-2430. [PMID: 35279704 DOI: 10.1038/s41388-022-02254-5] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/11/2021] [Revised: 02/05/2022] [Accepted: 02/16/2022] [Indexed: 01/29/2023]
Abstract
Discrimination of malignancy from thyroid nodules poses challenges in clinical practice. We aimed to identify the plasma metabolomic biomarkers in discriminating papillary thyroid cancer (PTC) from benign thyroid nodule (BTN). Metabolomics profiling of plasma was performed in two independent cohorts of 651 subjects of PTC (n = 215), BTN (n = 230), and healthy controls (n = 206). In addition, 132 patients with thyroid micronodules (<1 cm) and 44 patients with BTN suspected malignancy by ultrasound were used for biomarker validation. Recursive feature elimination algorithm was used for metabolic biomarkers selecting. Significant differential metabolites were demonstrated in patients with thyroid nodules (PTC and BTN) from healthy controls (P = 0.0001). A metabolic biomarker panel (17 differential metabolites) was identified to discriminate PTC from BTN with an AUC of 97.03% (95% CI: 95.28-98.79%), 91.89% sensitivity, and 92.63% specificity in discovery cohort. The panel had an AUC of 92.72% (95% CI: 87.46-97.99%), 86.57% sensitivity, and 92.50% specificity in validation cohort. The metabolic biomarker signature could correctly identify 84.09% patients whose nodules were suspected malignant by ultrasonography but finally histological benign. Moreover, high accuracy of 87.88% for diagnosis of papillary thyroid microcarcinoma was displayed by this panel and showed significant improvement in accuracy, AUC and specificity when compared with ultrasound. We identified a novel metabolic biomarker signature to discriminate PTC from BTN. The clinical use of this biomarker panel would have improved diagnosis stratification of thyroid microcarcinoma in comparison to ultrasound.
Collapse
|
8
|
Reinke SN, Naz S, Chaleckis R, Gallart-Ayala H, Kolmert J, Kermani NZ, Tiotiu A, Broadhurst DI, Lundqvist A, Olsson H, Ström M, Wheelock ÅM, Gómez C, Ericsson M, Sousa AR, Riley JH, Bates S, Scholfield J, Loza M, Baribaud F, Bakke PS, Caruso M, Chanez P, Fowler SJ, Geiser T, Howarth P, Horváth I, Krug N, Montuschi P, Behndig A, Singer F, Musial J, Shaw DE, Dahlén B, Hu S, Lasky-Su J, Sterk PJ, Chung KF, Djukanovic R, Dahlén SE, Adcock IM, Wheelock CE. Urinary metabotype of severe asthma evidences decreased carnitine metabolism independent of oral corticosteroid treatment in the U-BIOPRED study. Eur Respir J 2021; 59:13993003.01733-2021. [PMID: 34824054 PMCID: PMC9245194 DOI: 10.1183/13993003.01733-2021] [Citation(s) in RCA: 16] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/18/2021] [Accepted: 10/28/2021] [Indexed: 11/05/2022]
Abstract
INTRODUCTION Asthma is a heterogeneous disease with poorly defined phenotypes. Severe asthmatics often receive multiple treatments including oral corticosteroids (OCS). Treatment may modify the observed metabotype, rendering it challenging to investigate underlying disease mechanisms. Here, we aimed to identify dysregulated metabolic processes in relation to asthma severity and medication. METHODS Baseline urine was collected prospectively from healthy participants (n=100), mild-to-moderate asthmatics (n=87) and severe asthmatics (n=418) in the cross-sectional U-BIOPRED cohort; 12-18-month longitudinal samples were collected from severe asthmatics (n=305). Metabolomics data were acquired using high-resolution mass spectrometry and analysed using univariate and multivariate methods. RESULTS Ninety metabolites were identified, with 40 significantly altered (p<0.05, FDR<0.05) in severe asthma and 23 by OCS use. Multivariate modelling showed that observed metabotypes in healthy participants and mild-to-moderate asthmatics differed significantly from severe asthmatics (p=2.6×10-20), OCS-treated asthmatics differed significantly from non-treated (p=9.5×10-4), and longitudinal metabotypes demonstrated temporal stability. Carnitine levels evidenced the strongest OCS-independent decrease in severe asthma. Reduced carnitine levels were associated with mitochondrial dysfunction via decreases in pathway enrichment scores of fatty acid metabolism and reduced expression of the carnitine transporter SLC22A5 in sputum and bronchial brushings. CONCLUSIONS This is the first large-scale study to delineate disease- and OCS-associated metabolic differences in asthma. The widespread associations with different therapies upon the observed metabotypes demonstrate the necessity to evaluate potential modulating effects on a treatment- and metabolite-specific basis. Altered carnitine metabolism is a potentially actionable therapeutic target that is independent of OCS treatment, highlighting the role of mitochondrial dysfunction in severe asthma.
Collapse
Affiliation(s)
- Stacey N Reinke
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden.,Centre for Integrative Metabolomics & Computational Biology, School of Science, Edith Cowan University, Perth, Australia.,equal contribution
| | - Shama Naz
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden.,equal contribution
| | - Romanas Chaleckis
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden.,Gunma Initiative for Advanced Research (GIAR), Gunma University, Maebashi, Japan
| | - Hector Gallart-Ayala
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden
| | - Johan Kolmert
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden.,The Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
| | | | - Angelica Tiotiu
- National Heart and Lung Institute, Imperial College, London, U.K.,Department of Pulmonology, University Hospital of Nancy, Nancy, France
| | - David I Broadhurst
- Centre for Integrative Metabolomics & Computational Biology, School of Science, Edith Cowan University, Perth, Australia
| | - Anders Lundqvist
- Respiratory & Immunology, BioPharmaceuticals R&D, DMPK, Research and Early Development, AstraZeneca, Gothenburg, Sweden
| | - Henric Olsson
- Translational Science and Experimental Medicine, Research and Early Development, AstraZeneca, Gothenburg, Sweden
| | - Marika Ström
- Respiratory Medicine Unit, K2 Department of Medicine Solna and Center for Molecular Medicine, Karolinska Institutet, Stockholm, Sweden.,Department of Respiratory Medicine and Allergy, Karolinska University Hospital, Stockholm, Sweden
| | - Åsa M Wheelock
- Respiratory Medicine Unit, K2 Department of Medicine Solna and Center for Molecular Medicine, Karolinska Institutet, Stockholm, Sweden.,Department of Respiratory Medicine and Allergy, Karolinska University Hospital, Stockholm, Sweden
| | - Cristina Gómez
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden.,The Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
| | - Magnus Ericsson
- Department of Clinical Pharmacology, Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden
| | | | | | | | - James Scholfield
- Faculty of Medicine, Southampton University and NIHR Southampton Respiratory Biomedical Research Center, University Hospital Southampton, Southampton, U.K
| | - Matthew Loza
- Janssen Research and Development, High Wycombe, U.K
| | | | - Per S Bakke
- Institute of Medicine, University of Bergen, Bergen, Norway
| | - Massimo Caruso
- Department of Biomedical and Biotechnological Sciences and Department of Clinical and Experimental Medicine, University of Catania, Catania, Italy
| | - Pascal Chanez
- Assistance Publique des Hôpitaux de Marseille, Clinique des Bronches, Allergies et Sommeil, Aix Marseille Université, Marseille, France
| | - Stephen J Fowler
- Division of Infection, Immunity and Respiratory Medicine, Faculty of Biology, School of Biological Sciences, Medicine and Health, University of Manchester, and Manchester Academic Health Science Centre and NIHR Biomedical Research Centre, Manchester University Hospitals NHS Foundation Trust, Manchester, U.K
| | - Thomas Geiser
- Department of Pulmonary Medicine, University Hospital, University of Bern, Switzerland
| | - Peter Howarth
- Faculty of Medicine, Southampton University and NIHR Southampton Respiratory Biomedical Research Center, University Hospital Southampton, Southampton, U.K
| | - Ildikó Horváth
- Department of Pulmonology, Semmelweis University, Budapest, Hungary
| | - Norbert Krug
- Fraunhofer Institute for Toxicology and Experimental Medicine, Hannover, Germany
| | - Paolo Montuschi
- Pharmacology, Catholic University of the Sacred Heart, Rome, Italy
| | - Annelie Behndig
- Department of Public Health and Clinical Medicine, Section of Medicine, Umeå University, Umeå, Sweden
| | - Florian Singer
- Division of Paediatric Respiratory Medicine and Allergology, Department of Paediatrics, Inselspital, Bern University Hospital, University of Bern, Switzerland
| | - Jacek Musial
- Dept of Medicine, Jagiellonian University Medical College, Krakow, Poland
| | - Dominick E Shaw
- Nottingham NIHR Biomedical Research Centre, University of Nottingham, U.K
| | - Barbro Dahlén
- Department of Respiratory Medicine and Allergy, Karolinska University Hospital, Stockholm, Sweden
| | - Sile Hu
- Data Science Institute, Imperial College, London, U.K
| | - Jessica Lasky-Su
- Channing Division of Network Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA
| | - Peter J Sterk
- Department of Respiratory Medicine, Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands
| | - Kian Fan Chung
- National Heart and Lung Institute, Imperial College, London, U.K
| | - Ratko Djukanovic
- Faculty of Medicine, Southampton University and NIHR Southampton Respiratory Biomedical Research Center, University Hospital Southampton, Southampton, U.K
| | - Sven-Erik Dahlén
- The Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden.,Department of Respiratory Medicine and Allergy, Karolinska University Hospital, Stockholm, Sweden
| | - Ian M Adcock
- National Heart and Lung Institute, Imperial College, London, U.K
| | - Craig E Wheelock
- Division of Physiological Chemistry 2, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden .,Gunma Initiative for Advanced Research (GIAR), Gunma University, Maebashi, Japan.,Department of Respiratory Medicine and Allergy, Karolinska University Hospital, Stockholm, Sweden
| | | |
Collapse
|
9
|
Wieder C, Frainay C, Poupin N, Rodríguez-Mier P, Vinson F, Cooke J, Lai RPJ, Bundy JG, Jourdan F, Ebbels T. Pathway analysis in metabolomics: Recommendations for the use of over-representation analysis. PLoS Comput Biol 2021; 17:e1009105. [PMID: 34492007 PMCID: PMC8448349 DOI: 10.1371/journal.pcbi.1009105] [Citation(s) in RCA: 71] [Impact Index Per Article: 17.8] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/14/2021] [Revised: 09/17/2021] [Accepted: 08/23/2021] [Indexed: 11/18/2022] Open
Abstract
Over-representation analysis (ORA) is one of the commonest pathway analysis approaches used for the functional interpretation of metabolomics datasets. Despite the widespread use of ORA in metabolomics, the community lacks guidelines detailing its best-practice use. Many factors have a pronounced impact on the results, but to date their effects have received little systematic attention. Using five publicly available datasets, we demonstrated that changes in parameters such as the background set, differential metabolite selection methods, and pathway database used can result in profoundly different ORA results. The use of a non-assay-specific background set, for example, resulted in large numbers of false-positive pathways. Pathway database choice, evaluated using three of the most popular metabolic pathway databases (KEGG, Reactome, and BioCyc), led to vastly different results in both the number and function of significantly enriched pathways. Factors that are specific to metabolomics data, such as the reliability of compound identification and the chemical bias of different analytical platforms also impacted ORA results. Simulated metabolite misidentification rates as low as 4% resulted in both gain of false-positive pathways and loss of truly significant pathways across all datasets. Our results have several practical implications for ORA users, as well as those using alternative pathway analysis methods. We offer a set of recommendations for the use of ORA in metabolomics, alongside a set of minimal reporting guidelines, as a first step towards the standardisation of pathway analysis in metabolomics.
Collapse
Affiliation(s)
- Cecilia Wieder
- Section of Bioinformatics, Division of Systems Medicine, Department of Metabolism, Digestion, and Reproduction, Faculty of Medicine, Imperial College London, London, United Kingdom
| | - Clément Frainay
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
| | - Nathalie Poupin
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
| | - Pablo Rodríguez-Mier
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
| | - Florence Vinson
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
| | - Juliette Cooke
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
| | - Rachel PJ Lai
- Department of Infectious Disease, Faculty of Medicine, Imperial College London, London, United Kingdom
| | - Jacob G. Bundy
- Section of Biomolecular Medicine, Division of Systems Medicine, Department of Metabolism, Digestion, and Reproduction, Faculty of Medicine, Imperial College London, London, United Kingdom
| | - Fabien Jourdan
- Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRAE, ENVT, INP-Purpan, UPS, Toulouse, France
- MetaToul-MetaboHUB, National Infrastructure of Metabolomics and Fluxomics, Toulouse, France
| | - Timothy Ebbels
- Section of Bioinformatics, Division of Systems Medicine, Department of Metabolism, Digestion, and Reproduction, Faculty of Medicine, Imperial College London, London, United Kingdom
| |
Collapse
|
10
|
Jahagirdar S, Saccenti E. Evaluation of Single Sample Network Inference Methods for Metabolomics-Based Systems Medicine. J Proteome Res 2020; 20:932-949. [PMID: 33267585 PMCID: PMC7786380 DOI: 10.1021/acs.jproteome.0c00696] [Citation(s) in RCA: 9] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/08/2023]
Abstract
![]()
Networks
and network analyses are fundamental tools of systems
biology. Networks are built by inferring pair-wise relationships among
biological entities from a large number of samples such that subject-specific
information is lost. The possibility of constructing these sample
(individual)-specific networks from single molecular profiles might
offer new insights in systems and personalized medicine and as a consequence
is attracting more and more research interest. In this study, we evaluated
and compared LIONESS (Linear Interpolation to Obtain Network Estimates
for Single Samples) and ssPCC (single sample network based on Pearson
correlation) in the metabolomics context of metabolite–metabolite
association networks. We illustrated and explored the characteristics
of these two methods on (i) simulated data, (ii) data generated from
a dynamic metabolic model to simulate real-life observed metabolite
concentration profiles, and (iii) 22 metabolomic data sets and (iv)
we applied single sample network inference to a study case pertaining
to the investigation of necrotizing soft tissue infections to show
how these methods can be applied in metabolomics. We also proposed
some adaptations of the methods that can be used for data exploration.
Overall, despite some limitations, we found single sample networks
to be a promising tool for the analysis of metabolomics data.
Collapse
Affiliation(s)
- Sanjeevan Jahagirdar
- Laboratory of Systems and Synthetic Biology, Wageningen University & Research, Stippeneng 4, 6708 WE Wageningen, The Netherlands
| | - Edoardo Saccenti
- Laboratory of Systems and Synthetic Biology, Wageningen University & Research, Stippeneng 4, 6708 WE Wageningen, The Netherlands
| |
Collapse
|
11
|
Zhao H, Shen J, Ye Y, Wu X, Esteva FJ, Tripathy D, Chow WH. Validation of plasma metabolites associated with breast cancer risk among Mexican Americans. Cancer Epidemiol 2020; 69:101826. [PMID: 33010726 PMCID: PMC7710579 DOI: 10.1016/j.canep.2020.101826] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/06/2020] [Revised: 08/26/2020] [Accepted: 09/14/2020] [Indexed: 10/23/2022]
Abstract
In our previous breast cancer case control study in Hispanics, we found 14 metabolites whose levels differed between cases and controls. To validate the results, we carried out a nested case control study of 100 incident breast cancer and 100 matched healthy women identified from the Mano-A-Mano Mexican American Cohort study. With the adjustment of parity, education, birth place, language acculturation, BMI category, smoking, drinking, physical activity, and sitting time, 4 metabolites were associated with breast cancer risk: 3-hydroxyoctanoate (Odds ratio (OR) = 1.51, 95% confidence interval (CI): 1.10, 3.47), 3-hydroxybutyrate (BHBA) (OR = 1.42, 95%CI: 1.01, 3.72), linoleate (18:2n6) (OR = 1.39, 95% CI: 1.07, 4.04), and bilirubin (OR = 0.54, 95%CI: 0.42, 0.95). Then, we used 3 non-redundant metabolites, namely 3-hydroxyoctanoate, linoleate (18:2n6), and bilirubin, to generate a metabolic risk score. Increased metabolites risk score was associated with a 1.67-fold increased risk of breast cancer (OR = 1.67, 95%CI: 1.32, 3.94). And the significant association was more evident among those who were diagnosed with cancer earlier during the follow-up (≤ 5 years) than their counterparts. In conclusion, we identified four significant metabolites which may help elucidate metabolic pathways that contribute to breast carcinogenesis. Our findings warrant further replication efforts.
Collapse
Affiliation(s)
- Hua Zhao
- Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States; Department of Family Medicine and Population Health, School of Medicine, Virginia Commonwealth University, Richmond, VA, 23284, United States.
| | - Jie Shen
- Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States; Department of Family Medicine and Population Health, School of Medicine, Virginia Commonwealth University, Richmond, VA, 23284, United States
| | - Yuanqing Ye
- Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States; Department of Precision Health and DataScience, School of Public Health, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, 310058, PR China
| | - Xifeng Wu
- Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States; Department of Precision Health and DataScience, School of Public Health, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, 310058, PR China
| | - Francisco J Esteva
- Perlmutter Cancer Center at New York University Langone Health, New York, NY, 10016, United States
| | - Debasish Tripathy
- Breast Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States
| | - Wong-Ho Chow
- Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX, 77030, United States
| |
Collapse
|
12
|
Balasubramanian R, Demler O, Guasch-Ferré M, Paynter NP, Sheehan R, Liu S, Manson JE, Salas-Salvadó J, Martínez-Gonzalez MÁ, Hu FB, Clish C, Rexrode KM. Metabolomic Effects of Hormone Therapy and Associations With Coronary Heart Disease Among Postmenopausal Women. CIRCULATION. GENOMIC AND PRECISION MEDICINE 2020; 13:e002977. [PMID: 33141616 PMCID: PMC8824616 DOI: 10.1161/circgen.119.002977] [Citation(s) in RCA: 8] [Impact Index Per Article: 1.6] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
BACKGROUND In the WHI-HT trials (Women's Health Initiative Hormone Therapy), treatment with oral conjugated equine estrogens and medroxyprogesterone acetate (CEE+MPA) resulted in increased risk of coronary heart disease (CHD), whereas oral conjugated equine estrogens (CEE) did not. METHODS Four hundred eighty-one metabolites were measured at baseline and at 1-year in 503 and 431 participants in the WHI CEE and CEE+MPA trials, respectively. The effects of randomized HT on the metabolite profiles at 1-year was evaluated in linear models adjusting for baseline metabolite levels, age, body mass index, race, incident CHD, prevalent hypertension, and diabetes. Metabolites with discordant effects by HT type were evaluated for association with incident CHD in 944 participants (472 CHD cases) in the WHI-OS (Women's Health Initiative Observational Study), with replication in an independent cohort of 980 men and women at high risk for cardiovascular disease. RESULTS HT effects on the metabolome were profound; 62% of metabolites significantly changed with randomized CEE and 52% with CEE+MPA (false discovery rate-adjusted P value<0.05) in multivariable models. Concerted increases in abundance were seen within various metabolite classes including triacylglycerols, phosphatidylethanolamines, and phosphatidylcholines; decreases in abundance was observed for acylcarnitines, lysophosphatidylcholines, quaternary amines, and cholesteryl/cholesteryl esters. Twelve metabolites had discordant effects by HT type and were associated with incident CHD in the WHI-OS; a metabolite score estimated in a Least Absolute Shrinkage and Selection Operator regression was associated with CHD risk with an odds ratio of 1.47 per SD increase (95% CI, 1.27-1.70, P<10-6). All twelve metabolites were altered in the CHD protective direction by CEE treatment. One metabolite (lysine) was significantly altered in the direction of increased CHD risk by CEE+MPA; the remaining 11 metabolites were not significantly changed by CEE+MPA. The CHD associations of a subset of 4 metabolites including C58:11 triacylglycerol, C54:9 triacylglycerol, C36:1 phosphatidylcholine and sucrose replicated in an independent dataset of 980 participants in the PREDIMED trial (Prevención con Dieta Mediterránea). CONCLUSIONS Randomized treatment with oral HT resulted in large metabolome shifts that generally favored CEE alone over CEE+MPA in term of CHD risk. Discordant metabolite effects between HT regimens may partially mediate the differences in CHD risk between the 2 WHI-HT trials.
Collapse
Affiliation(s)
| | - Olga Demler
- Division of Preventive Medicine, Brigham and Women's Hospital, Harvard Medical School
| | - Marta Guasch-Ferré
- Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
| | - Nina P Paynter
- Division of Preventive Medicine, Brigham and Women's Hospital, Harvard Medical School
| | - Ryan Sheehan
- Department of Biostatistics & Epidemiology, University of Massachusetts-Amherst
| | - Simin Liu
- Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA
- Departments of Epidemiology & Medicine, Brown University, Providence, RI
| | - JoAnn E Manson
- Division of Preventive Medicine, Brigham and Women's Hospital, Harvard Medical School
- Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA
| | - Jordi Salas-Salvadó
- Universitat Rovira i Virgili, Departament de Bioquímica i Biotecnologia, Unitat de Nutrició Humana, Hospital Universitari San Joan de Reus
- Institut d'Investigació Pere Virgili (IISPV), Reus
- Consorcio CIBER, M.P. Fisiopatología de la Obesidad y Nutrición (CIBERObn), Instituto de Salud Carlos III (ISCIII), Madrid
| | - Miguel Á Martínez-Gonzalez
- Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
- Consorcio CIBER, M.P. Fisiopatología de la Obesidad y Nutrición (CIBERObn), Instituto de Salud Carlos III (ISCIII), Madrid
- Department of Preventive Medicine & Public Health, University of Navarra
- IdiSNA (Instituto de Investigación Sanitaria de Navarra), Pamplona, Spain
| | - Frank B Hu
- Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, MA
- Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA
| | - Clary Clish
- Broad Institute of the Massachusetts Institute of Technology & Harvard University, Cambridge, MA
| | - Kathryn M Rexrode
- Division of Women's Health, Brigham and Women's Hospital, Harvard Medical School
| |
Collapse
|
13
|
Edison AS, Colonna M, Gouveia GJ, Holderman NR, Judge MT, Shen X, Zhang S. NMR: Unique Strengths That Enhance Modern Metabolomics Research. Anal Chem 2020; 93:478-499. [DOI: 10.1021/acs.analchem.0c04414] [Citation(s) in RCA: 31] [Impact Index Per Article: 6.2] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/16/2022]
|
14
|
On the Use of Correlation and MI as a Measure of Metabolite-Metabolite Association for Network Differential Connectivity Analysis. Metabolites 2020; 10:metabo10040171. [PMID: 32344593 PMCID: PMC7241243 DOI: 10.3390/metabo10040171] [Citation(s) in RCA: 15] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/25/2020] [Revised: 04/15/2020] [Accepted: 04/22/2020] [Indexed: 02/06/2023] Open
Abstract
Metabolite differential connectivity analysis has been successful in investigating potential molecular mechanisms underlying different conditions in biological systems. Correlation and Mutual Information (MI) are two of the most common measures to quantify association and for building metabolite-metabolite association networks and to calculate differential connectivity. In this study, we investigated the performance of correlation and MI to identify significantly differentially connected metabolites. These association measures were compared on (i) 23 publicly available metabolomic data sets and 7 data sets from other fields, (ii) simulated data with known correlation structures, and (iii) data generated using a dynamic metabolic model to simulate real-life observed metabolite concentration profiles. In all cases, we found more differentially connected metabolites when using correlation indices as a measure for association than MI. We also observed that different MI estimation algorithms resulted in difference in performance when applied to data generated using a dynamic model. We concluded that there is no significant benefit in using MI as a replacement for standard Pearson's or Spearman's correlation when the application is to quantify and detect differentially connected metabolites.
Collapse
|
15
|
Metabolic Fingerprint of Turner Syndrome. J Clin Med 2020; 9:jcm9030664. [PMID: 32131408 PMCID: PMC7141341 DOI: 10.3390/jcm9030664] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/23/2020] [Revised: 02/24/2020] [Accepted: 02/26/2020] [Indexed: 01/12/2023] Open
Abstract
Girls with Turner syndrome (TS) are at increased risk of developing insulin resistance and coronary artery disease as a result of hypertension and obesity frequently seen in these patients. On the other hand, it is known that obesity is associated with increased serum levels of branched-chain amino acids (BCAAs: valine; leucine and isoleucine) and aromatic amino acids. The aim of the study is to compare the metabolic fingerprint of girls with TS to the metabolic fingerprint of girls with obesity. Metabolic fingerprinting using an untargeted metabolomic approach was examined in plasma from 46 girls with TS (study group) and 22 age-matched girls with obesity (control group). The mean values of BCAAs, methionine, phenylalanine, lysine, tryptophan, histidine, tyrosine, alanine and ornithine were significantly lower in the study group than in the control (p from 0.0025 to <0.000001). Strong significant correlation between BCAAs, phenylalanine, arginine, tyrosine, glutamic acid, citrulline and alanine, and body mass index expressed as standard deviation score BMI-SDS in the patients with obesity (p from 0.049 to 0.0005) was found. In contrast; there was no correlation between these amino acids and BMI-SDS in the girls with TS. It is suggested that obesity in patients with TS is not associated with altered amino acids metabolism.
Collapse
|
16
|
Mendez KM, Reinke SN, Broadhurst DI. A comparative evaluation of the generalised predictive ability of eight machine learning algorithms across ten clinical metabolomics data sets for binary classification. Metabolomics 2019; 15:150. [PMID: 31728648 PMCID: PMC6856029 DOI: 10.1007/s11306-019-1612-4] [Citation(s) in RCA: 104] [Impact Index Per Article: 17.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 09/22/2019] [Accepted: 11/05/2019] [Indexed: 12/18/2022]
Abstract
INTRODUCTION Metabolomics is increasingly being used in the clinical setting for disease diagnosis, prognosis and risk prediction. Machine learning algorithms are particularly important in the construction of multivariate metabolite prediction. Historically, partial least squares (PLS) regression has been the gold standard for binary classification. Nonlinear machine learning methods such as random forests (RF), kernel support vector machines (SVM) and artificial neural networks (ANN) may be more suited to modelling possible nonlinear metabolite covariance, and thus provide better predictive models. OBJECTIVES We hypothesise that for binary classification using metabolomics data, non-linear machine learning methods will provide superior generalised predictive ability when compared to linear alternatives, in particular when compared with the current gold standard PLS discriminant analysis. METHODS We compared the general predictive performance of eight archetypal machine learning algorithms across ten publicly available clinical metabolomics data sets. The algorithms were implemented in the Python programming language. All code and results have been made publicly available as Jupyter notebooks. RESULTS There was only marginal improvement in predictive ability for SVM and ANN over PLS across all data sets. RF performance was comparatively poor. The use of out-of-bag bootstrap confidence intervals provided a measure of uncertainty of model prediction such that the quality of metabolomics data was observed to be a bigger influence on generalised performance than model choice. CONCLUSION The size of the data set, and choice of performance metric, had a greater influence on generalised predictive performance than the choice of machine learning algorithm.
Collapse
Affiliation(s)
- Kevin M Mendez
- Centre for Metabolomics & Computational Biology, School of Science, Edith Cowan University, Joondalup, 6027, Australia
| | - Stacey N Reinke
- Centre for Metabolomics & Computational Biology, School of Science, Edith Cowan University, Joondalup, 6027, Australia
| | - David I Broadhurst
- Centre for Metabolomics & Computational Biology, School of Science, Edith Cowan University, Joondalup, 6027, Australia.
| |
Collapse
|
17
|
His M, Viallon V, Dossus L, Gicquiau A, Achaintre D, Scalbert A, Ferrari P, Romieu I, Onland-Moret NC, Weiderpass E, Dahm CC, Overvad K, Olsen A, Tjønneland A, Fournier A, Rothwell JA, Severi G, Kühn T, Fortner RT, Boeing H, Trichopoulou A, Karakatsani A, Martimianaki G, Masala G, Sieri S, Tumino R, Vineis P, Panico S, van Gils CH, Nøst TH, Sandanger TM, Skeie G, Quirós JR, Agudo A, Sánchez MJ, Amiano P, Huerta JM, Ardanaz E, Schmidt JA, Travis RC, Riboli E, Tsilidis KK, Christakoudi S, Gunter MJ, Rinaldi S. Prospective analysis of circulating metabolites and breast cancer in EPIC. BMC Med 2019; 17:178. [PMID: 31547832 PMCID: PMC6757362 DOI: 10.1186/s12916-019-1408-4] [Citation(s) in RCA: 86] [Impact Index Per Article: 14.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/02/2019] [Accepted: 08/13/2019] [Indexed: 01/12/2023] Open
Abstract
BACKGROUND Metabolomics is a promising molecular tool to identify novel etiologic pathways leading to cancer. Using a targeted approach, we prospectively investigated the associations between metabolite concentrations in plasma and breast cancer risk. METHODS A nested case-control study was established within the European Prospective Investigation into Cancer cohort, which included 1624 first primary incident invasive breast cancer cases (with known estrogen and progesterone receptor and HER2 status) and 1624 matched controls. Metabolites (n = 127, acylcarnitines, amino acids, biogenic amines, glycerophospholipids, hexose, sphingolipids) were measured by mass spectrometry in pre-diagnostic plasma samples and tested for associations with breast cancer incidence using multivariable conditional logistic regression. RESULTS Among women not using hormones at baseline (n = 2248), and after control for multiple tests, concentrations of arginine (odds ratio [OR] per SD = 0.79, 95% confidence interval [CI] = 0.70-0.90), asparagine (OR = 0.83 (0.74-0.92)), and phosphatidylcholines (PCs) ae C36:3 (OR = 0.83 (0.76-0.90)), aa C36:3 (OR = 0.84 (0.77-0.93)), ae C34:2 (OR = 0.85 (0.78-0.94)), ae C36:2 (OR = 0.85 (0.78-0.88)), and ae C38:2 (OR = 0.84 (0.76-0.93)) were inversely associated with breast cancer risk, while the acylcarnitine C2 (OR = 1.23 (1.11-1.35)) was positively associated with disease risk. In the overall population, C2 (OR = 1.15 (1.06-1.24)) and PC ae C36:3 (OR = 0.88 (0.82-0.95)) were associated with risk of breast cancer, and these relationships did not differ by breast cancer subtype, age at diagnosis, fasting status, menopausal status, or adiposity. CONCLUSIONS These findings point to potentially novel pathways and biomarkers of breast cancer development. Results warrant replication in other epidemiological studies.
Collapse
Affiliation(s)
- Mathilde His
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Vivian Viallon
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Laure Dossus
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Audrey Gicquiau
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - David Achaintre
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Augustin Scalbert
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Pietro Ferrari
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Isabelle Romieu
- Centre for Research on Population Health, National Institute of Public Health, Cuernavaca, Mexico
| | - N Charlotte Onland-Moret
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands
| | - Elisabete Weiderpass
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | | | - Kim Overvad
- Department of Public Health, Aarhus University, Aarhus, Denmark
- Department of Cardiology, Aalborg University Hospital, Aalborg, Denmark
| | - Anja Olsen
- Danish Cancer Society Research Center, Copenhagen, Denmark
| | - Anne Tjønneland
- Danish Cancer Society Research Center, Copenhagen, Denmark
- University of Copenhagen, Copenhagen, Denmark
| | - Agnès Fournier
- CESP, Université Paris-Sud, UVSQ, INSERM, Université Paris-Saclay, Villejuif, France
- Gustave Roussy, Villejuif, France
| | - Joseph A Rothwell
- CESP, Université Paris-Sud, UVSQ, INSERM, Université Paris-Saclay, Villejuif, France
- Gustave Roussy, Villejuif, France
| | - Gianluca Severi
- CESP, Université Paris-Sud, UVSQ, INSERM, Université Paris-Saclay, Villejuif, France
- Gustave Roussy, Villejuif, France
| | - Tilman Kühn
- Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany
| | - Renée T Fortner
- Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany
| | - Heiner Boeing
- Department of Epidemiology, German Institute of Human Nutrition Potsdam-Rehbruecke (DIfE), Arthur-Scheunert-Allee 114-116, 14558, Nuthetal, Germany
| | | | - Anna Karakatsani
- Hellenic Health Foundation, Athens, Greece
- Pulmonary Medicine Department, School of Medicine, National and Kapodistrian University of Athens, "ATTIKON" University Hospital, Haidari, Greece
| | | | - Giovanna Masala
- Cancer Risk Factors and Life-Style Epidemiology Unit, Institute for Cancer Research, Prevention and Clinical Network - ISPRO, Florence, Italy
| | - Sabina Sieri
- Epidemiology and Prevention Unit, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy
| | - Rosario Tumino
- Cancer Registry and Histopathology Department, "M.P.Arezzo"Hospital, ASP Ragusa, Ragusa, Italy
| | - Paolo Vineis
- Italian Institute for Genomic Medicine (IIGM), 10126, Turin, Italy
- MRC-PHE Centre for Environment and Health, School of Public Health, Imperial College London, London, UK
| | - Salvatore Panico
- Dipartimento di medicina clinica e chirurgia, Federico II University, Naples, Italy
| | - Carla H van Gils
- Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands
| | - Therese H Nøst
- Department of Community Medicine, UiT the Arctic University of Norway, Tromso, Norway
| | - Torkjel M Sandanger
- Department of Community Medicine, UiT the Arctic University of Norway, Tromso, Norway
| | - Guri Skeie
- Department of Community Medicine, UiT the Arctic University of Norway, Tromso, Norway
- Nutritional Epidemiology Group, School of Food Science and Nutrition, University of Leeds, Leeds, UK
| | | | - Antonio Agudo
- Unit of Nutrition and Cancer, Cancer Epidemiology Research Program, Catalan Institute of Oncology-IDIBELL, L'Hospitalet de Llobregat, Barcelona, Spain
| | - Maria-Jose Sánchez
- Escuela Andaluza de Salud Pública, Instituto de Investigación Biosanitaria ibs.GRANADA, Universidad de Granada, Granada, Spain
- CIBER Epidemiology and Public Health CIBERESP, Madrid, Spain
| | - Pilar Amiano
- CIBER Epidemiology and Public Health CIBERESP, Madrid, Spain
- Public Health Division of Gipuzkoa, BioDonostia Research Institute, San Sebastian, Spain
| | - José María Huerta
- CIBER Epidemiology and Public Health CIBERESP, Madrid, Spain
- Department of Epidemiology, Murcia Regional Health Council, IMIB-Arrixaca, Murcia, Spain
| | - Eva Ardanaz
- CIBER Epidemiology and Public Health CIBERESP, Madrid, Spain
- Navarra Public Health Institute, Pamplona, Spain
- IdiSNA, Navarra Institute for Health Research, Pamplona, Spain
| | - Julie A Schmidt
- Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK
| | - Ruth C Travis
- Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK
| | - Elio Riboli
- Department of Epidemiology and Biostatistics, Imperial College London, St Mary's Campus, Norfolk Place, London, W2 1PG, UK
| | - Konstantinos K Tsilidis
- Department of Epidemiology and Biostatistics, Imperial College London, St Mary's Campus, Norfolk Place, London, W2 1PG, UK
- Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina, Greece
| | - Sofia Christakoudi
- Department of Epidemiology and Biostatistics, Imperial College London, St Mary's Campus, Norfolk Place, London, W2 1PG, UK
- MRC Centre for Transplantation, King's College London, Great Maze Pond, London, SE1 9RT, UK
| | - Marc J Gunter
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France
| | - Sabina Rinaldi
- International Agency for Research on Cancer, 150 cours Albert Thomas, 69372, Lyon CEDEX 08, France.
| |
Collapse
|
18
|
Turkson S, Kloster A, Hamilton PJ, Neigh GN. Neuroendocrine drivers of risk and resilience: The influence of metabolism & mitochondria. Front Neuroendocrinol 2019; 54:100770. [PMID: 31288042 PMCID: PMC6886586 DOI: 10.1016/j.yfrne.2019.100770] [Citation(s) in RCA: 10] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/02/2019] [Revised: 06/20/2019] [Accepted: 07/03/2019] [Indexed: 02/07/2023]
Abstract
The manifestation of risk versus resilience has been considered from varying perspectives including genetics, epigenetics, early life experiences, and type and intensity of the challenge with which the organism is faced. Although all of these factors are central to determining risk and resilience, the current review focuses on what may be a final common pathway: metabolism. When an organism is faced with a perturbation to the environment, whether internal or external, appropriate energy allocation is essential to resolving the divergence from equilibrium. This review examines the potential role of metabolism in the manifestation of stress-induced neural compromise. In addition, this review details the current state of knowledge on neuroendocrine factors which are poised to set the tone of the metabolic response to a systemic challenge. The goal is to provide an essential framework for understanding stress in a metabolic context and appreciation for key neuroendocrine signals.
Collapse
Affiliation(s)
- Susie Turkson
- Department of Anatomy & Neurobiology, Virginia Commonwealth University, Richmond, VA, United States
| | - Alix Kloster
- Department of Anatomy & Neurobiology, Virginia Commonwealth University, Richmond, VA, United States
| | - Peter J Hamilton
- Department of Anatomy & Neurobiology, Virginia Commonwealth University, Richmond, VA, United States
| | - Gretchen N Neigh
- Department of Anatomy & Neurobiology, Virginia Commonwealth University, Richmond, VA, United States.
| |
Collapse
|