Multinomial Machine Learning Identifies Independent Biomarkers by Integrated Metabolic Analysis of Acute Coronary Syndrome

Overview

Journal Sci Rep

Specialty Science

Date 2023 Nov 23

PMID 37996510

Authors

Meijiao Fu

Ruhua He

Zhihan Zhang

Fuqing Ma

Libo Shen

Yu Zhang

Mingyu Duan

Yameng Zhang

Yifan Wang

Li Zhu

Jun He

Affiliations

Soon will be listed here.

Abstract

A multi-class classification model for acute coronary syndrome (ACS) remains to be constructed based on multi-fluid metabolomics. Major confounders may exert spurious effects on the relationship between metabolism and ACS. The study aims to identify an independent biomarker panel for the multiclassification of HC, UA, and AMI by integrating serum and urinary metabolomics. We performed a liquid chromatography-tandem mass spectrometry (LC-MS/MS)-based metabolomics study on 300 serum and urine samples from 44 patients with unstable angina (UA), 77 with acute myocardial infarction (AMI), and 29 healthy controls (HC). Multinomial machine learning approaches, including multinomial adaptive least absolute shrinkage and selection operator (LASSO) regression and random forest (RF), and assessment of the confounders were applied to integrate a multi-class classification biomarker panel for HC, UA and AMI. Different metabolic landscapes were portrayed during the transition from HC to UA and then to AMI. Glycerophospholipid metabolism and arginine biosynthesis were predominant during the progression from HC to UA and then to AMI. The multiclass metabolic diagnostic model (MDM) dependent on ACS, including 2-ketobutyric acid, LysoPC(18:2(9Z,12Z)), argininosuccinic acid, and cyclic GMP, demarcated HC, UA, and AMI, providing a C-index of 0.84 (HC vs. UA), 0.98 (HC vs. AMI), and 0.89 (UA vs. AMI). The diagnostic value of MDM largely derives from the contribution of 2-ketobutyric acid, and LysoPC(18:2(9Z,12Z)) in serum. Higher 2-ketobutyric acid and cyclic GMP levels were positively correlated with ACS risk and atherosclerosis plaque burden, while LysoPC(18:2(9Z,12Z)) and argininosuccinic acid showed the reverse relationship. An independent multiclass biomarker panel for HC, UA, and AMI was constructed using the multinomial machine learning methods based on serum and urinary metabolite signatures.

Citing Articles

Machine Learning Applications in Acute Coronary Syndrome: Diagnosis, Outcomes and Management.

Nie S, Zhang S, Zhao Y, Li X, Xu H, Wang Y Adv Ther. 2024; 42(2):636-665.

PMID: 39641854 DOI: 10.1007/s12325-024-03060-z.

Identification of novel hypertension biomarkers using explainable AI and metabolomics.

Sekaran K, Zayed H Metabolomics. 2024; 20(6):124.

PMID: 39489869 PMC: 11532322. DOI: 10.1007/s11306-024-02182-3.

Exploring the role of CBLB in acute myocardial infarction: transcriptomic, microbiomic, and metabolomic analyses.

You H, Chang F, Chen H, Wang Y, Han W J Transl Med. 2024; 22(1):654.

PMID: 39004726 PMC: 11247792. DOI: 10.1186/s12967-024-05425-y.

Combining the Strengths of the Explainable Boosting Machine and Metabolomics Approaches for Biomarker Discovery in Acute Myocardial Infarction.

Arslan A, Yagin F, Algarni A, Al-Hashem F, Ardigo L Diagnostics (Basel). 2024; 14(13).

PMID: 39001243 PMC: 11240568. DOI: 10.3390/diagnostics14131353.

References

Bui D, Ravasz D, Chinopoulos C . The Effect of 2-Ketobutyrate on Mitochondrial Substrate-Level Phosphorylation. Neurochem Res. 2019; 44(10):2301-2306. PMC: 6776489. DOI: 10.1007/s11064-019-02759-8. View

Knuplez E, Marsche G . An Updated Review of Pro- and Anti-Inflammatory Properties of Plasma Lysophosphatidylcholines in the Vascular System. Int J Mol Sci. 2020; 21(12). PMC: 7350010. DOI: 10.3390/ijms21124501. View

Wen L, Feil S, Wolters M, Thunemann M, Regler F, Schmidt K . A shear-dependent NO-cGMP-cGKI cascade in platelets acts as an auto-regulatory brake of thrombosis. Nat Commun. 2018; 9(1):4301. PMC: 6191445. DOI: 10.1038/s41467-018-06638-8. View

Forstermann U, Sessa W . Nitric oxide synthases: regulation and function. Eur Heart J. 2011; 33(7):829-37, 837a-837d. PMC: 3345541. DOI: 10.1093/eurheartj/ehr304. View

Cui S, Li L, Zhang Y, Lu J, Wang X, Song X . Machine Learning Identifies Metabolic Signatures that Predict the Risk of Recurrent Angina in Remitted Patients after Percutaneous Coronary Intervention: A Multicenter Prospective Cohort Study. Adv Sci (Weinh). 2021; 8(10):2003893. PMC: 8132066. DOI: 10.1002/advs.202003893. View

Gornik H, Creager M . Arginine and endothelial and vascular health. J Nutr. 2004; 134(10 Suppl):2880S-2887S. DOI: 10.1093/jn/134.10.2880S. View

Dodd M, Atherton H, Carr C, Stuckey D, West J, Griffin J . Impaired in vivo mitochondrial Krebs cycle activity after myocardial infarction assessed using hyperpolarized magnetic resonance spectroscopy. Circ Cardiovasc Imaging. 2014; 7(6):895-904. PMC: 4450075. DOI: 10.1161/CIRCIMAGING.114.001857. View

Wang X, Wang D, Wu J, Yu X, Lv J, Kong J . Metabolic Characterization of Myocardial Infarction Using GC-MS-Based Tissue Metabolomics. Int Heart J. 2017; 58(3):441-446. DOI: 10.1536/ihj.16-432. View

Fromentin S, Forslund S, Chechi K, Aron-Wisnewsky J, Chakaroun R, Nielsen T . Microbiome and metabolome features of the cardiometabolic disease spectrum. Nat Med. 2022; 28(2):303-314. PMC: 8863577. DOI: 10.1038/s41591-022-01688-4. View

10.

Riederer M, Ojala P, Hrzenjak A, Graier W, Malli R, Tritscher M . Acyl chain-dependent effect of lysophosphatidylcholine on endothelial prostacyclin production. J Lipid Res. 2010; 51(10):2957-66. PMC: 2936763. DOI: 10.1194/jlr.M006536. View

11.

Brkic L, Riederer M, Graier W, Malli R, Frank S . Acyl chain-dependent effect of lysophosphatidylcholine on cyclooxygenase (COX)-2 expression in endothelial cells. Atherosclerosis. 2012; 224(2):348-54. PMC: 3465554. DOI: 10.1016/j.atherosclerosis.2012.07.038. View

12.

Talmor-Barkan Y, Bar N, Shaul A, Shahaf N, Godneva A, Bussi Y . Metabolomic and microbiome profiling reveals personalized risk factors for coronary artery disease. Nat Med. 2022; 28(2):295-302. DOI: 10.1038/s41591-022-01686-6. View

13.

Khot U, Khot M, Bajzer C, Sapp S, Ohman E, Brener S . Prevalence of conventional risk factors in patients with coronary heart disease. JAMA. 2003; 290(7):898-904. DOI: 10.1001/jama.290.7.898. View

14.

Rocha A, Goldenstein S . Multiclass from binary: expanding one-versus-all, one-versus-one and ECOC-based approaches. IEEE Trans Neural Netw Learn Syst. 2014; 25(2):289-302. DOI: 10.1109/TNNLS.2013.2274735. View

15.

Barnett M, Doroudgar S, Khosraviani V, Ip E . Multiple comparisons: To compare or not to compare, that is the question. Res Social Adm Pharm. 2021; 18(2):2331-2334. DOI: 10.1016/j.sapharm.2021.07.006. View

16.

Bahamyirou A, Schnitzer M, Kennedy E, Blais L, Yang Y . Doubly robust adaptive LASSO for effect modifier discovery. Int J Biostat. 2022; 18(2):307-327. DOI: 10.1515/ijb-2020-0073. View

17.

Schaeffer D, Riazy M, Parhar K, Chen J, Duronio V, Sawamura T . LOX-1 augments oxLDL uptake by lysoPC-stimulated murine macrophages but is not required for oxLDL clearance from plasma. J Lipid Res. 2009; 50(8):1676-84. PMC: 2724041. DOI: 10.1194/jlr.M900167-JLR200. View

18.

Chouchani E, Pell V, Gaude E, Aksentijevic D, Sundier S, Robb E . Ischaemic accumulation of succinate controls reperfusion injury through mitochondrial ROS. Nature. 2014; 515(7527):431-435. PMC: 4255242. DOI: 10.1038/nature13909. View

19.

Meikle P, Formosa M, Mellett N, Jayawardana K, Giles C, Bertovic D . HDL Phospholipids, but Not Cholesterol Distinguish Acute Coronary Syndrome From Stable Coronary Artery Disease. J Am Heart Assoc. 2019; 8(11):e011792. PMC: 6585356. DOI: 10.1161/JAHA.118.011792. View

20.

Sumner L, Amberg A, Barrett D, Beale M, Beger R, Daykin C . Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics. 2013; 3(3):211-221. PMC: 3772505. DOI: 10.1007/s11306-007-0082-2. View