Deep Learning-Based Molecular Fingerprint Prediction for Metabolite Annotation DOI Creative Commons

Hoi Yan Katharine Chau,

Xinran Zhang, Habtom W. Ressom

et al.

Metabolites, Journal Year: 2025, Volume and Issue: 15(2), P. 132 - 132

Published: Feb. 14, 2025

Background/Objectives: Liquid chromatography coupled with mass spectrometry (LC-MS) is a commonly used platform for many metabolomics studies. However, metabolite annotation has been major bottleneck in these studies part due to the limited publicly available spectral libraries, which consist of tandem (MS/MS) data acquired from just fraction known compounds. Application deep learning methods increasingly reported as an alternative matching their ability map complex relationships between molecular fingerprints and spectrometric measurements. The objectives this study are investigate fingerprint based on MS/MS spectra rank putative IDs according similarity predicted fingerprints. Methods: We trained three types model spectra. Prior training, various processing steps, including scaling, binning, filtering, were performed obtained National Institute Standards Technology (NIST), MassBank North America (MoNA), Human Metabolome Database (HMDB). Furthermore, selection most relevant m/z bins was conducted. models evaluated ranking compound database challenges Critical Assessment Small Molecule Identification (CASMI) 2016, CASMI 2017, 2022 benchmark datasets. Results: Feature effectively reduced redundant features prior training. Deep truncated have shown comparable performances against CSI:FingerID IDs. Conclusion: results demonstrate promising potential annotation.

Language: Английский

MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation DOI Creative Commons
Zhiqiang Pang, Yao Lü,

Guangyan Zhou

et al.

Nucleic Acids Research, Journal Year: 2024, Volume and Issue: 52(W1), P. W398 - W406

Published: April 8, 2024

Abstract We introduce MetaboAnalyst version 6.0 as a unified platform for processing, analyzing, and interpreting data from targeted well untargeted metabolomics studies using liquid chromatography - mass spectrometry (LC–MS). The two main objectives in developing are to support tandem MS (MS2) processing annotation, the analysis of exposomics related experiments. Key features include: (i) significantly enhanced Spectra Processing module with MS2 asari algorithm; (ii) Peak Annotation based on comprehensive reference databases fragment-level annotation; (iii) new Statistical Analysis dedicated handling complex study design multiple factors or phenotypic descriptors; (iv) Causal estimating metabolite phenotype causal relations two-sample Mendelian randomization, (v) Dose-Response benchmark dose calculations. In addition, we have also improved MetaboAnalyst's visualization functions, updated its compound database sets, expanded pathway around 130 species. is freely available at https://www.metaboanalyst.ca.

Language: Английский

Citations

486
The underappreciated diversity of bile acid modifications DOI Creative Commons
Ipsita Mohanty, Helena Mannochio-Russo,

Joshua V. Schweer

et al.

Cell, Journal Year: 2024, Volume and Issue: 187(7), P. 1801 - 1818.e20

Published: March 1, 2024

The repertoire of modifications to bile acids and related steroidal lipids by host microbial metabolism remains incompletely characterized. To address this knowledge gap, we created a reusable resource tandem mass spectrometry (MS/MS) spectra filtering 1.2 billion publicly available MS/MS for bile-acid-selective ion patterns. Thousands are distributed throughout animal human bodies as well cultures. We employed library identify polyamine amidates, prevalent in carnivores. They present humans, their levels alter with diet change from Mediterranean typical American diet. This work highlights the existence many more acid than previously recognized value leveraging public large-scale untargeted metabolomics data discover metabolites. availability modification-centric will inform future studies investigating roles health disease.

Language: Английский

Citations

73
The changing metabolic landscape of bile acids – keys to metabolism and immune regulation DOI
Ipsita Mohanty, Celeste Allaband, Helena Mannochio-Russo

et al.

Nature Reviews Gastroenterology & Hepatology, Journal Year: 2024, Volume and Issue: 21(7), P. 493 - 516

Published: April 4, 2024

Language: Английский

Citations

65
MetaboAnalystR 4.0: a unified LC-MS workflow for global metabolomics DOI Creative Commons
Zhiqiang Pang, Lei Xu,

Charles Viau

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: May 1, 2024

Abstract The wide applications of liquid chromatography - mass spectrometry (LC-MS) in untargeted metabolomics demand an easy-to-use, comprehensive computational workflow to support efficient and reproducible data analysis. However, current tools were primarily developed perform specific tasks LC-MS based Here we introduce MetaboAnalystR 4.0 as a streamlined pipeline covering raw spectra processing, compound identification, statistical analysis, functional interpretation. key features includes auto-optimized feature detection quantification algorithm for LC-MS1 MS2 deconvolution identification data-dependent or data-independent acquisition, more accurate interpretation through integrated spectral annotation. Comprehensive validation studies using obtained from standards mixtures, dilution series clinical samples have shown its excellent performance across range common such peak picking, deconvolution, with good computing efficiency. Together existing analysis utilities, represents significant step toward unified, end-to-end global the open-source R environment.

Language: Английский

Citations

56
High-Resolution Mass Spectrometry for Human Exposomics: Expanding Chemical Space Coverage DOI Creative Commons
Yunjia Lai, Jeremy P. Koelmel, Douglas G. Walker

et al.

Environmental Science & Technology, Journal Year: 2024, Volume and Issue: 58(29), P. 12784 - 12822

Published: July 10, 2024

In the modern "omics" era, measurement of human exposome is a critical missing link between genetic drivers and disease outcomes. High-resolution mass spectrometry (HRMS), routinely used in proteomics metabolomics, has emerged as leading technology to broadly profile chemical exposure agents related biomolecules for accurate measurement, high sensitivity, rapid data acquisition, increased resolution space. Non-targeted approaches are increasingly accessible, supporting shift from conventional hypothesis-driven, quantitation-centric targeted analyses toward data-driven, hypothesis-generating exposome-wide profiling. However, HRMS-based exposomics encounters unique challenges. New analytical computational infrastructures needed expand analysis coverage through streamlined, scalable, harmonized workflows pipelines that permit longitudinal tracking, retrospective validation, multi-omics integration meaningful health-oriented inferences. this article, we survey literature on state-of-the-art technologies, review current informatic pipelines, provide an up-to-date reference exposomic chemists, toxicologists, epidemiologists, care providers, stakeholders health sciences medicine. We propose efforts benchmark fit-for-purpose platforms expanding space, including gas/liquid chromatography-HRMS (GC-HRMS LC-HRMS), discuss opportunities, challenges, strategies advance burgeoning field exposome.

Language: Английский

Citations

21
Energy-Confinement 3D Flower-Shaped Cages for AI-Driven Decoding of Metabolic Fingerprints in Cardiovascular Disease Diagnosis DOI
Zhiyu Li, Shuyu Zhang, Qianfeng Xiao

et al.

ACS Nano, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 7, 2025

Rapid and accurate detection plays a critical role in improving the survival prognosis of patients with cardiovascular disease, but traditional methods are far from ideal for those suspected conditions. Metabolite analysis based on nanomatrix-assisted laser desorption/ionization mass spectrometry (NMALDI-MS) is considered to be promising technique disease diagnosis. However, performance core nanomatrixes has limited its clinical application. In this study, we constructed 3D flower-shaped cages controllable structured metal-organic frameworks iron oxide nanoparticles low thermal conductivity significant photothermal effects. The elongation incident light path through multilayer reflection significantly enhances effective absorption area nanomatrixes. Concurrently, alternating layered structure confines energy, reducing losses. Moreover, increases affinity sites, expanding coverage. This approach effectively ionization desorption efficiency during LDI process. We applied technology analyze serum metabolomes myocardial infarction, heart failure, failure combined achieving cost-effective, high-throughput, highly accurate, user-friendly diseases. Subsequently, deep detected fingerprints via artificial intelligence models screens potential metabolic biomarkers, providing new paradigm diagnosis

Language: Английский

Citations

2
Open access repository-scale propagated nearest neighbor suspect spectral library for untargeted metabolomics DOI Creative Commons
Wout Bittremieux, Nicole E. Avalon, Sydney P. Thomas

et al.

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: Dec. 20, 2023

Abstract Despite the increasing availability of tandem mass spectrometry (MS/MS) community spectral libraries for untargeted metabolomics over past decade, majority acquired MS/MS spectra remain uninterpreted. To further aid in interpreting unannotated spectra, we created a nearest neighbor suspect library, consisting 87,916 annotated derived from hundreds millions originating published experiments. Entries this or “suspects,” were that could be linked molecular network to an spectrum. Annotations propagated unknowns based on structural relationships reference molecules using MS/MS-based spectrum alignment. We demonstrate broad relevance library through representative examples propagation-based annotation acylcarnitines, bacterial and plant natural products, drug metabolism. Our results also highlight how can help better understand Alzheimer’s brain phenotype. The is openly available download data analysis GNPS platform investigators hypothesize candidate structures unknown data.

Language: Английский

Citations

42
Soil metabolomics: Deciphering underground metabolic webs in terrestrial ecosystems DOI Creative Commons
Yang Song,

Shi Yao,

Xiaona Li

et al.

Eco-Environment & Health, Journal Year: 2024, Volume and Issue: 3(2), P. 227 - 237

Published: March 21, 2024

Soil metabolomics is an emerging approach for profiling diverse small molecule metabolites, i.e., metabolomes, in the soil. including fatty acids, amino lipids, organic sugars, and volatile compounds, often contain essential nutrients such as nitrogen, phosphorus, sulfur are directly linked to soil biogeochemical cycles driven by microorganisms. This paper presents overview of methods analyzing metabolites state-of-the-art relation nutrient cycling. We describe important applications studying carbon cycling sequestration, response pools changing environmental conditions. includes using provide new insights into close relationships between microbiome metabolome, well responses metabolome plant stresses contamination. also highlight advantage study elements suggest that future research needs better understand factors driving function health.

Language: Английский

Citations

9
ModiFinder: Tandem Mass Spectral Alignment Enables Structural Modification Site Localization DOI
Mohammad Reza Zare Shahneh,

Michael Strobel,

Giovanni Andrea Vitale

et al.

Journal of the American Society for Mass Spectrometry, Journal Year: 2024, Volume and Issue: unknown

Published: June 3, 2024

Untargeted tandem mass spectrometry (MS/MS) has become a high-throughput method to measure small molecules in complex samples. One key goal is the transformation of these MS/MS spectra into chemical structures. Computational techniques such as library search have enabled reidentification known compounds. Analog and molecular networking extend this identification unknown While there been advancements metrics for similarity structurally similar compounds, still lack automated methods provide site specific information about structural modifications. Here we introduce ModiFinder which leverages alignment peaks between related molecules. Specifically, focuses on shifted fragment alignment. These putatively represent substructures molecule that contain modification. synthesizes together scores likelihood each atom be modification site. We demonstrate manuscript how can effectively localize modifications extends capabilities analog searching accelerate discovery novel

Language: Английский

Citations

7
Bridging knowledge gaps in human chemical exposure via drinking water with non-target screening DOI Creative Commons
Davide Ciccarelli, Saer Samanipour, Helena Rapp-Wright

et al.

Critical Reviews in Environmental Science and Technology, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 25

Published: Sept. 1, 2024

Fundamental knowledge gaps still exist in the exposome, especially regarding analytical space coverage, mapping and prioritization of a very large number diverse chemical structures. This review focuses on contributions suspect non-target screening (NTS) to contaminants characterization toxicity assessment drinking water. A comprehensive publications from 2013-2024 revealed only 172 substances identified with certainty using NTS 17 countries. The approaches, their complementarity, effectiveness use compound identification frameworks are discussed. 'intelligent' tools (including machine learning) aid substance identification, is emerging. Strategies for integration epidemiology also considered, including re-use existing data. holds great potential exposure water its contribution exposome.

Language: Английский

Citations

7