Unknown Metabolite Identification Using Machine Learning Collision Cross-Section Prediction and Tandem Mass Spectrometry

Association between human blood metabolome and risk of myocarditis: a mendelian randomization study

Myocarditis is a common disease of the cardiovascular and immune systems, but the relationship between relevant blood metabolites and the risk of myocarditis has not been well-established. To identify potential biometabolic markers associated with myocarditis, we conducted a two-sample Mendelian randomization (MR) study. We performed preliminary MR analysis using the inverse variance weighted (IVW) method, supplemented by MR-Egger, weighted median, and weighted mode methods to adjust for false discovery rate (FDR). Confounders were screened using the GWAS Catalog website. Sensitivity analyses included Cochrane Q-test, Egger regression, Mendelian Randomization Pleiotropy RESidual Sum and Outlier (MR-PRESSO), scatterplots, funnel plots, and forest plots. For genetic and directional analysis, we employed co-localization analysis and the Steiger test. MR analysis was performed using the FinnGen database and meta-analysis was performed using the IEU database. MR analysis identified significant correlations for five metabolic biomarkers after FDR correction. These included four known metabolites: kynurenine, 1-stearoyl-GPE (18:0), deoxycarnitine, and 5-acetylamino-6-formylamino-3-methyluracil, as well as one unknown metabolite, X-25,422. Among these, kynurenine (OR = 1.441, 95%CI = 1.089-1.906, p-value = 0.018) and 1-stearoyl-GPE (18:0) (OR = 1.263, 95%CI = 1.029-1.550, p-value = 0.029) were identified as risk factors for myocarditis, while deoxycarnitine (OR = 0.813, 95%CI = 0.676-0.979, p-value = 0.029), 5-acetylamino-6-formylamino-3-methyluracil (OR = 0.864, 95% CI = 0.775-0.962, p-value = 0.018), and X-25,422 (OR = 0.721, 95%CI = 0.587-0.886, p-value = 0.009) were found to be protective factors. No evidence of heterogeneity, horizontal pleiotropy, or sensitivity issues was observed, and no shared genetic factors between exposure and outcome were detected. The causality was in the correct direction. Meta-analysis further confirmed the causal relationship between the five metabolites and myocarditis. This study identifies a causal relationship between five circulating metabolites and myocarditis. Kynurenine, 1-stearoyl-GPE (18:0), deoxycarnitine, X-25,422, and 5-acetylamino-6-formylamino-3-methyluracil may serve as potential drug targets for myocarditis, providing a theoretical basis for the prevention, diagnosis, and treatment of the condition.

Critical review on in silico methods for structural annotation of chemicals detected with LC/HRMS non-targeted screening

Non-targeted screening with liquid chromatography coupled to high-resolution mass spectrometry (LC/HRMS) is increasingly leveraging in silico methods, including machine learning, to obtain candidate structures for structural annotation of LC/HRMS features and their further prioritization. Candidate structures are commonly retrieved based on the tandem mass spectral information either from spectral or structural databases; however, the vast majority of the detected LC/HRMS features remain unannotated, constituting what we refer to as a part of the unknown chemical space. Recently, the exploration of this chemical space has become accessible through generative models. Furthermore, the evaluation of the candidate structures benefits from the complementary empirical analytical information such as retention time, collision cross section values, and ionization type. In this critical review, we provide an overview of the current approaches for retrieving and prioritizing candidate structures. These approaches come with their own set of advantages and limitations, as we showcase in the example of structural annotation of ten known and ten unknown LC/HRMS features. We emphasize that these limitations stem from both experimental and computational considerations. Finally, we highlight three key considerations for the future development of in silico methods.

Recent Developments in Machine Learning for Mass Spectrometry

Statistical analysis and modeling of mass spectrometry (MS) data have a long and rich history with several modern MS-based applications using statistical and chemometric methods. Recently, machine learning (ML) has experienced a renaissance due to advents in computational hardware and the development of new algorithms for artificial neural networks (ANN) and deep learning architectures. Moreover, recent successes of new ANN and deep learning architectures in several areas of science, engineering, and society have further strengthened the ML field. Importantly, modern ML methods and architectures have enabled new approaches for tasks related to MS that are now widely adopted in several popular MS-based subdisciplines, such as mass spectrometry imaging and proteomics. Herein, we aim to provide an introductory summary of the practical aspects of ML methodology relevant to MS. Additionally, we seek to provide an up-to-date review of the most recent developments in ML integration with MS-based techniques while also providing critical insights into the future direction of the field.

Predicting the Predicted: A Comparison of Machine Learning-Based Collision Cross-Section Prediction Models for Small Molecules

The application of machine learning (ML) to -omics research is growing at an exponential rate owing to the increasing availability of large amounts of data for model training. Specifically, in metabolomics, ML has enabled the prediction of tandem mass spectrometry and retention time data. More recently, due to the advent of ion mobility, new ML models have been introduced for collision cross-section (CCS) prediction, but those have been trained with different and relatively small data sets covering a few thousands of small molecules, which hampers their systematic comparison. Here, we compared four existing ML-based CCS prediction models and their capacity to predict CCS values using the recently introduced METLIN-CCS data set. We also compared them with simple linear models and with ML models that used fingerprints as regressors. We analyzed the role of structural diversity of the data on which the ML models are trained with and explored the practical application of these models for metabolite annotation using CCS values. Results showed a limited capability of the existing models to achieve the necessary accuracy to be adopted for routine metabolomics analysis. We showed that for a particular molecule, this accuracy could only be improved when models were trained with a large number of structurally similar counterparts. Therefore, we suggest that current annotation capabilities will only be significantly altered with models trained with heterogeneous data sets composed of large homogeneous hubs of structurally similar molecules to those being predicted.

Computational tools and algorithms for ion mobility spectrometry-mass spectrometry

Ion mobility spectrometry-mass spectrometry (IMS-MS or IM-MS) is a powerful analytical technique that combines the gas-phase separation capabilities of IM with the identification and quantification capabilities of MS. IM-MS can differentiate molecules with indistinguishable masses but different structures (e.g., isomers, isobars, molecular classes, and contaminant ions). The importance of this analytical technique is reflected by a staged increase in the number of applications for molecular characterization across a variety of fields, from different MS-based omics (proteomics, metabolomics, lipidomics, etc.) to the structural characterization of glycans, organic matter, proteins, and macromolecular complexes. With the increasing application of IM-MS there is a pressing need for effective and accessible computational tools. This article presents an overview of the most recent free and open-source software tools specifically tailored for the analysis and interpretation of data derived from IM-MS instrumentation. This review enumerates these tools and outlines their main algorithmic approaches, while highlighting representative applications across different fields. Finally, a discussion of current limitations and expectable improvements is presented.

Offline Two-Dimensional Liquid Chromatography-Mass Spectrometry for Deep Annotation of the Fecal Metabolome Following Fecal Microbiota Transplantation

Biological interpretation of untargeted LC-MS-based metabolomics data depends on accurate compound identification, but current techniques fall short of identifying most features that can be detected. The human fecal metabolome is complex, variable, incompletely annotated, and serves as an ideal matrix to evaluate novel compound identification methods. We devised an experimental strategy for compound annotation using multidimensional chromatography and semiautomated feature alignment and applied these methods to study the fecal metabolome in the context of fecal microbiota transplantation (FMT) for recurrent C. difficile infection. Pooled fecal samples were fractionated using semipreparative liquid chromatography and analyzed by an orthogonal LC-MS/MS method. The resulting spectra were searched against commercial, public, and local spectral libraries, and annotations were vetted using retention time alignment and prediction. Multidimensional chromatography yielded more than a 2-fold improvement in identified compounds compared to conventional LC-MS/MS and successfully identified several rare and previously unreported compounds, including novel fatty-acid conjugated bile acid species. Using an automated software-based feature alignment strategy, most metabolites identified by the new approach could be matched to features that were detected but not identified in single-dimensional LC-MS/MS data. Overall, our approach represents a powerful strategy to enhance compound identification and biological insight from untargeted metabolomics data.

Triboelectric Nanogenerators for the Masses: A Low-Cost Do-It-Yourself Pulsed Ion Source for Sample-Limited Applications

Triboelectric nanogenerators (TENG) are useful devices for converting mechanical motion into electric current using readily available materials. Though the applications for these devices span across many fields, TENG can be leveraged for mass spectrometry (MS) as inexpensive and effective power supplies for pulsed nanoelectrospray ionization (nESI). The inherently discontinuous spray provided by TENG is particularly useful in scenarios where high sample economy is imperative, as in the case of ultraprecious samples. Previous work has shown the utility of TENG MS as a highly sensitive technique capable of yielding quality spectra from only a few microliters of sample at low micromolar concentrations. As the field of miniaturized, fieldable mass spectrometers grows, it remains critical to develop advanced ion sources with similarly small power requirements and footprints. Here, we present a redesigned TENG ion source with a sub-1000 USD material cost, lower power consumption, reduced footprint, and improved capabilities. We validate the performance of this new device for a diverse set of applications, including lipid double bond localization and native protein analysis.

HILIC-IM-MS for Simultaneous Lipid and Metabolite Profiling of Bacteria

Although MALDI-ToF platforms for microbial identifications have found great success in clinical microbiology, the sole use of protein fingerprints for the discrimination of closely related species, strain-level identifications, and detection of antimicrobial resistance remains a challenge for the technology. Several alternative mass spectrometry-based methods have been proposed to address the shortcomings of the protein-centric approach, including MALDI-ToF methods for fatty acid/lipid profiling and LC-MS profiling of metabolites. However, the molecular diversity of microbial pathogens suggests that no single "ome" will be sufficient for the accurate and sensitive identification of strain- and susceptibility-level profiling of bacteria. Here, we describe the development of an alternative approach to microorganism profiling that relies upon both metabolites and lipids rather than a single class of biomolecule. Single-phase extractions based on butanol, acetonitrile, and water (the BAW method) were evaluated for the recovery of lipids and metabolites from Gram-positive and -negative microorganisms. We found that BAW extraction solutions containing 45% butanol provided optimal recovery of both molecular classes in a single extraction. The single-phase extraction method was coupled to hydrophilic interaction liquid chromatography (HILIC) and ion mobility-mass spectrometry (IM-MS) to resolve similar-mass metabolites and lipids in three dimensions and provide multiple points of evidence for feature annotation in the absence of tandem mass spectrometry. We demonstrate that the combined use of metabolites and lipids can be used to differentiate microorganisms to the species- and strain-level for four of the ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Acinetobacter baumannii, and Pseudomonas aeruginosa) using data from a single ionization mode. These results present promising, early stage evidence for the use of multiomic signatures for the identification of microorganisms by liquid chromatography, ion mobility, and mass spectrometry that, upon further development, may improve upon the level of identification provided by current methods.

MOCCal: A Multiomic CCS Calibrator for Traveling Wave Ion Mobility Mass Spectrometry

Ion mobility mass spectrometry (IM-MS) is a rapid, gas-phase separation technology that can resolve ions on the basis of their size-to-charge and mass-to-charge ratios. Since each class of biomolecule has a unique relationship between size and mass, IM-MS spectra of complex biological samples are organized into trendlines that each contain one type of biomolecule (i.e., lipid, peptide, metabolite). These trendlines can aid in the identification of unknown ions by providing a general classification, while more specific identifications require the conversion of IM arrival times to collision cross section (CCS) values to minimize instrument-to-instrument variability. However, the process of converting IM arrival times to CCS values varies between the different IM devices. Arrival times from traveling wave ion mobility (TWIM) devices must undergo a calibration process to obtain CCS values, which can impart biases if the calibrants are not structurally similar to the analytes. For multiomic mixtures, several different types of calibrants must be used to obtain the most accurate CCS values from TWIM platforms. Here we describe the development of a multiomic CCS calibration tool, MOCCal, to automate the assignment of unknown features to the power law calibration that provides the most accurate CCS value. MOCCal calibrates every experimental arrival time with up to three class-specific calibration curves and uses the difference (in Å2) between the calibrated TWCCSN2 value and DTCCSN2 vs m/z regression lines to determine the best calibration curve. Using real and simulated multiomic samples, we demonstrate that MOCCal provides accurately calibrated TWCCSN2 values for small molecules, lipids, and peptides.

Application of Ion Mobility Spectrometry and the Derived Collision Cross Section in the Analysis of Environmental Organic Micropollutants

Ion mobility spectrometry (IMS) is a rapid gas-phase separation technique, which can distinguish ions on the basis of their size, shape, and charge. The IMS-derived collision cross section (CCS) can serve as additional identification evidence for the screening of environmental organic micropollutants (OMPs). In this work, we summarize the published experimental CCS values of environmental OMPs, introduce the current CCS prediction tools, summarize the use of IMS and CCS in the analysis of environmental OMPs, and finally discussed the benefits of IMS and CCS in environmental analysis. An up-to-date CCS compendium for environmental contaminants was produced by combining CCS databases and data sets of particular types of environmental OMPs, including pesticides, drugs, mycotoxins, steroids, plastic additives, per- and polyfluoroalkyl substances (PFAS), polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), and polybrominated diphenyl ethers (PBDEs), as well as their well-known transformation products. A total of 9407 experimental CCS values from 4170 OMPs were retrieved from 23 publications, which contain both drift tube CCS in nitrogen (DTCCSN2) and traveling wave CCS in nitrogen (TWCCSN2). A selection of publicly accessible and in-house CCS prediction tools were also investigated; the chemical space covered by the training set and the quality of CCS measurements seem to be vital factors affecting the CCS prediction accuracy. Then, the applications of IMS and the derived CCS in the screening of various OMPs were summarized, and the benefits of IMS and CCS, including increased peak capacity, the elimination of interfering ions, the separation of isomers, and the reduction of false positives and false negatives, were discussed in detail. With the improvement of the resolving power of IMS and enhancements of experimental CCS databases, the practicability of IMS in the analysis of environmental OMPs will continue to improve.

Investigating the Potential of Drift Tube Ion Mobility for the Analysis of Oxidized Lipids

Epilipids, a subset of the lipidome that comprises oxidized, nitrated, and halogenated lipid species, show important biochemical activity in the regulation of redox lipid metabolism by influencing cell fate decisions, including death, health, and aging. Due to the large chemical diversity, reversed-phase liquid chromatography-high-resolution mass spectrometry (RPLC-HRMS) methods have only a limited ability to separate numerous isobaric and isomeric epilipids. Ion mobility spectrometry (IMS) is a gas-phase separation technique that can be combined with LC-HRMS to improve the overall peak capacity of the analytical platform. Here, we illustrate the advantages and discuss the current limitations of implementing IMS in LC-HRMS workflows for the analysis of oxylipins and oxidized complex lipids. Using isomeric mixtures of oxylipins, we demonstrated that while deprotonated ions of eicosanoids were poorly resolved by IMS, sodium acetate and metal adducts (e.g., Li, Na, Ag, Ba, K) of structural isomers often showed ΔCCS% above 1.4% and base peak separation with high-resolution demultiplexing (HRDm). The knowledge of the IM migration order was also used as a proof of concept to help in the annotation of oxidized complex lipids using HRDm and all-ion fragmentation spectra. Additionally, we used a mixture of deuterium-labeled lipids for a routine system suitability test with the purpose of improving harmonization and interoperability of IMS data sets in (epi)lipidomics.

Collision cross section measurement and prediction methods in omics

Omics studies such as metabolomics, lipidomics, and proteomics have become important for understanding the mechanisms in living organisms. However, the compounds detected are structurally different and contain isomers, with each structure or isomer leading to a different result in terms of the role they play in the cell or tissue in the organism. Therefore, it is important to detect, characterize, and elucidate the structures of these compounds. Liquid chromatography and mass spectrometry have been utilized for decades in the structure elucidation of key compounds. While prediction models of parameters (such as retention time and fragmentation pattern) have also been developed for these separation techniques, they have some limitations. Moreover, ion mobility has become one of the most promising techniques to give a fingerprint to these compounds by determining their collision cross section (CCS) values, which reflect their shape and size. Obtaining accurate CCS enables its use as a filter for potential analyte structures. These CCS values can be measured experimentally using calibrant-independent and calibrant-dependent approaches. Identification of compounds based on experimental CCS values in untargeted analysis typically requires CCS references from standards, which are currently limited and, if available, would require a large amount of time for experimental measurements. Therefore, researchers use theoretical tools to predict CCS values for untargeted and targeted analysis. In this review, an overview of the different methods for the experimental and theoretical estimation of CCS values is given where theoretical prediction tools include computational and machine modeling type approaches. Moreover, the limitations of the current experimental and theoretical approaches and their potential mitigation methods were discussed.

Array-Based Machine Learning for Functional Group Detection in Electron Ionization Mass Spectrometry

Mass spectrometry is a ubiquitous technique capable of complex chemical analysis. The fragmentation patterns that appear in mass spectrometry are an excellent target for artificial intelligence methods to automate and expedite the analysis of data to identify targets such as functional groups. To develop this approach, we trained models on electron ionization (a reproducible hard fragmentation technique) mass spectra so that not only the final model accuracies but also the reasoning behind model assignments could be evaluated. The convolutional neural network (CNN) models were trained on 2D images of the spectra using transfer learning of Inception V3, and the logistic regression models were trained using array-based data and Scikit Learn implementation in Python. Our training dataset consisted of 21,166 mass spectra from the United States' National Institute of Standards and Technology (NIST) Webbook. The data was used to train models to identify functional groups, both specific (e.g., amines, esters) and generalized classifications (aromatics, oxygen-containing functional groups, and nitrogen-containing functional groups). We found that the highest final accuracies on identifying new data were observed using logistic regression rather than transfer learning on CNN models. It was also determined that the mass range most beneficial for functional group analysis is 0-100 m/z. We also found success in correctly identifying functional groups of example molecules selected from both the NIST database and experimental data. Beyond functional group analysis, we also have developed a methodology to identify impactful fragments for the accurate detection of the models' targets. The results demonstrate a potential pathway for analyzing and screening substantial amounts of mass spectral data.

MassCCS: A High-Performance Collision Cross-Section Software for Large Macromolecular Assemblies

Ion mobility mass spectrometry (IM-MS) techniques have become highly valued as a tool for structural characterization of biomolecular systems since they yield accurate measurements of the rotationally averaged collision cross-section (CCS) against a buffer gas. Despite its enormous potential, IM-MS data interpretation is often challenging due to the conformational isomerism of metabolites, lipids, proteins, and other biomolecules in the gas phase. Therefore, reliable and fast CCS calculations are needed to help interpret IM-MS data. In this work, we present MassCCS, a parallelized open-source code for computing CCS of molecules ranging from small organic compounds to massive protein assemblies at the trajectory method level of description using atomic and molecular buffer gas particles. The performance of the code is comparable to other available software for small molecules and proteins but is significantly faster for larger macromolecular assemblies. We performed extensive tests regarding accuracy, performance, and scalability with system size and number of CPU cores. MassCCS has proven highly accurate and efficient, with execution times under a few minutes, even for large (84.87 MDa) virus capsid assemblies with very modest computational resources. MassCCS is freely available at https://github.com/cces-cepid/massccs.

Ion Mobility Mass Spectrometry (IM-MS) for Structural Biology: Insights Gained by Measuring Mass, Charge, and Collision Cross Section

The investigation of macromolecular biomolecules with ion mobility mass spectrometry (IM-MS) techniques has provided substantial insights into the field of structural biology over the past two decades. An IM-MS workflow applied to a given target analyte provides mass, charge, and conformation, and all three of these can be used to discern structural information. While mass and charge are determined in mass spectrometry (MS), it is the addition of ion mobility that enables the separation of isomeric and isobaric ions and the direct elucidation of conformation, which has reaped huge benefits for structural biology. In this review, where we focus on the analysis of proteins and their complexes, we outline the typical features of an IM-MS experiment from the preparation of samples, the creation of ions, and their separation in different mobility and mass spectrometers. We describe the interpretation of ion mobility data in terms of protein conformation and how the data can be compared with data from other sources with the use of computational tools. The benefit of coupling mobility analysis to activation via collisions with gas or surfaces or photons photoactivation is detailed with reference to recent examples. And finally, we focus on insights afforded by IM-MS experiments when applied to the study of conformationally dynamic and intrinsically disordered proteins.