Flavor is the main factor driving consumers acceptance of food products. However, tracking the biochemistry of flavor is a formidable challenge due to the complexity of food composition. Current methodologies for linking individual molecules to flavor in foods and beverages are expensive and time-consuming. Predictive models based on machine learning (ML) are emerging as an alternative to speed up this process. Nonetheless, the optimal approach to predict flavor features of molecules remains elusive. In this work we present FlavorMiner, an ML-based multilabel flavor predictor. FlavorMiner seamlessly integrates different combinations of algorithms and mathematical representations, augmented with class balance strategies to address the inherent class of the input dataset. Notably, Random Forest and K-Nearest Neighbors combined with Extended Connectivity Fingerprint and RDKit molecular descriptors consistently outperform other combinations in most cases. Resampling strategies surpass weight balance methods in mitigating bias associated with class imbalance. FlavorMiner exhibits remarkable accuracy, with an average ROC AUC score of 0.88. This algorithm was used to analyze cocoa metabolomics data, unveiling its profound potential to help extract valuable insights from intricate food metabolomics data. FlavorMiner can be used for flavor mining in any food product, drawing from a diverse training dataset that spans over 934 distinct food products.Scientific Contribution FlavorMiner is an advanced machine learning (ML)-based tool designed to predict molecular flavor features with high accuracy and efficiency, addressing the complexity of food metabolomics. By leveraging robust algorithmic combinations paired with mathematical representations FlavorMiner achieves high predictive performance. Applied to cocoa metabolomics, FlavorMiner demonstrated its capacity to extract meaningful insights, showcasing its versatility for flavor analysis across diverse food products. This study underscores the transformative potential of ML in accelerating flavor biochemistry research, offering a scalable solution for the food and beverage industry.

Mapping the chemical space of compounds to chemical structures remains a challenge in metabolomics. Despite the advancements in untargeted liquid chromatography-mass spectrometry (LC–MS) to achieve a high-throughput profile of metabolites from complex biological resources, only a small fraction of these metabolites can be annotated with confidence. Many novel computational methods and tools have been developed to enable chemical structure annotation to known and unknown compounds such as in silico generated spectra and molecular networking. Here, we present an automated and reproducible Metabolome Annotation Workflow (MAW) for untargeted metabolomics data to further facilitate and automate the complex annotation by combining tandem mass spectrometry (MS2) input data pre-processing, spectral and compound database matching with computational classification, and in silico annotation. MAW takes the LC-MS2 spectra as input and generates a list of putative candidates from spectral and compound databases. The databases are integrated via the R package Spectra and the metabolite annotation tool SIRIUS as part of the R segment of the workflow (MAW-R). The final candidate selection is performed using the cheminformatics tool RDKit in the Python segment (MAW-Py). Furthermore, each feature is assigned a chemical structure and can be imported to a chemical structure similarity network. MAW is following the FAIR (Findable, Accessible, Interoperable, Reusable) principles and has been made available as the docker images, maw-r and maw-py. The source code and documentation are available on GitHub (https://github.com/zmahnoor14/MAW). The performance of MAW is evaluated on two case studies. MAW can improve candidate ranking by integrating spectral databases with annotation tools like SIRIUS which contributes to an efficient candidate selection procedure. The results from MAW are also reproducible and traceable, compliant with the FAIR guidelines. Taken together, MAW could greatly facilitate automated metabolite characterization in diverse fields such as clinical metabolomics and natural product discovery.

Compound (or chemical) databases are an invaluable resource for many scientific disciplines. Exposomics researchers need to find and identify relevant chemicals that cover the entirety of potential (chemical and other) exposures over entire lifetimes. This daunting task, with over 100 million chemicals in the largest chemical databases, coupled with broadly acknowledged knowledge gaps in these resources, leaves researchers faced with too much—yet not enough—information at the same time to perform comprehensive exposomics research. Furthermore, the improvements in analytical technologies and computational mass spectrometry workflows coupled with the rapid growth in databases and increasing demand for high throughput “big data” services from the research community present significant challenges for both data hosts and workflow developers. This article explores how to reduce candidate search spaces in non-target small molecule identification workflows, while increasing content usability in the context of environmental and exposomics analyses, so as to profit from the increasing size and information content of large compound databases, while increasing efficiency at the same time. In this article, these methods are explored using PubChem, the NORMAN Network Suspect List Exchange and the in silico fragmentation approach MetFrag. A subset of the PubChem database relevant for exposomics, PubChemLite, is presented as a database resource that can be (and has been) integrated into current workflows for high resolution mass spectrometry. Benchmarking datasets from earlier publications are used to show how experimental knowledge and existing datasets can be used to detect and fill gaps in compound databases to progressively improve large resources such as PubChem, and topic-specific subsets such as PubChemLite. PubChemLite is a living collection, updating as annotation content in PubChem is updated, and exported to allow direct integration into existing workflows such as MetFrag. The source code and files necessary to recreate or adjust this are jointly hosted between the research parties (see data availability statement). This effort shows that enhancing the FAIRness (Findability, Accessibility, Interoperability and Reusability) of open resources can mutually enhance several resources for whole community benefit. The authors explicitly welcome additional community input on ideas for future developments.

AbstractWe report the major conclusions of the online open-access workshop “Computational Applications in Secondary Metabolite Discovery (CAiSMD)” that took place from 08 to 10 March 2021. Invited speakers from academia and industry and about 200 registered participants from five continents (Africa, Asia, Europe, South America, and North America) took part in the workshop. The workshop highlighted the potential applications of computational methodologies in the search for secondary metabolites (SMs) or natural products (NPs) as potential drugs and drug leads. During 3 days, the participants of this online workshop received an overview of modern computer-based approaches for exploring NP discovery in the “omics” age. The invited experts gave keynote lectures, trained participants in hands-on sessions, and held round table discussions. This was followed by oral presentations with much interaction between the speakers and the audience. Selected applicants (early-career scientists) were offered the opportunity to give oral presentations (15 min) and present posters in the form of flash presentations (5 min) upon submission of an abstract. The final program available on the workshop website (https://caismd.indiayouth.info/) comprised of 4 keynote lectures (KLs), 12 oral presentations (OPs), 2 round table discussions (RTDs), and 5 hands-on sessions (HSs). This meeting report also references internet resources for computational biology in the area of secondary metabolites that are of use outside of the workshop areas and will constitute a long-term valuable source for the community. The workshop concluded with an online survey form to be completed by speakers and participants for the goal of improving any subsequent editions.

Liquid chromatography coupled with high-resolution mass spectrometry (LC-HRMS) is increasingly popular for the non-targeted exploration of complex samples, where tandem mass spectrometry (MS/MS) is used to characterize the structure of unknown compounds. However, mass spectra do not always contain sufficient information to unequivocally identify the correct structure. This study investigated how much additional information can be gained using hydrogen deuterium exchange (HDX) experiments. The exchange of “easily exchangeable” hydrogen atoms (connected to heteroatoms), with predominantly [M+D]+ ions in positive mode and [M-D]− in negative mode was observed. To enable high-throughput processing, new scoring terms were incorporated into the in silico fragmenter MetFrag. These were initially developed on small datasets and then tested on 762 compounds of environmental interest. Pairs of spectra (normal and deuterated) were found for 593 of these substances (506 positive mode, 155 negative mode spectra). The new scoring terms resulted in 29 additional correct identifications (78 vs 49) for positive mode and an increase in top 10 rankings from 80 to 106 in negative mode. Compounds with dual functionality (polar head group, long apolar tail) exhibited dramatic retention time (RT) shifts of up to several minutes, compared with an average 0.04 min RT shift. For a smaller dataset of 80 metabolites, top 10 rankings improved from 13 to 24 (positive mode, 57 spectra) and from 14 to 31 (negative mode, 63 spectra) when including HDX information. The results of standard measurements were confirmed using targets and tentatively identified surfactant species in an environmental sample collected from the river Danube near Novi Sad (Serbia). The changes to MetFrag have been integrated into the command line version available at http://c-ruttkies.github.io/MetFrag and all resulting spectra and compounds are available in online resources and in the Electronic Supplementary Material (ESM).

Ceramides (CERs) are the backbone of the intercellular lipid lamellae of the stratum corneum (SC), the outer layer of the skin. Skin diseases such as atopic dermatitis, psoriasis, and aged skin are characterized by dysfunctional skin barrier and dryness which are associated with reduced levels of CERs. Replenishing the depleted epidermal CERs with exogenous CERs has been shown to have beneficial effects in improving the skin barrier and hydration. The exogenous CERs such as phyto-derived CERs (PhytoCERs) can be delivered deep into the SC using novel topical formulations. This, however, requires investigating the rate and extent of skin permeation of CERs. In this study, an LC/APCI-MS method to detect and quantify PhytoCERs in different layers of the skin has been developed and validated. The method was used to investigate the skin permeation of PhytoCERs using Franz diffusion cells after applying an amphiphilic cream containing PhytoCERs to the surface of ex vivo human skin. As plant-specific CERs are not commercially available, well-characterized CERs isolated from oat (Avena abyssinica) were used as reference standards for the development and validation of the method. The method was linear over the range of 30–1050 ng/mL and sensitive with limit of detection and quantification of 10 and 30 ng/mL, respectively. The method was also selective, accurate, and precise with minimal matrix effect (with mean matrix factor around 100%). Even if more than 85% of oat CERs in the cream remained in the cream after the incubation periods of 30, 100, and 300 min, it was possible to quantify the small quantities of oat CERs distributed across the SC, epidermis, and dermis of the skin indicating the method’s sensitivity. Therefore, the method can be used to investigate the skin permeation of oat CERs from the various pharmaceutical and cosmeceutical products without any interference from the skin constituents such as the epidermal lipids.

In nontarget screening, structure elucidation of small molecules from high resolution mass spectrometry (HRMS) data is challenging, particularly the selection of the most likely candidate structure among the many retrieved from compound databases. Several fragmentation and retention prediction methods have been developed to improve this candidate selection. In order to evaluate their performance, we compared two in silico fragmenters (MetFrag and CFM-ID) and two retention time prediction models (based on the chromatographic hydrophobicity index (CHI) and on log D). A set of 78 known organic micropollutants was analyzed by liquid chromatography coupled to a LTQ Orbitrap HRMS with electrospray ionization (ESI) in positive and negative mode using two fragmentation techniques with different collision energies. Both fragmenters (MetFrag and CFM-ID) performed well for most compounds, with average ranking the correct candidate structure within the top 25% and 22 to 37% for ESI+ and ESI− mode, respectively. The rank of the correct candidate structure slightly improved when MetFrag and CFM-ID were combined. For unknown compounds detected in both ESI+ and ESI−, generally positive mode mass spectra were better for further structure elucidation. Both retention prediction models performed reasonably well for more hydrophobic compounds but not for early eluting hydrophilic substances. The log D prediction showed a better accuracy than the CHI model. Although the two fragmentation prediction methods are more diagnostic and sensitive for candidate selection, the inclusion of retention prediction by calculating a consensus score with optimized weighting can improve the ranking of correct candidates as compared to the individual methods.

Chemical database searching has become a fixture in many non-targeted identification workflows based on high-resolution mass spectrometry (HRMS). However, the form of a chemical structure observed in HRMS does not always match the form stored in a database (e.g., the neutral form versus a salt; one component of a mixture rather than the mixture form used in a consumer product). Linking the form of a structure observed via HRMS to its related form(s) within a database will enable the return of all relevant variants of a structure, as well as the related metadata, in a single query. A Konstanz Information Miner (KNIME) workflow has been developed to produce structural representations observed using HRMS (“MS-Ready structures”) and links them to those stored in a database. These MS-Ready structures, and associated mappings to the full chemical representations, are surfaced via the US EPA’s Chemistry Dashboard (https://comptox.epa.gov/dashboard/). This article describes the workflow for the generation and linking of ~ 700,000 MS-Ready structures (derived from ~ 760,000 original structures) as well as download, search and export capabilities to serve structure identification using HRMS. The importance of this form of structural representation for HRMS is demonstrated with several examples, including integration with the in silico fragmentation software application MetFrag. The structures, search, download and export functionality are all available through the CompTox Chemistry Dashboard, while the MetFrag implementation can be viewed at https://msbi.ipb-halle.de/MetFragBeta/.

BackgroundThe fourth round of the Critical Assessment of Small Molecule Identification (CASMI) Contest (www.casmi-contest.org) was held in 2016, with two new categories for automated methods. This article covers the 208 challenges in Categories 2 and 3, without and with metadata, from organization, participation, results and post-contest evaluation of CASMI 2016 through to perspectives for future contests and small molecule annotation/identification.ResultsThe Input Output Kernel Regression (CSI:IOKR) machine learning approach performed best in “Category 2: Best Automatic Structural Identification—In Silico Fragmentation Only”, won by Team Brouard with 41% challenge wins. The winner of “Category 3: Best Automatic Structural Identification—Full Information” was Team Kind (MS-FINDER), with 76% challenge wins. The best methods were able to achieve over 30% Top 1 ranks in Category 2, with all methods ranking the correct candidate in the Top 10 in around 50% of challenges. This success rate rose to 70% Top 1 ranks in Category 3, with candidates in the Top 10 in over 80% of the challenges. The machine learning and chemistry-based approaches are shown to perform in complementary ways.ConclusionsThe improvement in (semi-)automated fragmentation methods for small molecule identification has been substantial. The achieved high rates of correct candidates in the Top 1 and Top 10, despite large candidate numbers, open up great possibilities for high-throughput annotation of untargeted analysis for “known unknowns”. As more high quality training data becomes available, the improvements in machine learning methods will likely continue, but the alternative approaches still provide valuable complementary information. Improved integration of experimental context will also improve identification success further for “real life” annotations. The true “unknown unknowns” remain to be evaluated in future CASMI contests.

Advanced glycation end products (AGEs) are posttranslational modifications formed non-enzymatically from the reaction of carbohydrates and their degradation products with proteins. Accumulation of AGEs is associated with the progression of severe diabetic complications, for example, and elevated tissue levels of AGEs might even predict these pathologies. As AGE formation is often site-specific, mapping of these modification sites may reveal more sensitive and specific markers than the global tissue level. Here, 42 AGE modifications were identified in a bottom-up proteomic approach by tandem mass spectrometry, which corresponded to 36 sites in 22 high to medium abundant proteins in individual plasma samples obtained from type 2 diabetes mellitus (T2DM) patients with long disease duration (>10 years). Major modifications were glarg (11 modification sites) and carboxymethylation (5) of arginine and formylation (8), acetylation (7), and carboxymethylation (7) of lysine residues. Relative quantification of these sites in plasma samples obtained from normoglycemic individuals (n = 47) and patients with T2DM being newly diagnosed (n = 47) or of medium (2–5 years, n = 20) and long disease duration (>10 years, n = 20) did not reveal any significant differences.