Untargeted high-resolution paired mass distance data mining for retrieving general chemical relationships

The impact of sampling time point on the lipidome composition

Lipidomic profiling has been reported as an effective approach for characterizing and differentiating brain tumors. However, since lipids can undergo non-specific enzymatic and nonenzymatic reactions due to tissue disruption, it is critical to consider the preanalytical phase of the diagnostic process (e.g., optimizing the sampling time and sampling conditions). Thus, this study assesses the ways in which the time point of sampling impacts the lipidome composition of brain tumors. Two histologically distinct brain tumors-namely, meningiomas and gliomas-were sampled using solid-phase microextraction (SPME) fibers at two time points: on-site directly after removal, and after 12 months of storage at -30 °C. The samples were analyzed via HILIC chromatography coupled with HRMS, which enabled the detection of a wide range of features, including phospholipids and sphingolipids, as well as changes in the profiles of these compounds. The samples obtained from the stored tissues tended to have elevated levels of analytes with lower m/z values. In addition, the samples obtained from the fresh and stored tissues were easily distinguished based on their lipidome compositions, regardless of the histological tumor type. Notably, while storage did not affect the possibility of differentiating meningiomas and gliomas, the biological interpretation of the obtained results were prone to bias.

Active Molecular Network Discovery Links Lifestyle Variables to Breast Cancer in the Long Island Breast Cancer Study Project

A healthy lifestyle has been associated with decreased risk of developing breast cancer. Using untargeted metabolomics profiling, which provides unbiased information regarding lifestyle choices such as diet and exercise, we aim to identify the molecular mechanisms connecting lifestyle and breast cancer through network analysis. A total of 100 postmenopausal women, 50 with breast cancer and 50 cancer-free controls, were selected from the Long Island Breast Cancer Study Project (LIBCSP). We measured untargeted plasma metabolomics using liquid chromatography-high-resolution mass spectrometry (LC-HRMS). Using the "enet" package, we retained highly correlated metabolites representing active molecular network (AMN) clusters for analysis. LASSO was used to examine associations between cancer status and AMN metabolites and covariates such as BMI, age, and reproductive factors. LASSO was then repeated to examine associations between AMN metabolites and 10 lifestyle-related variables including smoking, physical activity, alcohol consumption, meat consumption, fruit and vegetable consumption, and supplemental vitamin use. Results were displayed as a network to uncover biological pathways linking lifestyle factors to breast cancer status. After filtering, 851 "active" metabolites out of 1797 metabolomics were retained in 197 correlation AMN clusters. Using LASSO, breast cancer status was associated with 71 "active" metabolites. Several of these metabolites were associated with lifestyle variables including meat consumption, alcohol consumption, and supplemental β-carotene, B12, and folate use. Those metabolites could potentially serve as molecular-level biological intermediaries connecting healthy lifestyle factors to breast cancer, even though direct associations between breast cancer and the investigated lifestyles at the phenotype level are not evident. In particular, DiHODE, a metabolite linked with inflammation, was associated with breast cancer status and connected to β-carotene supplement usage through an AMN. We found several plasma metabolites associated with lifestyle factors and breast cancer status. Future studies investigating the mechanistic role of inflammation in linking supplement usage to breast cancer status are warranted.

Machine learning-enhanced molecular network reveals global exposure to hundreds of unknown PFAS

Unknown forever chemicals like per- and polyfluoroalkyl substances (PFASs) are difficult to identify. Current platforms designed for metabolites and natural products cannot capture the diverse structural characteristics of PFAS. Here, we report an automatic PFAS identification platform (APP-ID) that screens for PFAS in environmental samples using an enhanced molecular network and identifies unknown PFAS structures using machine learning. Our networking algorithm, which enhances characteristic fragment matches, has lower false-positive rate (0.7%) than current algorithms (2.4 to 46%). Our support vector machine model identified unknown PFAS in test set with 58.3% accuracy, surpassing current software. Further, APP-ID detected 733 PFASs in real fluorochemical wastewater, 39 of which are previously unreported in environmental media. Retrospective screening of 126 PFASs against public data repository from 20 countries show PFAS substitutes are prevalent worldwide.

Exploring Prenatal Exposure to Halogenated Compounds and Its Relationship with Birth Outcomes Using Nontarget Analysis

Halogenated organic compounds (HOCs) are a class of contaminants showing high toxicity, low biodegradability, and high bioaccumulation potential, especially chlorinated and brominated HOCs (Cl/Br-HOCs). Knowledge gaps exist on whether novel Cl/Br-HOCs could penetrate the placental barrier and cause adverse birth outcomes. Herein, 326 cord blood samples were collected in a hospital in Jinan, Shandong Province from February 2017 to January 2022, and 44 Cl/Br-HOCs were identified with communicating confidence level above 4 based on a nontarget approach, covering veterinary drugs, pesticides, and their transformation products, pharmaceutical and personal care products, disinfection byproducts, and so on. To our knowledge, the presence of closantel, bromoxynil, 4-hydroxy-2,5,6-trichloroisophthalonitrile, 2,6-dibromo-4-nitrophenol, and related components in cord blood samples was reported for the first time. Both multiple linear regression (MLR) and Bayesian kernel machine regression (BKMR) models were applied to evaluate the relationships of newborn birth outcomes (birth weight, length, and ponderal index) with individual Cl/Br-HOC and Cl/Br-HOCs mixture exposure, respectively. A significantly negative association was observed between pentachlorophenol exposure and newborn birth length, but the significance vanished after the false discovery rate correction. The BKMR analysis showed that Cl/Br-HOCs mixture exposure was significantly associated with reduced newborn birth length, indicating higher risks of fetal growth restriction. Our findings offer an overview of Cl/Br-HOCs exposome during the early life stage and enhance the understanding of its exposure risks.

Unveiling intricate transformation pathways of emerging contaminants during wastewater treatment processes through simplified network analysis

As the key stage for purifying wastewater, elimination of emerging contaminants (ECs) is found to be fairly low in wastewater treatment plants (WWTPs). However, less knowledge is obtained regarding the transformation pathways between various chemical structures of ECs under different treatment processes. This study unveiled the transformation pathways of ECs with different structures in 15 WWTPs distributed across China by simplified network analysis (SNA) we proposed. After treatment, the molecular weight of the whole component of wastewater decreased and the hydrophilicity increased. There are significant differences in the structure of eliminated, consistent and formed pollutants. Amino acids, peptides, and analogues (AAPAs) were detected most frequently and most removable. Benzenoids were refractory. Triazoles were often produced. The high-frequency reactions in different WWTPs were similar, (de)methylation and dehydration occurred most frequently. Different biological treatment processes performed similarly, while some advanced treatment processes differed, such as a significant increase of -13.976 (2HO reaction) paired mass distances (PMDs) in the chlorine alone process. Further, the common structural transformation was uncovered. 4 anti-hypertensive drugs, including irbesartan, valsartan, olmesartan, and losartan, were identified, along with 22 transformation products (TPs) of them. OH2 and H2O PMDs occurred most frequently and in 80.81 % of the parent-transformation product pairs, the intensity of the product was higher than parent in effluents, whose risk should be considered in future assessment activity. Together our results provide a macrography perspective on the transformation processes of ECs in WWTPs. In the future, selectively adopting wastewater treatment technology according to structures is conductive for eliminating recalcitrant ECs in WWTPs.

Deciphering Microbe-Mediated Dissolved Organic Matter Reactome in Wastewater Treatment Plants Using Directed Paired Mass Distance

Understanding the reaction mechanism of dissolved organic matter (DOM) during wastewater biotreatment is crucial for optimal DOM control. Here, we develop a directed paired mass distance (dPMD) method that constructs a molecular network displaying the reaction pathways of DOM. It couples direction inference and PMD analysis to extract the substrate-product relationships and delta masses of potentially paired reactants directly from sequential mass spectrometry data without formula assignment. Using this method, we analyze the influent and effluent samples from the bioprocesses of 12 wastewater treatment plants (WWTPs) and build a dPMD network to characterize the core reactome of DOM. The network shows that the first step of the transformation triggers reaction cascades that diversify the DOM, but the highly overlapped subsequent reaction pathways result in similar effluent DOM compositions across WWTPs despite varied influents. Mass changes exhibit consistent gain/loss preferences (e.g., +3.995 and -16.031) but different occurrences across WWTPs. Combined with genome-centric metatranscriptomics, we reveal the associations among dPMDs, enzymes, and microbes. Most enzymes are involved in oxygenation, (de)hydrogenation, demethylation, and hydration-related reactions but with different target substrates and expressed by various taxa, as exemplified by Proteobacteria, Actinobacteria, and Nitrospirae. Therefore, a functionally diverse community is pivotal for advanced DOM degradation.

Interpretable Machine Learning and Reactomics Assisted Isotopically Labeled FT-ICR-MS for Exploring the Reactivity and Transformation of Natural Organic Matter during Ultraviolet Photolysis

Isotopically labeled FT-ICR-MS combined with multiple post-analyses, including interpretable machine learning (IML) and a paired mass distance (PMD) network, was employed to unravel the reactivity and transformation of natural organic matter (NOM) during ultraviolet (UV) irradiation. FT-ICR-MS analysis was used to assign formulas, which were classified on the basis of their molecular compositions and structural categories. Isotope (deuterium, D) labeling was utilized to unequivocally determine the photochemical products and examine the development of OD radical-mediated NOM transformation. With regard to the reactive molecular formulas, CHOS formulas exhibited the highest reactivity (86.5% of precursors disappeared) followed by CHON (53.4%) and CHO (24.6%) formulas. With regard to structural categories, the degree of reactivity decreased in the following order: tannins > condensed aromatics > lignin/CRAMs. The IML algorithm demonstrated that the crucial features governing the reactivity of formulas were the molecular weight, DBE-O, NOSC, and the presence of heteroatoms (i.e., N and S), suggesting that the large and unsaturated compounds containing S and N are more prone to photodegradation. The reactomics approach using the PMD network further indicated that 11 specific molecular formulas in the CHOS and CHO class served as hubs, implying a higher photoreactivity and participation in a range of transformations. The isotope labeling analyses also found that, among the reactions observed, hydroxylation (i.e., +OD) is dominant for lignin/CRAMs and condensed aromatics, and formulas containing ≤10 D atoms were developed. Overall, this study, by adopting rigorous and interpretable techniques, could provide in-depth insights into the molecular-level dynamics of NOM under UV irradiation.

Exploring the Complexities of Dissolved Organic Matter Photochemistry from the Molecular Level by Using Machine Learning Approaches

Dissolved organic matter (DOM) sustains a substantial part of the organic matter transported seaward, where photochemical reactions significantly affect its transformation and fate. The irradiation experiments can provide valuable information on the photochemical reactivity (photolabile, photoresistant, and photoproduct) of molecules. However, the inconsistency of the fate of irradiated molecules among different experiments curtailed our understanding of the roles the photochemical reactions have played, which cannot be properly addressed by traditional approaches. Here, we conducted irradiation experiments for samples from two large estuaries in China. Molecules that occurred in irradiation experiments were characterized by the Fourier transform ion cyclotron resonance mass spectrometry and assigned probabilistic labels to define their photochemical reactivity. These molecules with probabilistic labels were used to construct a learning database for establishing a suitable machine learning (ML) model. We further applied our well-trained ML model to "un-matched" (i.e., not detected in our irradiation experiments) molecules from five estuaries worldwide, to predict their photochemical reactivity. Results showed that numerous molecules with strong photolability can be captured solely by the ML model. Moreover, comparing DOM photochemical reactivity in five estuaries revealed that the riverine DOM chemistry largely determines their subsequent photochemical transformation. We offer an expandable and renewable approach based on ML to compatibly integrate existing irradiation experiments and shed insight into DOM transformation and degradation processes.

Tracking the transformation pathway of dissolved organic matters (DOMs) in biochars under sludge pyrolysis via reactomics and molecular network analysis

This work examined the transformation pathways of sludge biochar-derived dissolved organic matters (SBC-derived DOMs) under sludge pyrolysis via FT-ICR-MS-based reactomics and molecular network analysis. Lignin/carboxylic-rich alicyclic molecules, proteins/aliphatic, and lipids of SBC-derived DOMs did not contribute equally to the overall pyrolytic reactions. Reactomics suggested that the pyrolysis reactions of SBC-derived DOMs consist of multiple cascade reactions involving the elimination of assemblages of reactive fragments during each pyrolysis reaction region, and the overall pyrolysis process was divided into three stages according to cascade reaction variations. Especially, cascade reactions at 400-500 °C produced potential environmental risk substances of N-containing, carbonyl-containing, and phenolic compounds. Besides, network analysis unraveled the complexity and number of molecular reaction pairs of SBC-derived DOMs decreased with the increase in pyrolytic temperatures. Keystone molecules and pathways results indicated that the pyrolytic temperature of the sludge pyrolysis process should be controlled at temperatures above 500 °C according to the harmful substances generation pattern in reaction products. Overall, the possible transformation pathways of SBC-derived DOMs during sludge pyrolysis treatment were proposed. This study elucidated the underlying mechanisms in generating SBC-derived DOMs and provided theoretical support for process optimization and harmful substances control of sludge pyrolysis.

Deep Characterization of Serum Metabolome Based on the Segment-Optimized Spectral-Stitching Direct-Infusion Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Approach

Direct-infusion Fourier transform ion cyclotron resonance mass spectrometry (DI-FTICR MS) shows great promise for metabolomic analysis due to ultrahigh mass accuracy and resolution. However, most of the DI-FTICR MS approaches focused on high-throughput metabolomics analysis at the expense of sensitivity and resolution and the potential for metabolome characterization has not been fully explored. Here, we proposed a novel deep characterization approach of serum metabolome using a segment-optimized spectral-stitching DI-FTICR MS method integrated with high-confidence and database-independent formula assignments. With varied acquisition parameters for each segment, a highly efficient acquisition was achieved for the whole mass range with sub-ppm mass accuracy. In a pooled human serum sample, thousands of features were assigned with unambiguous formulas and possible candidates based on highly accurate mass measurements. Furthermore, a reaction network was used to select confidently unique formulas from possible candidates, which was constructed by unambiguous formulas and possible candidates connected by the formula differences resulting from biochemical and MS transformation. Compared with full-range and conventional segment acquisition, 8- and 1.2-fold increases in observed features were achieved, respectively. Assignment accuracy was 93-94% for both a standard mixture containing 190 metabolites and a spiked serum sample with the root mean square mass error of 0.15-0.16 ppm. In total, 3534 unequivocal neutral molecular formulas were assigned in the pooled serum sample, 35% of which are contained in the HMDB. This method offers great enhancement in the deep characterization of serum metabolome by DI-FTICR MS.

Comprehensive understanding of DOM reactivity in anaerobic fermentation of persulfate-pretreated sewage sludge via FT-ICR mass spectrometry and reactomics analysis

Understanding the composition and reactivity of dissolved organic matter (DOM) at molecular level is vital for deciphering potential regulators or indicators relating to anaerobic process performance, though it was hardly achieved by traditional analyses. Here, the DOM composition, molecular reactivity and transformation in the enhanced sludge fermentation process were comprehensively elucidated using high-resolution mass spectrometry measurement, and data mining with machine learning and paired mass distance (PMD)-based reactomics. In the fermentation process for dewatered sludge, persulfate (PDS) pretreatment presented its highest performance in improving volatile fatty acids (VFAs) production with the increase from 2,711 mg/L to 3,869 mg/L, whereas its activation in the presence of Fe (as well as the hybrid of Fe and activated carbon) led to the decreased VFAs production performance. In addition to the conventional view of improved decomposition and solubilization of N-containing structures from sludge under the sole PDS pretreatment, the improved VFAs production was associated with the alternation of DOM molecular compositions such as humification generating molecules with high O/C, N/C, S/C and aromatic index (AI_mod). Machine learning was capable of predicting the DOM reactivity classes with 74-76 % accuracy and found that these molecular parameters in addition to nominal oxidation state of carbon (NOSC) were among the most important variables determining the generation or disappearance of bio-resistant molecules in the PDS pretreatment. The constructed PMD-based network suggested that highly connected molecular network with long path length and high diameter was in favor of VFAs production. Especially, -NH related transformation was found to be active under the enhanced fermentation process. Moreover, network topology analysis revealed that CHONS compounds (e.g., C₁₃H₂₇O₈N₁S₁) can be the keystone molecules, suggesting that the presence of sulfur related molecules (e.g., cysteine-like compounds) should be paid more attention as potential regulators or indicators for controlling sludge fermentation performance. This study also proposed the non-targeted DOM molecular analysis and downstream data mining for extending our understanding of DOM transformation at molecular level.

AI/ML-driven advances in untargeted metabolomics and exposomics for biomedical applications

Metabolomics describes a high-throughput approach for measuring a repertoire of metabolites and small molecules in biological samples. One utility of untargeted metabolomics, unbiased global analysis of the metabolome, is to detect key metabolites as contributors to, or readouts of, human health and disease. In this perspective, we discuss how artificial intelligence (AI) and machine learning (ML) have promoted major advances in untargeted metabolomics workflows and facilitated pivotal findings in the areas of disease screening and diagnosis. We contextualize applications of AI and ML to the emerging field of high-resolution mass spectrometry (HRMS) exposomics, which unbiasedly detects endogenous metabolites and exogenous chemicals in human tissue to characterize exposure linked with disease outcomes. We discuss the state of the science and suggest potential opportunities for using AI and ML to improve data quality, rigor, detection, and chemical identification in untargeted metabolomics and exposomics studies.

Molecular Gatekeeper Discovery: Workflow for Linking Multiple Exposure Biomarkers to Metabolomics

The exposome reflects multiple exposures across the life-course that can affect health. Metabolomics can reveal the underlying molecular basis linking exposures to health conditions. Here, we explore the concept and general data analysis framework of "molecular gatekeepers"─key metabolites that link single or multiple exposure biomarkers with correlated clusters of endogenous metabolites─to inform health-relevant biological targets. We performed untargeted metabolomics on plasma from 152 adolescent girls participating in the Growing Up Healthy Study in New York City. We then performed network analysis to link metabolites to exposure biomarkers including five trace elements (Cd, Mn, Pb, Se, and Hg) and five perfluorinated chemicals (PFCs; n-PFOS, Sm-PFOS, n-PFOA, PFHxS, and PFNA). We found 144 molecular gatekeepers and annotated 22 of them. Lysophosphatidylcholine (16:0) and taurodeoxycholate were correlated with both n-PFOA and n-PFOS, suggesting a shared dysregulation from multiple xenobiotic exposures. Sphingomyelin (d18:2/14:0) was significantly associated with age at menarche; yet, no direct association was detected between any exposure biomarkers and age at menarche. Thus, molecular gatekeepers can also discover molecular linkages between exposure biomarkers and health outcomes that may otherwise be obscured by complex interactions in direct measurements.

Novel insight into dissolved organic nitrogen (DON) transformation along wastewater treatment processes with special emphasis on endogenous-source DON

Knowledge of endogenous-source dissolved organic nitrogen (esDON) produced in wastewater treatment processes is critical for evaluating its potential impacts on receiving waters because esDON is a recognized concern, as it causes eutrophication. However, differentiating esDON from influent residual DON in real wastewater is always a challenge. Here, we deciphered esDON information in DON transformation processes along a full-scale wastewater treatment train by combining multiple chemometric tools with ion-mobility separation quadrupole time-of-flight mass spectrometry (IMS-QTOF MS) analyses. In total, DON became more refractory and compact with shorter carbon chains and fewer nitrogen atoms, and esDON composed a nonnegligible fraction that dominated DON transformation and characteristics. New esDON produced in treatment processes constituted a crucial part (>35.5%) of wastewater DON, and its contributions to wastewater DON are augmented along the train. Evidence of molecular conformations further confirmed dominant roles of esDON in DON characteristics. Moreover, esDON participated in 46.7% of core biochemical reaction networks, explaining the importance of esDON in DON transformation. Our study offers a tool to gain esDON characteristics and transformation mechanisms, and highlights the importance to control esDON for alleviating adverse influences from DON in receiving waters.

Reproducible untargeted metabolomics workflow for exhaustive MS2 data acquisition of MS1 features

Unknown features in untargeted metabolomics and non-targeted analysis (NTA) are identified using fragment ions from MS/MS spectra to predict the structures of the unknown compounds. The precursor ion selected for fragmentation is commonly performed using data dependent acquisition (DDA) strategies or following statistical analysis using targeted MS/MS approaches. However, the selected precursor ions from DDA only cover a biased subset of the peaks or features found in full scan data. In addition, different statistical analysis can select different precursor ions for MS/MS analysis, which make the post-hoc validation of ions selected following a secondary analysis impossible for precursor ions selected by the original statistical method. Here we propose an automated, exhaustive, statistical model-free workflow: paired mass distance-dependent analysis (PMDDA), for reproducible untargeted mass spectrometry MS2 fragment ion collection of unknown compounds found in MS1 full scan. Our workflow first removes redundant peaks from MS1 data and then exports a list of precursor ions for pseudo-targeted MS/MS analysis on independent peaks. This workflow provides comprehensive coverage of MS2 collection on unknown compounds found in full scan analysis using a “one peak for one compound” workflow without a priori redundant peak information. We compared pseudo-spectra formation and the number of MS2 spectra linked to MS1 data using the PMDDA workflow to that obtained using CAMERA and RAMclustR algorithms. More annotated compounds, molecular networks, and unique MS/MS spectra were found using PMDDA compared with CAMERA and RAMClustR. In addition, PMDDA can generate a preferred ion list for iterative DDA to enhance coverage of compounds when instruments support such functions. Finally, compounds with signals in both positive and negative modes can be identified by the PMDDA workflow, to further reduce redundancies. The whole workflow is fully reproducible as a docker image xcmsrocker with both the original data and the data processing template.

In Vivo Solid-Phase Microextraction and Applications in Environmental Sciences

Solid-phase microextraction (SPME) is a well-established sample-preparation technique for environmental studies. The application of SPME has extended from the headspace extraction of volatile compounds to the capture of active components in living organisms via the direct immersion of SPME probes into the tissue (in vivo SPME). The development of biocompatible coatings and the availability of different calibration approaches enable the in vivo sampling of exogenous and endogenous compounds from the living plants and animals without the need for tissue collection. In addition, new geometries such as thin-film coatings, needle-trap devices, recession needles, coated tips, and blades have increased the sensitivity and robustness of in vivo sampling. In this paper, we detail the fundamentals of in vivo SPME, including the various extraction modes, coating geometries, calibration methods, and data analysis methods that are commonly employed. We also discuss recent applications of in vivo SPME in environmental studies and in the analysis of pollutants in plant and animal tissues, as well as in human saliva, breath, and skin analysis. As we show, in vivo SPME has tremendous potential for the targeted and untargeted screening of small molecules in living organisms for environmental monitoring applications.

Strategies for structure elucidation of small molecules based on LC–MS/MS data from complex biological samples

LC–MS/MS is a major analytical platform for metabolomics, which has become a recent hotspot in the research fields of life and environmental sciences. By contrast, structure elucidation of small molecules based on LC–MS/MS data remains a major challenge in the chemical and biological interpretation of untargeted metabolomics datasets. In recent years, several strategies for structure elucidation using LC–MS/MS data from complex biological samples have been proposed, these strategies can be simply categorized into two types, one based on structure annotation of mass spectra and for the other on retention time prediction. These strategies have helped many scientists conduct research in metabolite-related fields and are indispensable for the development of future tools. Here, we summarized the characteristics of the current tools and strategies for structure elucidation of small molecules based on LC–MS/MS data, and further discussed the directions and perspectives to improve the power of the tools or strategies for structure elucidation.

Metabolite discovery through global annotation of untargeted metabolomics data

Liquid chromatography-high-resolution mass spectrometry (LC-MS)-based metabolomics aims to identify and quantify all metabolites, but most LC-MS peaks remain unidentified. Here we present a global network optimization approach, NetID, to annotate untargeted LC-MS metabolomics data. The approach aims to generate, for all experimentally observed ion peaks, annotations that match the measured masses, retention times and (when available) tandem mass spectrometry fragmentation patterns. Peaks are connected based on mass differences reflecting adduction, fragmentation, isotopes, or feasible biochemical transformations. Global optimization generates a single network linking most observed ion peaks, enhances peak assignment accuracy, and produces chemically informative peak-peak relationships, including for peaks lacking tandem mass spectrometry spectra. Applying this approach to yeast and mouse data, we identified five previously unrecognized metabolites (thiamine derivatives and N-glucosyl-taurine). Isotope tracer studies indicate active flux through these metabolites. Thus, NetID applies existing metabolomic knowledge and global optimization to substantially improve annotation coverage and accuracy in untargeted metabolomics datasets, facilitating metabolite discovery.

Microbial Roles in Dissolved Organic Matter Transformation in Full-Scale Wastewater Treatment Processes Revealed by Reactomics and Comparative Genomics

Understanding the degradation of dissolved organic matter (DOM) is vital for optimizing DOM control. However, the microbe-mediated DOM transformation during wastewater treatment remains poorly characterized. Here, microbes and DOM along full-scale biotreatment processes were simultaneously characterized using comparative genomics and high-resolution mass spectrometry-based reactomics. Biotreatments significantly increased DOM's aromaticity and unsaturation due to the overproduced lignin and polyphenol analogs. DOM was diversified by over five times in richness, with thousands of nitrogenous and sulfur-containing compounds generated through microbe-mediated oxidoreduction, functional group transfer, and C-N and C-S bond formation. Network analysis demonstrated microbial division of labor in DOM transformation. However, their roles were determined by their functional traits rather than taxa. Specifically, network and module hubs exhibited rapid growth potentials and broad substrate affinities but were deficient in xenobiotics-metabolism-associated genes. They were mainly correlated to liable DOM consumption and its transformation to recalcitrant compounds. In contrast, connectors and peripherals were potential degraders of recalcitrant DOM but slow in growth. They showed specialized associations with fewer DOM molecules and probably fed on metabolites of hub microbes. Thus, developing technologies (e.g., carriers) to selectively enrich peripheral degraders and consequently decouple the liable and recalcitrant DOM transformation processes may advance DOM removal.