Statistical Variable Selection: An Alternative Prioritization Strategy during the Nontarget Analysis of LC-HR-MS Data

qPeaks: A Linear Regression-Based Asymmetric Peak Model for Parameter-Free Automatized Detection and Characterization of Chromatographic Peaks in Non-Target Screening Data

We present qPeaks (quality peaks), a novel, user-parameter-free algorithm for peak detection and peak characterization applicable to chromatographic data. The algorithm is based on a linearizable regression model that analyzes asymmetric peaks and estimates the specific uncertainties associated with the peak regression parameters. The uncertainties of the parameters are used to derive a data quality score DQSpeak, rendering low reliability results more transparent during processing and allowing for the prioritization of generated features. High DQSpeak chromatographic peaks have a lower chance of being classified as false-positive and show higher repeatability over multiple measurements. The high efficiency of the algorithm makes it particularly useful for application within processing routines of nontarget screening through chromatography coupled with high-resolution mass spectrometry. qPeaks is integrated into the qAlgorithms nontarget screening processing toolbox and appends a parameter-free chromatographic peak detection and characterization step to it. With qAlgorithms, now high-resolution mass spectra are centroided using the qCentroids algorithms, centroids are clustered to form extracted ion chromatograms (EICs) with the qBinning algorithm, and chromatographic peaks are found on the generated EICs with qPeaks. However, all tools from qAlgorithms can also be used independently.

Critical Assessment of the Chemical Space Covered by LC-HRMS Non-Targeted Analysis

Non-targeted analysis (NTA) has emerged as a valuable approach for the comprehensive monitoring of chemicals of emerging concern (CECs) in the exposome. The NTA approach can theoretically identify compounds with diverse physicochemical properties and sources. Even though they are generic and have a wide scope, non-targeted analysis methods have been shown to have limitations in terms of their coverage of the chemical space, as the number of identified chemicals in each sample is very low (e.g., ≤5%). Investigating the chemical space that is covered by each NTA assay is crucial for understanding the limitations and challenges associated with the workflow, from the experimental methods to the data acquisition and data processing techniques. In this review, we examined recent NTA studies published between 2017 and 2023 that employed liquid chromatography-high-resolution mass spectrometry. The parameters used in each study were documented, and the reported chemicals at confidence levels 1 and 2 were retrieved. The chosen experimental setups and the quality of the reporting were critically evaluated and discussed. Our findings reveal that only around 2% of the estimated chemical space was covered by the NTA studies investigated for this review. Little to no trend was found between the experimental setup and the observed coverage due to the generic and wide scope of the NTA studies. The limited coverage of the chemical space by the reviewed NTA studies highlights the necessity for a more comprehensive approach in the experimental and data processing setups in order to enable the exploration of a broader range of chemical space, with the ultimate goal of protecting human and environmental health. Recommendations for further exploring a wider range of the chemical space are given.

Predicting RP-LC retention indices of structurally unknown chemicals from mass spectrometry data

Non-target analysis combined with liquid chromatography high resolution mass spectrometry is considered one of the most comprehensive strategies for the detection and identification of known and unknown chemicals in complex samples. However, many compounds remain unidentified due to data complexity and limited number structures in chemical databases. In this work, we have developed and validated a novel machine learning algorithm to predict the retention index (r[Formula: see text]) values for structurally (un)known chemicals based on their measured fragmentation pattern. The developed model, for the first time, enabled the predication of r[Formula: see text] values without the need for the exact structure of the chemicals, with an [Formula: see text] of 0.91 and 0.77 and root mean squared error (RMSE) of 47 and 67 r[Formula: see text] units for the NORMAN ([Formula: see text]) and amide ([Formula: see text]) test sets, respectively. This fragment based model showed comparable accuracy in r[Formula: see text] prediction compared to conventional descriptor-based models that rely on known chemical structure, which obtained an [Formula: see text] of 0.85 with an RMSE of 67.

A Novel 4-Set Venn Diagram Model Based on High-Resolution Mass Spectrometry To Monitor Wastewater Treatment

A 4-set Venn diagram model oriented to high-resolution mass spectrometry (HRMS) data was developed to decipher the fate of dissolved organic matters (DOM) in three-stage continuous wastewater treatment processes. In total, 24 typical wastewater treatment modes conceptualized into a combination of three stages were generalized so that this model can be applied to all common types of actual wastewater treatment processes. As a result, eight kinds of native DOM and seven kinds of wastewater-produced (WW-produced) DOM separately represented by each proper subset of the 4-set Venn diagram could be identified so as to offer a molecular profile of DOM transformation. The 15 proper subsets of the 4-set Venn diagram could then explain how different wastewater treatment units work. Transformation rates of each DOM molecular formula can be estimated as a semiquantitative result. We further discussed the relationship between the transformation rates and proper subsets. As a proof of concept, the 4-set Venn diagram model was successfully applied in a complicated full-scale mature leachate treatment process with nine treatment units. This model can help to overcome the challenging task of data mining when applying HRMS and reduce the workload of data screening in the subsequent structural annotation.

From Centroided to Profile Mode: Machine Learning for Prediction of Peak Width in HRMS Data

Centroiding is one of the major approaches used for size reduction of the data generated by high-resolution mass spectrometry. During centroiding, performed either during acquisition or as a pre-processing step, the mass profiles are represented by a single value (i.e., the centroid). While being effective in reducing the data size, centroiding also reduces the level of information density present in the mass peak profile. Moreover, each step of the centroiding process and their consequences on the final results may not be completely clear. Here, we present Cent2Prof, a package containing two algorithms that enables the conversion of the centroided data to mass peak profile data and vice versa. The centroiding algorithm uses the resolution-based mass peak width parameter as the first guess and self-adjusts to fit the data. In addition to the m/z values, the centroiding algorithm also generates the measured mass peak widths at half-height, which can be used during the feature detection and identification. The mass peak profile prediction algorithm employs a random-forest model for the prediction of mass peak widths, which is consequently used for mass profile reconstruction. The centroiding results were compared to the outputs of the MZmine-implemented centroiding algorithm. Our algorithm resulted in rates of false detection ≤5% while the MZmine algorithm resulted in 30% rate of false positive and 3% rate of false negative. The error in profile prediction was ≤56% independent of the mass, ionization mode, and intensity, which was 6 times more accurate than the resolution-based estimated values.

Non-target trend analysis for the identification of transformation products during ozonation experiments of citalopram and four of its biodegradation products

During ozonation in wastewater treatment plants, ozone reacts with emerging pollutants, which are partially removed through the secondary treatment, as long as, with their biotransformation products, triggering the formation of ozonation transformation products (TPs). Although the transformation of parent compounds (PCs) and their metabolites has been reported in the literature, the probable transformation of biotransformation products has not been investigated so far. This study evaluates the fate of citalopram (CTR) and four of its biotransformation products (DESCTR, CTRAM, CTRAC and CTROXO) during ozonation experiments. A Gaussian curve-based trend analysis was performed for the first time for the automated detection of TPs in ozone concentrations ranging from 0.06 to 12 mg/L. In total 46 ozonation TPs were detected; 7 TPs of CTR, 10 of DESCTR, 9 of CTRAM, 12 of CTRAC and 8 of CTROXO and were structurally elucidated based on their high resolution tandem mass spectra interpretation and tandem mass spectra similarity with the respective PC. Results have demonstrated that the examined compounds follow common transformation pathways in reaction with ozone and that common TPs were formed through the ozonation of different structurally-alike compounds. Moreover, the toxicity of the identified TPs was predicted with an in-house risk assessment program.

Inter-laboratory mass spectrometry dataset based on passive sampling of drinking water for non-target analysis

Non-target analysis (NTA) employing high-resolution mass spectrometry is a commonly applied approach for the detection of novel chemicals of emerging concern in complex environmental samples. NTA typically results in large and information-rich datasets that require computer aided (ideally automated) strategies for their processing and interpretation. Such strategies do however raise the challenge of reproducibility between and within different processing workflows. An effective strategy to mitigate such problems is the implementation of inter-laboratory studies (ILS) with the aim to evaluate different workflows and agree on harmonized/standardized quality control procedures. Here we present the data generated during such an ILS. This study was organized through the Norman Network and included 21 participants from 11 countries. A set of samples based on the passive sampling of drinking water pre and post treatment was shipped to all the participating laboratories for analysis, using one pre-defined method and one locally (i.e. in-house) developed method. The data generated represents a valuable resource (i.e. benchmark) for future developments of algorithms and workflows for NTA experiments.

Measurement(s)	chemical • drinking water
Technology Type(s)	high resolution mass spectrometry • non-target analysis • Interlaboratory
Factor Type(s)	method
Sample Characteristic - Environment	laboratory environment

Machine-accessible metadata file describing the reported data: 10.6084/m9.figshare.15028665

Suspect and non-target screening: the last frontier in environmental analysis

Suspect and non-target screening (SNTS) techniques are arising as new analytical strategies useful to disentangle the environmental occurrence of the thousands of exogenous chemicals present in our ecosystems. The unbiased discovery of the wide number of substances present over environmental analysis needs to find a consensus with powerful technical and computational requirements, as well as with the time-consuming unequivocal identification of discovered analytes. Within these boundaries, the potential applications of SNTS include the studies of environmental pollution in aquatic, atmospheric, solid and biological samples, the assessment of new compounds, transformation products and metabolites, contaminant prioritization, bioremediation or soil/water treatment evaluation, and retrospective data analysis, among many others. In this review, we evaluate the state of the art of SNTS techniques going over the normalized workflow from sampling and sample treatment to instrumental analysis, data processing and a brief review of the more recent applications of SNTS in environmental occurrence and exposure to xenobiotics. The main issues related to harmonization and knowledge gaps are critically evaluated and the challenges of their implementation are assessed in order to ensure a proper use of these promising techniques in the near future.

Environmental effects of offshore produced water discharges: A review focused on the Norwegian continental shelf

Produced water (PW), a large byproduct of offshore oil and gas extraction, is reinjected to formations or discharged to the sea after treatment. The discharges contain dispersed crude oil, polycyclic aromatic hydrocarbons (PAHs), alkylphenols (APs), metals, and many other constituents of environmental relevance. Risk-based regulation, greener offshore chemicals and improved cleaning systems have reduced environmental risks of PW discharges, but PW is still the largest operational source of oil pollution to the sea from the offshore petroleum industry. Monitoring surveys find detectable exposures in caged mussel and fish several km downstream from PW outfalls, but biomarkers indicate only mild acute effects in these sentinels. On the other hand, increased concentrations of DNA adducts are found repeatedly in benthic fish populations, especially in haddock. It is uncertain whether increased adducts could be a long-term effect of sediment contamination due to ongoing PW discharges, or earlier discharges of oil-containing drilling waste. Another concern is uncertainty regarding the possible effect of PW discharges in the sub-Arctic Southern Barents Sea. So far, research suggests that sub-arctic species are largely comparable to temperate species in their sensitivity to PW exposure. Larval deformities and cardiac toxicity in fish early life stages are among the biomarkers and adverse outcome pathways that currently receive much attention in PW effect research. Herein, we summarize the accumulated ecotoxicological knowledge of offshore PW discharges and highlight some key remaining knowledge needs.

Evaluation of Nontarget Long-Term LC-HRMS Time Series Data Using Multivariate Statistical Approaches

The use of liquid chromatography coupled with high-resolution mass spectrometry (LC-HRMS) has steadily increased in many application fields ranging from metabolomics to environmental science. HRMS data are frequently used for nontarget screening (NTS), i.e., the search for compounds that are not previously known and where no reference substances are available. However, the large quantity of data produced by NTS analytical workflows makes data interpretation and time-dependent monitoring of samples very sophisticated and necessitates exploiting chemometric data processing techniques. Consequently, in this study, a prioritization method to handle time series in nontarget data was established. As proof of concept, industrial wastewater was investigated. As routine industrial wastewater analyses monitor the occurrence of a limited number of targeted water contaminants, NTS provides the opportunity to detect also unknown trace organic compounds (TrOCs) that are not in the focus of routine target analysis. The developed prioritization method enables reducing raw data and including identification of prioritized unknown contaminants. To that end, a five-month time series for industrial wastewaters was utilized, analyzed by liquid chromatography-time-of-flight mass spectrometry (LC-qTOF-MS), and evaluated by NTS. Following peak detection, alignment, grouping, and blank subtraction, 3303 features were obtained of wastewater treatment plant (WWTP) influent samples. Subsequently, two complementary ways for exploratory time trend detection and feature prioritization are proposed. Therefore, following a prefiltering step, featurewise principal component analysis (PCA) and groupwise PCA (GPCA) of the matrix (temporal wise) were used to annotate trends of relevant wastewater contaminants. With sparse factorization of data matrices using GPCA, groups of correlated features/mass fragments or adducts were detected, recovered, and prioritized. Similarities and differences in the chemical composition of wastewater samples were observed over time to reveal hidden factors accounting for the structure of the data. The detected features were reduced to 130 relevant time trends related to TrOCs for identification. Exemplarily, as proof of concept, one nontarget pollutant was identified as N-methylpyrrolidone. The developed chemometric strategies of this study are not only suitable for industrial wastewater but also could be efficiently employed for time trend exploration in other scientific fields.

Application of a non-target workflow for the identification of specific contaminants using the example of the Nidda river basin

Non-target screening of water samples from the Nidda river basin in central Germany was conducted with the goal to identify previously unknown chemical contaminants and their emission sources. The focus was on organic, water-borne contaminants which were not typical to municipal wastewater. Grab samples of river water from 13 locations on the Nidda and 15 of its tributaries, in sum 112 samples, were analysed with high resolution LC-QToF-MS/MS. To facilitate the identification of substances, features originating from the same compound such as adducts and isotopologues as well as in-source fragments and species with multiple charge states were registered and grouped by a componentization step utilizing both retention times and peak shapes of the features to combine them in a single component. This led to a reduction of the number of features by an average of 1235 per sample (46%). These grouped features were prioritized if these were detected only in specific tributaries or specific river sections, reducing the number of components by an average of 913 per sample (78%). In addition, grouped features were labelled as typically found in municipal wastewater by combining data from 16 wastewater treatment plants located across Germany and Switzerland and comparing this to components detected in the Nidda basin. These were removed, leading to a further reduction of components by an average of 72 per sample (30%) for an average total reduction of 2536 per sample (93%). Finally, nine compounds, with emission sources in three specific tributaries, were identified, including the textile additive Nylostab S-EED®, which was previously not known to be an environmental contaminant, as well as naturally occurring compounds such as highly toxic microcystins.

Concentration and Distribution of Naphthenic Acids in the Produced Water from Offshore Norwegian North Sea Oilfields

Naphthenic acids (NAs) constitute one of the toxic components of the produced water (PW) from offshore oil platforms discharged into the marine environment. We employed liquid chromatography (LC) coupled to high-resolution mass spectrometry with electrospray ionization (ESI) in negative mode for the comprehensive chemical characterization and quantification of NAs in PW samples from six different Norwegian offshore oil platforms. In total, we detected 55 unique NA isomer groups, out of the 181 screened homologous groups, across all tested samples. The frequency of detected NAs in the samples varied between 14 and 44 isomer groups. Principal component analysis (PCA) indicated a clear distinction of the PW from the tested platforms based on the distribution of NAs in these samples. The averaged total concentration of NAs varied between 6 and 56 mg L^-1, among the tested platforms, whereas the concentrations of the individual NA isomer groups ranged between 0.2 and 44 mg L^-1. Based on both the distribution and the concentration of NAs in the samples, the C₈H₁₄O₂ isomer group appeared to be a reasonable indicator of the presence and the total concentration of NAs in the samples with a Pearson correlation coefficient of 0.89.

Nontarget analysis: A new tool for the evaluation of wastewater processes

Strategies to determine the removal efficiency of micropollutants in wastewater treatment plants (WWTPs) are widely discussed. Especially the evaluation of the potential benefit of further advanced treatment steps such as an additional tertiary treatment based on ozonation or activated carbon have come into focus. Such evaluation strategies are often based on the removal behavior of known micropollutants via target or suspected analysis. The utilization of nontarget analysis is considered to lead to a more comprehensive picture as also unknown or not expected micropollutants are analyzed. Here, the results of an evaluation via target and nontarget analysis were compared for biological treatment (BT) processes of eleven full-scale WWTPs and three different post-treatments (PTs): one sand filter (SF) and two granular activated carbon (GAC) filters. The similarity of the determined removals from target and nontarget analysis of the BTs increased significantly by excluding easily degradable "features" from the nontarget evaluation. A similar ranking of the removal trends for the BTs could also be achieved by comparing this new subset of nontarget features with a set of nine readily to moderately biodegradable micropollutants. This observation suggests that a performance ranking of BTs based either on target or nontarget analysis is plausible. In contrast to the BTs, the evaluation of the three PTs revealed that the difference of feature removal between SF and the two GACs was small, but large for the target analytes with substantially higher removal effciencies for the GACs compared to the SF. In addition to the removal behavior, the nontarget analysis provided further information about the number and quantity of transformation products (TPs) in the effluent from the BTs. For all BTs more than half (55-67%) of the features detected in the effluent were not found in the influent. A comparable proportion of TPs was also detected after GAC and sand filtration due to their microbial activities.

Expanding phytoremediation to the realms of known and unknown organic chemicals of concern

Recent advancements in analytical chemistry and data analyses via high-resolution mass spectrometry (HRMS) are evolving scientific understanding of the potential totality of organic chemical exposure and pollutant risk. This review addresses the importance of HRMS approaches, namely suspect screening and nontarget chemical analyses, to the realm of phytoremediation. These analytical approaches are not without caveats and constraints, but they provide an opportunity to understand in greater totality how plant-based technologies contribute, mitigate, and reduce organic chemical exposure across scales of experimental and system-level studies. These analytical tools can enlighten the complexity and efficacy of plant-contaminant system design and expand our understanding of biogenic and anthropogenic chemicals at work in phytoremediation systems. Advances in data analytics from biological sciences, such as metabolomics, are crucial to HRMS analysis. This review provides an overview of targeted, suspect screening, and nontarget HRMS approaches, summarizes the expanding knowledge of regulated and unregulated organic chemicals in the environment, addresses requisite HRMS instrumentation, analysis cost, uncertainty, and data processing techniques, and offers potential bridges of HRMS analyses to phytoremediation research and application.

The effect of extraction methodology on the recovery and distribution of naphthenic acids of oilfield produced water

Comprehensive chemical characterization of naphthenic acids (NAs) in oilfield produced water is a challenging task due to sample complexity. The recovery of NAs from produced water, and the corresponding distribution of detectable NAs are strongly influenced by sample extraction methodologies. In this study, we evaluated the effect of the extraction method on chemical space (i.e. the total number of chemicals present in a sample), relative recovery, and the distribution of NAs in a produced water sample. Three generic and pre-established extraction methods (i.e. liquid-liquid extraction (Lq), and solid phase extraction using HLB cartridges (HLB), and the combination of ENV+ and C8 (ENV) cartridges) were employed for our evaluation. The ENV method produced the largest number of detected NAs (134 out of 181) whereas the HLB and Lq methods produced 108 and 91 positive detections, respectively, in the tested produced water sample. For the relative recoveries, the ENV performed better than the other two methods. The uni-variate and multi-variate statistical analysis of our results indicated that the ENV and Lq methods explained most of the variance observed in our data. When looking at the distribution of NAs in our sample the ENV method appeared to provide a more complete picture of the chemical diversity of NAs in that sample. Finally, the results are further discussed.

Exploring the Potential of a Global Emerging Contaminant Early Warning Network through the Use of Retrospective Suspect Screening with High-Resolution Mass Spectrometry

A key challenge in the environmental and exposure sciences is to establish experimental evidence of the role of chemical exposure in human and environmental systems. High resolution and accurate tandem mass spectrometry (HRMS) is increasingly being used for the analysis of environmental samples. One lauded benefit of HRMS is the possibility to retrospectively process data for (previously omitted) compounds that has led to the archiving of HRMS data. Archived HRMS data affords the possibility of exploiting historical data to rapidly and effectively establish the temporal and spatial occurrence of newly identified contaminants through retrospective suspect screening. We propose to establish a global emerging contaminant early warning network to rapidly assess the spatial and temporal distribution of contaminants of emerging concern in environmental samples through performing retrospective analysis on HRMS data. The effectiveness of such a network is demonstrated through a pilot study, where eight reference laboratories with available archived HRMS data retrospectively screened data acquired from aqueous environmental samples collected in 14 countries on 3 different continents. The widespread spatial occurrence of several surfactants (e.g., polyethylene glycols ( PEGs ) and C12AEO-PEGs ), transformation products of selected drugs (e.g., gabapentin-lactam, metoprolol-acid, carbamazepine-10-hydroxy, omeprazole-4-hydroxy-sulfide, and 2-benzothiazole-sulfonic-acid), and industrial chemicals (3-nitrobenzenesulfonate and bisphenol-S) was revealed. Obtaining identifications of increased reliability through retrospective suspect screening is challenging, and recommendations for dealing with issues such as broad chromatographic peaks, data acquisition, and sensitivity are provided.

Combining a Deconvolution and a Universal Library Search Algorithm for the Nontarget Analysis of Data-Independent Acquisition Mode Liquid Chromatography-High-Resolution Mass Spectrometry Results

Nontarget analysis is considered one of the most comprehensive tools for the identification of unknown compounds in a complex sample analyzed via liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS). Due to the complexity of the data generated via LC-HRMS, the data-dependent acquisition mode, which produces the MS² spectra of a limited number of the precursor ions, has been one of the most common approaches used during nontarget screening. However, data-independent acquisition mode produces highly complex spectra that require proper deconvolution and library search algorithms. We have developed a deconvolution algorithm and a universal library search algorithm (ULSA) for the analysis of complex spectra generated via data-independent acquisition. These algorithms were validated and tested using both semisynthetic and real environmental data. A total of 6000 randomly selected spectra from MassBank were introduced across the total ion chromatograms of 15 sludge extracts at three levels of background complexity for the validation of the algorithms via semisynthetic data. The deconvolution algorithm successfully extracted more than 60% of the added ions in the analytical signal for 95% of processed spectra (i.e., 3 complexity levels multiplied by 6000 spectra). The ULSA ranked the correct spectra among the top three for more than 95% of cases. We further tested the algorithms with 5 wastewater effluent extracts for 59 artificial unknown analytes (i.e., their presence or absence was confirmed via target analysis). These algorithms did not produce any cases of false identifications while correctly identifying ∼70% of the total inquiries. The implications, capabilities, and the limitations of both algorithms are further discussed.

LC-HRMS Data Processing Strategy for Reliable Sample Comparison Exemplified by the Assessment of Water Treatment Processes

The behavior of micropollutants in water treatment is an important aspect in terms of water quality. Nontarget screening by liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS) offers the opportunity to comprehensively assess water treatment processes by comparing the signal heights of all detectable compounds before and after treatment. Without preselection of known target compounds, all accessible information is used to describe changes across processes and thus serves as a measure for the treatment efficiency. In this study, we introduce a novel LC-HRMS data processing strategy for the reliable classification of signals based on the observed fold changes. An approach for filtering detected features was developed and, after parameter adjustment, validated for its recall and precision. As proof of concept, the fate of 411 target compounds in a 0.1 μg/L standard mix was tracked throughout the data processing stages, where 406 targets were successfully recognized and retained during filtering. Potential pitfalls in signal classification were addressed. We found the recursive peak integration to be a key point for the reliable classification of signal changes across a process. For evaluating the repeatability, a combinatorial approach was conducted to verify the consistency of the final outcome using technical replicates of influent and effluent samples taken from an ozonation process during drinking water treatment. The results showed sufficient repeatability and thus emphasized the applicability of nontarget screening for the assessment of water treatment processes. The developed data processing strategies may be transferred to other research fields where sample comparisons are conducted.