Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect
ACS Publications. Most Trusted. Most Cited. Most Read
My Activity
CONTENT TYPES

KairosMS: A New Solution for the Processing of Hyphenated Ultrahigh Resolution Mass Spectrometry Data

Cite this: Anal. Chem. 2020, 92, 5, 3775–3786
Publication Date (Web):January 28, 2020
https://doi.org/10.1021/acs.analchem.9b05113

Copyright © 2020 American Chemical Society. This publication is licensed under CC-BY.

  • Open Access

Article Views

2502

Altmetric

-

Citations

LEARN ABOUT THESE METRICS
PDF (7 MB)
Supporting Info (2)»

Abstract

The use of hyphenated Fourier transform mass spectrometry (FTMS) methods affords additional information about complex chemical mixtures. Coeluted components can be resolved thanks to the ultrahigh resolving power, which also allows extracted ion chromatograms (EICs) to be used for the observation of isomers. As such data sets can be large and data analyses laborious, improved tools are needed for data analyses and extraction of key information. The typical workflow for this type of data is based upon manually dividing the total ion chromatogram (TIC) into several windows of usually equal retention time, averaging the signal of each window to create a single mass spectrum, extracting a peak list, performing the compositional assignments, visualizing the results, and repeating the process for each window. Through removal of the need to manually divide a data set into many time windows and analyze each one, a time-consuming workflow has been significantly simplified. An environmental sample from the oil sands region of Alberta, Canada, and dissolved organic matter samples from the Suwannee River Fulvic Acid (SRFA) and marine waters (Marine DOM) were used as a test bed for the new method. A complete solution named KairosMS was developed in the R language utilizing the Tidyverse packages and Shiny for the user interface. KairosMS imports raw data from common file types, processes it, and exports a mass list for compositional assignments. KairosMS then incorporates those assignments for analysis and visualization. The present method increases the computational speed while reducing the manual work of the analysis when compared to other current methods. The algorithm subsequently incorporates the assignments into the processed data set, generating a series of interactive plots, EICs for individual components or entire compound classes, and can export raw data or graphics for off-line use. Using the example of petroleum related data, it is then visualized according to heteroatom class, carbon number, double bond equivalents, and retention time. The algorithm also gives the ability to screen for isomeric contributions and to follow homologous series or compound classes, instead of individual components, as a function of time.

Complex mixtures such as petroleum, petroleum related samples, and dissolved organic matter (DOM) are among the most complex and heterogeneous mixtures found in nature. (1,2) The study of these complex mixtures is crucial to improve refining techniques (3−5) and assess their environmental impacts. (6−8) The ultrahigh resolution of Fourier transform mass spectrometry (FTMS) (9−13) has been beneficial to their study. (14−16) More recently, Orbitrap instruments were successfully used for oil-sand related samples (17,18) and DOM. (19) While Orbitrap instruments are more widely available and have lower costs, Fourier transform ion cyclotron mass spectrometry (FTICR MS) offers the highest performance for the study of complex mixtures. (18,19) The use of FT-based mass spectrometers have enabled researchers to observe previously unresolved molecules and gain a deep understanding of their composition. (20) Nevertheless, there still remains unexploited information to be extracted (21,22) and new techniques to observe record number of molecules are regularly developed. (23) Specifically, FTMS techniques allow researchers to determine the masses of thousands of ions with very high accuracy, but do not give any structural information that could be used for distinguishing molecules with the same mass but a different structure (isomers). (24) To address this issue, recent publications have used an online system of gas chromatography (GC), (24) liquid chromatography (LC), (25,26) and trapped ion mobility spectrometry (TIMS) (27) coupled to FTMS instruments.
While chromatography coupled to FTMS instruments is now well-established, the data processing pipeline is struggling to keep pace with the instrumentation advances. The tools developed so far struggle to perform well with the ultrahigh resolution and the complexity of the acquired spectra. Most software currently available such as OpenChrom, (28) MS-Dial, (29) XCMS, (30) MZmine, (31) MetAlign, (32) MathDAMP, (33) and MS Resolver (Pattern Recognition Systems, Bergen, Norway) (34) use m/z binning to create a data matrix with m/z and retention time (tR) as axes and intensity of the peak recorded by the analyzer. This strategy allows researchers to process the data rapidly, match peaks across samples, and perform downstream analysis such as group comparisons, clustering, principal component analysis, etc. However, these methods have not been designed to tackle the challenges posed by the natural variations of the m/z induced by the space-charge effects of the FTMS instruments along with the density of complex samples. Indeed, none offer the ability to recalibrate to tackle the space-charge effects. In addition, they have not been developed to work with low signal-to-noise (S/N) data and the need for denoising. This forces the user to raise the S/N threshold leading to potentially omitting informative peaks, losing some of the benefits of the ultrahigh resolution.
The m/z binning method is based upon the assumption that each molecule m/z is far enough from any other so that sufficiently large bins can be used while avoiding having two different molecules in a single bin. Each bin will then contain an extracted ion chromatogram (EIC) that comprises peaks at given m/z values from scans spanning a retention time range. The peaks within an EIC of a molecular composition are defined by a unique m/z, intensity, and retention time. Analyzing complex mixtures requires the ultrahigh resolution to be able to separate ions present within a very narrow m/z width. (4,35,36) For this reason, the use of large bins is detrimental as there is a high probability of having several EICs to appear in the same bin, losing the benefits of the ultrahigh resolution. The use of small bins also poses great challenges as it increases the risk of excluding parts of the EIC, especially with FT instruments which are subject to space-charge effects resulting in shifting m/z during the experiment. The width of the bins would need to be dynamic as different sample complexities and instruments would influence the viable bin width. For example, analyzing data with an m/z error range of 10 parts per million (ppm) does not pose the same challenges as techniques yielding m/z errors of less than 1 ppm. Similarly, analyzing samples with hundreds of different molecules does not pose the same challenges as analyzing hundreds of thousands of different molecules. The majority of the software cited earlier was developed with the aim of the characterization of other sample types (e.g., biomolecules) using lower resolution instrumentation; hence, they present significantly different characteristics and very different visualization tools. One issue is that they require the conversion of data to the mzXML format, which multiplies the file size; an example of a hyphenated ultrahigh resolution data set in the region of 20–30 GB can become almost 100 GB, leading to increased computational overheads and making successful processing unviable. Other file export methods can sometimes place restrictions on the maximum number of peaks captured, which is not suitable for complex mixture analysis. Furthermore, the software does not allow incorporation of molecular assignments determined using external methods (e.g., in-house algorithms or commercial software) which may be required for a researcher’s workflow, especially for work in specialized fields. As a consequence, the current tools do not scale well for the particularly large and complex data sets often associated with hyphenated ultrahigh resolution experiments.
Presumably, for these reasons, recent papers using hyphenated techniques with FTMS on complex mixtures have not made extensive use of the available software described previously to analyze their data. (24,25,37) MZmine was used by Barrow et al. (24) to obtain a 3D representation of the data but not for the molecular composition analysis. Instead, we can distinguish two methods being employed to analyze hyphenated complex mixture data and another one which has not yet been applied on complex mixtures analyzed by FTMS. The first strategy was employed by Barrow et al. (24) and Patriarca et al. (25) and relied on summing the signal for several time frames to generate peak lists for different time ranges. Those peak lists were analyzed as individual mass spectra and molecular assignments generated for each. The information resulting from those assignments was used to create the plots for each peak list which were then used to follow the molecular evolution of the sample over time. This technique has the advantage of relying on an established workflow to analyze individual spectra but is labor-intensive and induces a loss of temporal resolution, since large time frames are being grouped (e.g. 1 minute windows), meaning that variation within each averaged time frame may be lost.
The second strategy was used by Rüger et al. (37) and relies on an extensive signal processing routine coded for MATLAB which requires long computational time on a server (90–120 min with 20 to 60 GB of RAM) and a MATLAB license. The method has the advantage of not relying on other software and performs the processing starting from raw signal. Strict filtering is applied based on the expected molecular properties (38) and makes use of a modified region of interest (ROI) algorithm to extract the EICs. (39) This ROI method works best in the absence of noise, and because it uses the recorded intensities to detect ROIs, low-intensity regions are unlikely to be detected.
The final method to isolate the EICs has been described using Kalman tracking, (40) although it has not been tested on hyphenated FTMS complex mixtures. This method relies on evaluating the probable position of the next data point using centroid data (discrete m/z with zero line widths) of plasma samples. While the Kalman tracking appears to perform well in the presence of hundreds of EICs, its performance has not yet been demonstrated with millions of data points, which are routinely obtained in centroid mode with complex mixtures.
Previous work demonstrated that peak list data and R can be used to develop new processing methods which improve the quality of the data. (41) So, to address these issues, we developed a method in the open source language R that can run from either a personal computer (PC) or a web application named KairosMS. The name derives from Kairos, an ancient Greek word used to describe time along with Chronos. The method uses developments in signal treatment, peak-picking, and molecular assignments for complex mixtures analyzed with FTMS. It performs, when necessary, an m/z correction to compensate for the space-charge effects in complex mixtures, a quick matching of EICs together, and discarding of noise.
Once the EICs have been matched, a single mass list is generated where each EIC has been reduced to an m/z and an intensity to facilitate assigning peaks to molecules using standard software. The short computing times allow for the trialing and optimization of different settings quickly. Peak assignments are then imported into our workflow, where we developed a suite of tools for data visualization and exploration. Standard figures such as double bond equivalent (DBE) plots, (42−44) class distributions, or van Krevelen diagrams (45,46) for any specified time ranges, down to a scan-by-scan basis, can be generated within seconds. A high level of information is retained using this method, enabling new visualizations to be developed, such as the contribution of specific heteroatom classes and homologous series over each scan during the complete elution process.

Methodology

ARTICLE SECTIONS
Jump To

Sample Preparation

One oil sands process-affected water (OSPW) and two groundwater samples (G1 and G2) were obtained from the Athabasca region along a groundwater flow path. (24) The samples were filtered under vacuum, acidified to pH 4.5, and extracted using Strata-X-A solid phase extraction sorbent (Phenomenex Torrance, CA, United States). The extracts were then methylated using BF3-methanol prior to analysis.
The reference material Suwannee River Fulvic Acid (SRFA) and a marine sample taken at 674 m depth from the North Pacific Ocean at the Natural Energy Laboratory of Hawaii Authority (NELHA) (47,48) used for analysis were acidified (0.01 M HCl), desalted, and concentrated by solid phase extraction. (25) The marine sample is hereafter referred as Marine DOM. The SRFA sample was diluted with ultrapure water and enriched with 0.1% formic acid to a final concentration of 500 ppm in 5% methanol, 94.9% water, and 0.1% formic acid. The freeze-dried SRFA powder was weighed and diluted to 500 ppm with 5% acetonitrile, 94.9% water, and 0.1% formic acid.
A crude pyrolysis bio-oil sample with humidity less than 10 wt % was produced using a mixture of softwoods as original material. (49) The samples were dissolved in acetone at a final concentration of 3 ppm, and 1 μL was injected into a 30 m DB-5 column (0.25 mmID, 0.25 μm).

Instrumentation

KairosMS capabilities and visualization tools were explored for the analysis of six hyphenated data sets acquired with different ultrahigh resolution mass spectrometers. The experimental parameters and instrumentation are briefly described as follows:
GC-APCI-FTICR MS: The OSPW, G1, and G2 samples were analyzed using a 7890A GC instrument (Agilent Technologies, Santa Clara, California, United States) connected to an atmospheric pressure chemical ionization (APCI) source (Bruker Daltonik GmbH, Bremen, Germany) in positive mode which was itself used for the ionization method and connected to a 12 T solariX FTICR mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany) equipped with an Infinity Cell. The temperature was first held at 40 °C and increased at a rate of 20 °C min–1 until a final temperature of 280 °C was reached and held for 20 min. Broadband mass spectra in magnitude mode were acquired, and a single zero fill and Sine-Bell apodization were applied before usage of a Fourier transform.

LC-Orbitrap

The SRFA and Marine DOM were obtained from Patriarca et al. (25) and acquired using an LTQ-Velos-Pro Orbitrap MS (Thermo Scientific, Germany) using an electrospray ionization source (ESI) in negative ion mode. The chromatography was performed using an Agilent PLRP-S poly(styrene/divinylbenzene) column fitted with a precolumn filter (0.5 μm, Supelco Column Saver). After injection, the acetonitrile percentage was increased from 5 to 20% over 2 min and maintained constant for 10 min before being increased to 40% at 13 min and held isocratic until 22 min. Finally, the acetonitrile percentage was increased up to 90% and maintained for 10 min.

GC-APCI-FTICR MS 2xR

The bio-oil mass spectra were acquired using a Bruker 450 GC instrument (Bruker Daltonik GmbH, Bremen, Germany) connected to an APCI ion source in positive mode coupled to a 7 T solariX 2xR FTICR mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany) equipped with a ParaCell. It is worth noting that a 7 T FTICR equipped with a 2 ω detection has performance capabilities comparable to those of a 15 T instrument when operated at similar detection conditions of ω. The oven temperature was initialized at 60 °C and increased at a rate of 6 °C min–1 until a final temperature of 300 °C was reached. The oven was then maintained at 300 °C for 9 min. Broadband mass spectra were acquired, where a single zero fill and Sine-Bell apodization were applied before usage of a Fourier transform. In ftmsControl, a processing was applied which removes 95% of the data points due to the removal of the electronic noise.

Statistical Processing

ARTICLE SECTIONS
Jump To

Overview

The algorithm developed reads a mass list where each peak is defined by its m/z, intensity, and the retention time of the scan in which it was detected (Step 1). The mass list can be refined by cutting beginnings and/or ends of the retention time (Step 2) and/or low intensity peaks (Step 3). A method to detect and separate the EICs is applied (Step 4). After detection of the EICs, it is possible to apply a recalibration method to compensate for any space-charge effect (Step 5). A final EIC matching is performed on the recalibrated mass list (Step 6), and a peak list is created containing only one pair of m/z and intensity for each EIC and used for molecular assignment (Step 7). The assigned peaks’ information is merged with their corresponding EIC, and a table containing all of the EICs (assigned and unassigned) is created and used to create a large series of interactive figures (Step 8). The KairosMS workflow is shown in Figure 1, and further details of each step are described as follows. Throughout, the term “intensity” is used to refer to absolute abundances and “relative intensity” is used to refer to relative abundances.

Figure 1

Figure 1. KairosMS workflow. In Step 1, a mass list containing an identification number, m/z, intensity, and retention time is extracted using the instrument manufacturer’s software. The data processing is then performed during Steps 2–6 in KairosMS. KairosMS generates a single data set for compositional assignments (Step 7). Finally, a single data set or multiple data sets can be opened in KairosMS for interactive visualization and further data analysis (Step 8). All data and graphics are exportable.

Step 1: Extract the Data

FTICR MS data were opened with Bruker DataAnalysis 4.2 (DA) software, and the FTMS peak-picking method was used alongside a script to automatically perform a peak-picking for each mass spectrum recorded over time. The FTMS peak picking algorithm involves the setting of a minimum S/N threshold, thus providing an initial level of noise filtering. For Orbitrap data, the “.raw” data file was converted to mzXML format and read directly into KairosMS. The information was structured into a matrix, where each row corresponds to a peak and columns to respectively retention time, m/z, and intensity. The R code used for the processing made extensive use of the Tidyverse (50) packages and was implemented in a Shiny (51) interface.
The mass list is composed of peaks i = 1, ..., n, n being the number of peaks present. Each peak has an m/z, intensity, and retention time respectively noted Mi, Ii, and Ti, where Tit1, t2, t3, ..., tm and m is the number of scans.

Step 2: Trim the TIC

In chromatography, the beginning of acquisition often corresponds to a baseline signal of noise and can optionally be removed. In consequence, we offer the user the possibility to provide start and end points for the elution and discard any peaks with Ti outside this range. This step will help to discard unnecessary information which will speed up the processing and reduce file sizes. Part of it can be retained if the user wishes to later apply any baseline subtraction.

Step 3: Intensity Filter

Following an initial noise filtering performed on the basis of the S/N in Step 1, a second filtering can be performed on the basis of peak intensities; Zhurov et al. (52) demonstrated that it is possible to discriminate between noise and genuine peaks using the log of intensities. A density plot of the log(Ii) was created to optionally help the user decide on a level of intensity filtering. Peaks with log(Ii) lower than the threshold specified by the user are discarded. Removing parts of the lowest-intensity peaks may be necessary to improve the downstream separation between noise and EICs.

Step 4: EIC Matching Algorithm

The Themis algorithm (41) was adapted to work through an additional dimension to perform the denoising and extract each EIC. As previously described, the m/z consistency was used but this time between scans to isolate the EICs. The difference was that due to intensity variations inherent to the chromatography elution, the intensity parameter had to be excluded from the equation used by Gavard et al. (41) The method performed well in those conditions, but a threshold for the minimum number of consecutive peaks had to be implemented to reduce false positive EICs arising from the combination of too few data points. The user can adjust this parameter by considering the experiment hardware, the sample, and the conditions of acquisition: in some experimental conditions combining GC and simple oil-related samples, an EIC of a low abundance species can be as short as 3 to 5 scans. As described in Themis, a population separation threshold was automatically calculated, but control was given to the user to change this value if deemed necessary.

Step 5: Recalibration

This recalibration method relies on the intra-EIC variations; the first step is to perform a primary matching of the EICs as described in Step 4. The m/z was reconverted into Hz using an adaptation of eq 1 (53) taken from Barry et al. (54)
(1)
The frequency was calculated using eq 2, derived from eq 1, using instrument-specific values A and B provided by the user. For a Bruker FTICR MS data set, the A and B values are respectively named ML1 and ML2 and can be found within the method file within each data directory.
(2)
Let be the highest frequency in peak in the jth EIC. For each peak i within EIC j, we define the frequency shift . For each time t ∈ {t1, ..., tm} we compute the mean frequency shift S(t)
(3)
A LOESS model (55) was fitted to the relation between the scan total intensity and the mean frequency shift S(t). We denoted the new S(t) predicted using the LOESS model . The modeling helps to ensure that if some scans were too shifted to be picked up, they would still get the appropriate corrections in regards to their expected shift because of the total intensity of the scan.
We search for any which is . Any peak i within the previously identified is corrected by calculating
(4)
The remaining peaks have . All of the Mi values are subsequently updated using (1) with the corrected frequency . Using the updated Mi, the EIC matching described in Step 4 is performed again using the same parameters. We now calculate similarly to S(t), but based on instead of . The final frequencies are obtained by calculating and the corresponding m/z was calculated. The updated m/z are then used in Step 6.
If the A and B coefficients from eq 1 are not available, an equivalent procedure can be applied without going into the frequency domain by calculating the shift in ppm and applying the correction directly on the m/z. The described recalibration method, however, does not rely on prior knowledge of the true m/z of one or more peaks.

Step 6: Processing

If recalibration was performed in step 5, the density plot observed previously might have changed. In consequence, KairosMS offers the user the opportunity to change the settings used for the pairing (Step 4). Once the pairing described in Step 4 has been performed, the user has an overview of the number of isolated EICs. The number of EICs which had two or more peaks from the same retention time and went through an additional refinement is also presented, and a high value will indicate that the previous settings needs to be tightened.

Step 7: Molecular Assignment

Once the EICs were isolated, a mass list was created using the sum of intensities within each EIC and the mean m/z of each EIC. This standard mass list can be read into third party molecular assignment software (e.g. Composer, PetroOrg, in-house scripts, etc.), depending on the type of sample. The assignments for each EIC were merged with the data for the peaks within the EIC and stored as an R data table object called a tibble. (56) The columns containing the information from the assignment remained empty for peaks within the unassigned EICs. No information is therefore removed from the original peak list, and the assignments could be redone later if necessary.

Step 8: Data Analysis Tools

Currently, KairosMS produces a suite of visualization tools commonly used in petroleomics due to the need to visualize complex mixture data. These include displaying the DBE vs carbon number, percentage intensity contribution of the different classes, evolution of the intensity over time for each class, homologous series of molecules, van Krevelen diagrams, breakdown of the contribution of each atom present in the sample, area under the curve (quantification) for heteroatom classes to molecules, and principal component analysis. Note that in addition to using data from hyphenated mass spectrometry experiments, direct infusion data can also be analyzed, visualized, and compared. Steps 1 to 7 are performed on each sample individually, leading to the characterization of the majority of EICs. The comparison between samples is then based on the use of molecular assignments, which are determined, merged with the EICs, and compared during steps 7 and 8. Comparisons between several hyphenated samples that have been analyzed using KairosMS are also provided in Figures 5 and 6. KairosMS was coded in R and implemented into a Shiny interactive interface, allowing the user to see the plots as the analysis proceeds through the process and adjust the parameters accordingly. KairosMS can be run either locally on a personal computer or online through a server or a local network.

Results and Discussion

ARTICLE SECTIONS
Jump To

Data Processing

A screenshot presenting the interface of KairosMS is depicted in Figure S1. A detailed description of the processing steps in KairosMS for the OSPW sample are detailed below.
To process the OSPW data set, the TIC between 0 and 9 min was trimmed, and the intensities below 73 130 were filtered out. This led to a reduction from 2 963 880 to 2 086 325 data points (29.61% removed). The denoising and EIC extraction method from Step 3 was applied to enable the recalibration method described in Step 4.
The optional recalibration step allowed us to correct for the space-charge effects without using any prior knowledge about the sample. The calibration was performed using matching conditions of 20 consecutive peaks. Figure 2 shows the evolution of the average frequency shift for each retention time. One can notice the similarities with the profile of the original f shift compared to the TIC in Figure 4. The recalibration performed attenuates the space-charge effects, as the calculated frequency shifts after recalibration shown in black in Figure 2. Before recalibration, the RMS error, which was calculated using the difference between the assigned m/z and the experimental m/z of each peak of each EIC, was 2 ppm. After recalibration, the RMS error decreased to 0.4 ppm (Figure 3).

Figure 2

Figure 2. Average frequency shift in Hz within EICs for each retention time before recalibration (red) and after (black) for the OSPW sample.

Figure 3

Figure 3. Histogram of the mass error in ppm of each peak for each EIC with the assigned m/z.

Using a minimum EIC length of 20 scans, 1 473 579 of the 2 086 325 peaks were kept (29.37% removed). The complete process typically takes tens of seconds to perform and could be improved further with the use of parallel computing. A total of 6540 distinct EICs were isolated with this process. We compared the TIC before and after processing to make sure that no critical features were discarded and that the shape of the TIC had been preserved. Figure 4 shows there were no noticeable differences before and after processing, ensuring that all major peaks had been preserved and matched to an EIC.

Figure 4

Figure 4. Comparison of TICs before (top line, black) and after (bottom line, red) processing.

Because current petroleomics software for molecular assignments (Composer, PetroOrg) were designed to assign molecular composition on a single spectrum, each EIC was summarized by a single m/z and intensity. The average m/z of each EIC was used, while the intensities of the EICs were individually summed. The resulting spectrum for the OSPW sample is presented in Figure S2.
Once the assignments were performed using Composer, the molecular composition for each peak assigned was reattributed to their respective EIC, and a special data frame format (tibble) was used to store all of the information. Tibbles make subsetting easier than traditional data-frames and allow the mixing of several types of data (e.g. characters, numeric, factor). Subsetting is crucial at a later stage as we analyze the data and explore specific classes, retention times, m/z. The unassigned EICs were preserved so that the processed data, saved as.csv, could be reprocessed in the future.
The data processing described above for OSPW can be performed in about 5 min. The raw GC data obtained for the OSPW sample is about 24 GB in size and is reduced after steps 1–7 in KairosMS to a final data set size of 163.4 MB (0.68% of the original data set size). Similarly, the processing steps were performed to extract the EICs of the remaining data sets: G1, G2, SRFA, Marine DOM and the bio-oil. A final data set size in the range of 93–117 MB was obtained for each sample.

Data Analysis Tools

These data sets contain the assigned and unassigned EICs and are used for further data analysis in the final step in KairosMS. KairosMS enables the user to interpret hyphenated data and to study the molecular composition more efficiently than before; a list of some of the visualization tools available are listed below:
  • TIC and mass spectra visualization at a desired retention time.

  • Interactive DBE versus carbon number plots: individual or multiple classes with the possibility of the extraction of the EICs of each data point of the DBE plot. The DBE plots can be visualized in different retention time frames as desired for the user.

  • Interactive class distribution: heteroatomic classes can be visualized in different retention time ranges. The heteroatomic class distribution can be fixed or automatically updated for desired retention times.

  • Mass spectra, DBE, and carbon number distribution: the sum intensities versus carbon number and DBE can be plotted by individual heteroatomic class.

  • Interactive van Krevelen diagrams: van Krevelen diagrams can be plotted as a function of time, and the heteroatomic classes can be selected for the user. The user can also extract the EICs in each data point in the van Krevelen diagram.

  • Interactive EIC visualization: a total EIC per heteroatomic class or homologous series can be extracted. The total area under the curve (AUC) per class or homologous series is calculated by KairosMS and can be exported in a .csv file. Additionally, the user can define either: an m/z, assigned molecular formula, or a custom molecular formula to be visualized from the data set. The EICs can be visualized for a particular or multiple heteroatomic classes within a certain ranges of carbon number and DBEs as desired by the user.

  • Filters: the data visualization includes data filters either by class, DBE, isotopic compositions, retention time, adduct type, or by sample.

  • Plot settings: all plots can be individually exported in .png, .pdf, .eps, or .tiff format. The figures can be faceted by the sample-identifying name. Additionally, the data point size in DBE plots can be changed or plotted in a log10 scale. Class distribution figures can be plotted in stack bars or bar charts, and the coordinates can be flipped. The data in EICs can be visualized and exported with a defined dot size, and the EICs can be exported in a defined retention time domain. Alternatively, the figures can be generated in an external software by downloading the data from KairosMS in a .csv file. The figure format of the graphic can be changed by using different color schemes, different graphic resolution, figure size, and data legend size.

These capabilities are shown in the Movie provided in the Supporting Information.

Applications: Data Analysis Visualization

ARTICLE SECTIONS
Jump To

GC-FTICR MS for the Analysis of OSPW and Groundwater

As it often becomes necessary to compare data sets, (24,25) we extended the capabilities of KairosMS so that one can compare several samples after they have been processed due to the level of detail of information retained during the processing. No limits have been set to the number of data sets to compare, but using more files requires longer computation times and more memory. The previous OSPW sample was compared to two groundwater samples (G1 and G2). The first step was to use the class contribution function to observe the key differences between the samples (see Figure 5A). As shown in Figure 5A, the class distribution of all samples is shifted toward higher oxygen-containing species at higher retention time. Additionally, it is noticeable that the oxygen content of the OSPW sample is comparatively different to the groundwater sample. For instance, at low retention time, the relative abundance of the oxygenated classes is higher in the OSPW in comparison to the groundwater samples, and lower relative oxygen content species elute from the GC column at higher retention time in comparison with the groundwater samples.

Figure 5

Figure 5. (A) Heteroatom class distribution for retention times 12.10–12 min and 31.10–32 min. (B) EICs of selected homologous series from the O2[H] class. (C) EICs of selected molecular composition for the OSPW and two groundwater samples (G1 and G2).

Using the observations made in the class distribution in Figure 5A, the O2[H] class was selected for further analysis and further broken down to observe independently each homologous series, where it can be seen that the predominant DBEs were 2.5, 3.5, and 4.5 (Figure 5B). An enlarged version of Figure 5B with the complete retention time is available in the SI as Figure S5. The AUC of the homologous series shows an increased contribution at higher retention times as the DBE increases. Thus, species with higher DBE have increased boiling point and therefore elute from the GC column at higher retention time.
Because we observed differences in the O2 [H] class, we explored it further and observed its evolution with a scan by scan resolution. In Figure S3, we notice that G1 and G2 display the exact same elution profile, while the OSPW has remarkable differences. In Figure S4, the isotopes for this particular molecular composition were also observed.
KairosMS enables a rapid screening of EICs for each molecular composition identified by giving the capability to display all of the EICs matching specific features such as a specific m/z or m/z range, belong to a specific heteroatom class, DBE range or range of carbon numbers. Once differences are identified for a specific heteroatom class, it becomes possible to display the EIC for individual molecular composition within that class.
After screening EICs corresponding to the O2[H] class, noticeable differences were found between the groundwater and OSPW EICs. For instance, the EICs corresponding to C13H18O2[H] and C14H16O2[H] in Figure 5C show a distribution of peaks at low retention time in the groundwater samples that were not detected in the OSPW. This indicates the presence of different isomers and different ratios between isomers which can then be isolated and further investigated.

LC-Orbitrap: Dissolved Organic Matter

KairosMS capabilities to handle dissolved organic matter analyzed in an online LC-Orbitrap system were tested using Marine DOM and SRFA samples. Each data set was first processed using KairosMS and the results exported as .csv files. Finally, both files were loaded into KairosMS for data exploration and comparison. The TIC and the mass spectra at two different retention times of SRFA and the marine samples can be seen in Figure 6A. As shown in this figure, the compositions with higher m/z elute at higher retention time in both samples. Figure 6B and C shows the differences in molecular composition between the two samples for the complete run. The class distribution shown in Figure 6C can be modified within KairosMS to show any retention time range to highlight the difference of composition at any stage of the acquisition. In contrast to the GC experiment, the components eluting from this LC column at higher retention time correspond to species with lower oxygen containing species. A major new capability enabled by this work is the ability to track specific heteroatom classes across every scan acquired. For instance, Figure 6B shows the heteroatom class O4[H] and O10[H] intensity across the complete retention time, allowing the user to immediately highlight the differences between the two samples. In similarity with the EICs by homologous series, the AUC of the classes is calculated in KairosMS.

Figure 6

Figure 6. (A) Total ion chromatogram (TIC) with the mass spectra for retention times 10.72 min and 20.57 min. (B) EICs for the heteroatom classes O4[H] and O7[H] with associated AUC. (C) Heteroatom class distribution for retention times 10.72–10.80 min and 20.57–20.66 min for the SRFA and Marine DOM samples.

To further explore the differences between the two samples, it is also possible to plot side by side or to overlap the van Krevelen diagrams for each sample (Figure S8). As in Figure S6, this plot can be refined to observe any specific retention time range. The figure demonstrates significant differences between the two samples, especially in the region below H/C of 1.

GC FTICR MS 2xR

A bio-oil sample was analyzed using a solariX 2xR FTICR MS. The 2ω detection from this instrument allows to either operate the instrument at twice the speed for the same resolving power or to double the resolving power if the speed is kept the same as compared to conventional 2ω instruments. For gas chromatography, acquisition at twice the speed is particularly useful for samples presenting a large number of isomers.
The van Krevelen diagram and the DBE plots of the classes O2[H], O3[H], O4[H], and O5[H] obtained within the total retention time in the GC column (see Figure 7A and 7B–C, respectively). The EICs of the compositions in both type of plots can be visualized by using KairosMS. In contrast to DBE plots, where each data point corresponds to a single composition, van Krevelen data points can correspond to multiple EICs of different compositions with the same H/C and O/C values (e.g., C8H12O4[H] and C12H18O6[H]. In comparison with the previous samples, the bio-oil has a remarkable number of potential isomers. For instance, the composition C8H12O4[H] shows the presence of at least 71 potential isomers. It is important to note that species with higher carbon number, higher DBE, and higher oxygen content eluted from the column at higher retention times.

Figure 7

Figure 7. (A) van Krevelen diagram with EICs displayed for two different points on the plot. (B) DBE vs carbon number of the O2[H] and O5[H] heteroatom classes with associated EICs for two data points on the plot. (C) DBE vs carbon number of the O3[H] and O4[H] heteroatom classes with associated EICs for two data points. Note that each data point in a van Krevelen diagram represents the sum of multiple molecular formulas, while each data point in a DBE vs carbon number plot is a single molecular formula.

The EIC shown in Figure S9 shows the presence of at least 35 potential isomers. To assess KairosMS capabilities to isolate such challenging EIC, we’ve overlapped the EIC obtained using DA and KairosMS and it showed a complete overlap between the two with only minor differences within the noise baseline due to the necessary intensity threshold resulting from the peak picking.

Other Applications

KairosMS also provides the ability to search for any specific EIC of an identified molecule. The user can search using m/z or molecular composition but can also display all of the EICs with specific features such as heteroatom class, carbon number, and DBE. This allows a researcher to quickly determine differences between elution profiles at the molecular level.
Finally, by using the intensity information on all of the assigned EICs, the calculation of the elemental contribution within each sample can be swiftly obtained, and the percentages for each elements can be calculated as depicted in Figure S7.
KairosMS was also used to process peptide digest data, analyzed by LC-FTICR MS, as pictured in Figure S10. Even without providing molecular assignments, in this particular case, KairosMS was able to quickly and accurately calculate the area under the curve for all isolated EICs, providing useful quantification data.

Conclusion

ARTICLE SECTIONS
Jump To

Using hyphenated Fourier transform mass spectrometry, additional separation methods such as chromatography provide further insights about complex chemical mixtures, especially for the observation of isomers. The data analysis for such experiments previously relied on long and laborious manual work. The typical workflow for this type of data is based upon manually merging mass spectra over a series of retention time ranges, extraction of each peak list, assigning compositions, visualizing the results, and repeating the process for each retention time range. KairosMS addresses those issues by removing the need to manually divide a data set into many time windows and analyze each one, while also preserving the time resolution. Data are first extracted as a mass list to reduce computing time, relying on peak picking and centroid detection of existing algorithms. KairosMS then processes the data, can attenuate space-charge effects, and existing peak assignments methods are reused. The recalibration is currently best suited for FTICR MS instruments analyzing complex mixtures and needs to be optimized for data sets such as biomolecules. The method could be further improved by implementing prior knowledge about one or more peaks. KairosMS demonstrated its abilities over a wide range of samples (petroleum, environmental, dissolved organic matter results, and biomolecules) from different types of analyzers and types of chromatography, turning hyphenated ultrahigh resolution mass spectrometry into a regular tool for those samples. The capability to quickly visualize EICs from any class, homologous series, m/z, or molecular assignment helps the user to fully exploit the information enabled by the chromatography, especially in the presence of multiple isomers, paving the way to shift from relying solely on molecular compositions to understanding the structure of the molecules in complex mixtures. Among the features for comparing many complex data sets, KairosMS also includes options for hierarchical clustering and principal component analysis. It should be noted KairosMS can be used for data analysis, visualization, and sample comparison for direct infusion data in addition to hyphenated data sets. Using the same simple file format for the processed data, we were able to simultaneously browse and compare samples, saving the user from repetitive tasks. It is expected that with such information made available, additional new visualization methods can be developed to help tackle the challenges posed by the volume of data.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.9b05113.

  • Figure S1: Screenshot presenting KairosMS interface; Figure S2: Mass spectrum created using the EICs extracted to be used for molecular assignments; Figure S3: Comparison of the elution of the O2[H] class contribution between an OSPW and two groundwater samples using a scan by scan resolution; Figure S4: EICs for the monoisotopic form and isotopologues of C16H26O2 [H] for the G1, G2, and OSPW samples; Figure S5: Elution of DBE series (homologous series) comprising the O2[H] class; Figure S6: Percentage of contribution to the total signal for all of the classes identified in the SRFA and Marine DOM samples; Figure S7: Elemental contributions for the samples SRFA and Marine DOM based on all of the assigned EICs; Figure S8: van Krevelen diagram of the H/C ratio vs O/C ratio for the Marine DOM and SRFA samples; Figure S9: Comparison of the EIC of the same molecular assignment as seen in DA and KairosMS after peak picking at S/N 1; Figure S10: EIC from a peptide digest of ubiquitin analyzed by LC-FTICR MS (PDF)

  • Video showing interactive data visualization using KairosMS (MP4)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To

  • Corresponding Author
  • Authors
    • Remy Gavard - MAS CDT, University of Warwick, Coventry CV4 7AL, United KingdomOrcidhttp://orcid.org/0000-0001-5899-3058
    • Hugh E. Jones - Department of Chemistry, University of Warwick, Coventry CV4 7AL, United Kingdom
    • Diana Catalina Palacio Lozano - Department of Chemistry, University of Warwick, Coventry CV4 7AL, United KingdomOrcidhttp://orcid.org/0000-0001-5315-5792
    • Mary J. Thomas - MAS CDT, University of Warwick, Coventry CV4 7AL, United KingdomOrcidhttp://orcid.org/0000-0002-6744-5413
    • David Rossell - Department of Statistics, University of Warwick, Coventry CV4 7AL, United KingdomDepartment of Economics & Business, Universitat Pompeu Fabra, Barcelona 08005, Spain
    • Simon E. F. Spencer - Department of Statistics, University of Warwick, Coventry CV4 7AL, United Kingdom
  • Notes
    The authors declare no competing financial interest.

Acknowledgments

ARTICLE SECTIONS
Jump To

R.G., H.E.J., and M.J.T. thank EPSRC for a PhD studentship through the EPSRC Centre for Doctoral Training in Molecular Analytical Science, grant number EP/L015307/1. M.P.B. and R.G. thank Warwick Ventures for the HEIF Impact Fund. D.C.P.L. is thankful for the Newton Fund award (reference number 275910721), Research Agreement No. 5211770 UIS-ICP, and COLCIENCIAS (project No. FP44842-039-2015) for funding. M.P.B. thanks John V. Headley, Kerry M. Peru, and Jason Ahad for the groundwater and OSPW samples. R.G. thanks Jeffrey Hawkes and Claudia Patriarca for the SRFA and Marine DOM data. D.R. was partially supported by Ramon y Cajal Fellowship RYC-2015-18544 from Ministerio de Economia y Competitividad (Government of Spain), Ayudas Investigación Cientifica Big Data (Fundación BBVA), and Programa Estatal I+D+i I+D+i PGC2018-101643–B-I00 (Government of Spain). M.P.B. and R.G. also thank David Stranz (Sierra Analytics) for his valuable contributions.

References

ARTICLE SECTIONS
Jump To

This article references 56 other publications.

  1. 1
    Hertkorn, N.; Ruecker, C.; Meringer, M.; Gugisch, R.; Frommberger, M.; Perdue, E. M.; Witt, M.; Schmitt-Kopplin, P. High-precision frequency measurements: Indispensable tools at the core of the molecular-level analysis of complex systems. Anal. Bioanal. Chem. 2007, 389, 13111327,  DOI: 10.1007/s00216-007-1577-4
  2. 2
    Barrow, M. P. Petroleomics: study of the old and the new. Biofuels 2010, 1, 651655,  DOI: 10.4155/bfs.10.55
  3. 3
    Mullins, O. C.; Sheu, E. Y.; Hammami, A.; Marshall, A. G.; Eds. Asphaltenes, Heavy Oils, and Petroleomics; Springer: New York, NY, 2007.
  4. 4
    Marshall, A. G.; Rodgers, R. P. Petroleomics: Chemistry of the underworld. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 1809018095,  DOI: 10.1073/pnas.0805069105
  5. 5
    Hsu, C. S.; Hendrickson, C. L.; Rodgers, R. P.; McKenna, A. M.; Marshall, A. G. Petroleomics: Advanced molecular probe for petroleum heavy ends. J. Mass Spectrom. 2011, 46, 337343,  DOI: 10.1002/jms.1893
  6. 6
    Kellerman, A. M.; Dittmar, T.; Kothawala, D. N.; Tranvik, L. J. Chemodiversity of dissolved organic matter in lakes driven by climate and hydrology. Nat. Commun. 2014, 5. DOI: 10.1038/ncomms4804
  7. 7
    Stubbins, A.; Dittmar, T. Illuminating the deep: Molecular signatures of photochemical alteration of dissolved organic matter from North Atlantic Deep Water. Mar. Chem. 2015, 177, 318324,  DOI: 10.1016/j.marchem.2015.06.020
  8. 8
    Headley, J. V.; Peru, K. M.; Barrow, M. P. Advances in mass spectrometric characterization of naphthenic acids fraction compounds in oil sands environmental samples and crude oil - A review. Mass Spectrom. Rev. 2016, 35, 311328,  DOI: 10.1002/mas.21472
  9. 9
    Comisarow, M. B.; Marshall, A. G. Fourier transform ion cyclotron resonance spectroscopy. Chem. Phys. Lett. 1974, 25, 282283,  DOI: 10.1016/0009-2614(74)89137-2
  10. 10
    Comisarow, M. B.; Marshall, A. G. Selective-phase ion cyclotron resonance spectroscopy. Can. J. Chem. 1974, 52, 19971999,  DOI: 10.1139/v74-288
  11. 11
    Comisarow, M. B.; Marshall, A. G. Frequency-sweep fourier transform ion cyclotron resonance spectroscopy. Chem. Phys. Lett. 1974, 26, 489490,  DOI: 10.1016/0009-2614(74)80397-0
  12. 12
    Amster, I. J. Fourier transform mass spectrometry. J. Mass Spectrom. 1996, 31, 13251337,  DOI: 10.1002/(SICI)1096-9888(199612)31:12<1325::AID-JMS453>3.0.CO;2-W
  13. 13
    Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Fourier Transform Ion Cyclotron Resonance Mass Spectromeyry: A Primer. Mass Spectrom. Rev. 1998, 17, 135,  DOI: 10.1002/(SICI)1098-2787(1998)17:1<1::AID-MAS1>3.0.CO;2-K
  14. 14
    Barrow, M. P.; Burkitt, W. I.; Derrick, P. J. Principles of Fourier transform ion cyclotron resonance mass spectrometry and its application in structural biology. Analyst 2005, 130, 18,  DOI: 10.1039/b403880k
  15. 15
    Koch, B. P.; Ludwichowski, K. U.; Kattner, G.; Dittmar, T.; Witt, M. Advanced characterization of marine dissolved organic matter by combining reversed-phase liquid chromatography and FT-ICR-MS. Mar. Chem. 2008, 111, 233241,  DOI: 10.1016/j.marchem.2008.05.008
  16. 16
    Dittmar, T.; Stubbins, A. Treatise Geochemistry, 2nd ed.; Elsevier Ltd., 2013; Vol. 12; pp 125156.
  17. 17
    Headley, J. V.; Peru, K. M.; Janfada, A.; Fahlman, B.; Gu, C.; Hassan, S. Characterization of oil sands acids in plant tissue using Orbitrap ultra-high resolution mass spectrometry with electrospray ionization. Rapid Commun. Mass Spectrom. 2011, 25, 459462,  DOI: 10.1002/rcm.4877
  18. 18
    Zhurov, K. O.; Kozhinov, A. N.; Tsybin, Y. O. Evaluation of high-field orbitrap fourier transform mass spectrometer for petroleomics. Energy Fuels 2013, 27, 29742983,  DOI: 10.1021/ef400203g
  19. 19
    Hawkes, J. A.; Dittmar, T.; Patriarca, C.; Tranvik, L.; Bergquist, J. Evaluation of the Orbitrap Mass Spectrometer for the Molecular Fingerprinting Analysis of Natural Dissolved Organic Matter. Anal. Chem. 2016, 88, 76987704,  DOI: 10.1021/acs.analchem.6b01624
  20. 20
    Cho, E.; Witt, M.; Hur, M.; Jung, M. J.; Kim, S. Application of FT-ICR MS Equipped with Quadrupole Detection for Analysis of Crude Oil. Anal. Chem. 2017, 89, 1210112107,  DOI: 10.1021/acs.analchem.7b02644
  21. 21
    Krajewski, L. C.; Rodgers, R. P.; Marshall, A. G. Assigned Chemical Formulas from an Atmospheric Pressure Photoionization 9.4 T Fourier Transform Positive Ion Cyclotron Resonance Mass Spectrum. Anal. Chem. 2017, 89, 11318,  DOI: 10.1021/acs.analchem.7b02004
  22. 22
    Smith, D. F.; Podgorski, D. C.; Rodgers, R. P.; Blakney, G. T.; Hendrickson, C. L. 21 T FT-ICR Mass Spectrometer for Ultrahigh-Resolution Analysis of Complex Organic Mixtures. Anal. Chem. 2018, 90, 20412047,  DOI: 10.1021/acs.analchem.7b04159
  23. 23
    Palacio Lozano, D. C.; Gavard, R.; Arenas-Diaz, J. P.; Thomas, M. J.; Stranz, D. D.; Mejía-Ospino, E.; Guzman, A.; Spencer, S. E. F.; Rossell, D.; Barrow, M. P. Pushing the analytical limits: new insights into complex mixtures using mass spectra segments of constant ultrahigh resolving power. Chem. Sci. 2019, 10, 69666978,  DOI: 10.1039/C9SC02903F
  24. 24
    Barrow, M. P.; Peru, K. M.; Headley, J. V. An added dimension: GC atmospheric pressure chemical ionization FTICR MS and the athabasca oil sands. Anal. Chem. 2014, 86, 82818288,  DOI: 10.1021/ac501710y
  25. 25
    Patriarca, C.; Bergquist, J.; Sjoberg, P. J. R.; Tranvik, L.; Hawkes, J. A. Online HPLC-ESI-HRMS Method for the Analysis and Comparison of Different Dissolved Organic Matter Samples. Environ. Sci. Technol. 2018, 52, 20912099, PMID: 29241333.  DOI: 10.1021/acs.est.7b04508
  26. 26
    Kim, S.; Kim, D.; Kim, S.; Son, S.; Jung, M. J. Application of Online Liquid Chromatography 7 T FT-ICR Mass Spectrometer Equipped with Quadrupolar Detection for Analysis of Natural Organic Matter. Anal. Chem. 2019, 91, 76907697,  DOI: 10.1021/acs.analchem.9b00689
  27. 27
    Benigni, P.; Thompson, C. J.; Ridgeway, M. E.; Park, M. A.; Fernandez-Lima, F. Targeted high-resolution ion mobility separation coupled to ultrahigh-resolution mass spectrometry of endocrine disruptors in complex mixtures. Anal. Chem. 2015, 87, 43214325,  DOI: 10.1021/ac504866v
  28. 28
    Wenig, P.; Odermatt, J. OpenChrom: a cross-platform open source software for the mass spectrometric analysis of chromatographic data. BMC Bioinf. 2010, 11. DOI: 10.1186/1471-2105-11-405
  29. 29
    Tsugawa, H.; Cajka, T.; Kind, T.; Ma, Y.; Higgins, B.; Ikeda, K.; Kanazawa, M.; VanderGheynst, J.; Fiehn, O.; Arita, M. MS-DIAL: data-independent MS/MS deconvolution for comprehensive metabolome analysis. Nat. Methods 2015, 12, 523,  DOI: 10.1038/nmeth.3393
  30. 30
    Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. XCMS: Processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Anal. Chem. 2006, 78, 779787,  DOI: 10.1021/ac051437y
  31. 31
    Katajamaa, M.; Miettinen, J.; Orešič, M. MZmine: Toolbox for processing and visualization of mass spectrometry based molecular profile data. Bioinformatics 2006, 22, 634636,  DOI: 10.1093/bioinformatics/btk039
  32. 32
    Tolstikov, V. V.; Lommen, A.; Nakanishi, K.; Tanaka, N.; Fiehn, O. Monolithic Silica-Based Capillary Reversed-Phase Liquid Chromatography/Electrospray Mass Spectrometry for Plant Metabolomics. Anal. Chem. 2003, 75, 67376740,  DOI: 10.1021/ac034716z
  33. 33
    Baran, R.; Kochi, H.; Saito, N.; Suematsu, M.; Soga, T.; Nishioka, T.; Robert, M.; Tomita, M. MathDAMP: A package for differential analysis of metabolite profiles. BMC Bioinf. 2006, 7, 530,  DOI: 10.1186/1471-2105-7-530
  34. 34
    Idborg-Björkman, H.; Edlund, P. O.; Kvalheim, O. M.; Schuppe-Koistinen, I.; Jacobsson, S. P. Screening of biomarkers in rat urine using LC/electrospray ionization-MS and two-way data analysis. Anal. Chem. 2003, 75, 47844792,  DOI: 10.1021/ac0341618
  35. 35
    Marshall, A. G.; Rodgers, R. P. Petroleomics: The Next Grand Challenge for Chemical Analysis. Acc. Chem. Res. 2004, 37, 5359,  DOI: 10.1021/ar020177t
  36. 36
    Hur, M.; Oh, H. B.; Kim, S. Optimized automatic noise level calculations for broadband FT-ICR mass spectra of petroleum give more reliable and faster peak picking results. Bull. Korean Chem. Soc. 2009, 30, 26652668,  DOI: 10.5012/bkcs.2009.30.11.2665
  37. 37
    Rüger, C. P.; Schwemer, T.; Sklorz, M.; O’Connor, P. B.; Barrow, M. P.; Zimmermann, R. Comprehensive chemical comparison of fuel composition and aerosol particles emitted from a ship diesel engine by gas chromatography atmospheric pressure chemical ionisation ultra-high resolution mass spectrometry with improved data processing routines. Eur. J. Mass Spectrom. 2017, 23, 2839,  DOI: 10.1177/1469066717694286
  38. 38
    Schwemer, T.; Rüger, C. P.; Sklorz, M.; Zimmermann, R. Gas Chromatography Coupled to Atmospheric Pressure Chemical Ionization FT-ICR Mass Spectrometry for Improvement of Data Reliability. Anal. Chem. 2015, 87, 1195711961,  DOI: 10.1021/acs.analchem.5b02114
  39. 39
    Tautenhahn, R.; Böttcher, C.; Neumann, S. Highly sensitive feature detection for high resolution LC/MS. BMC Bioinf. 2008,. DOI: 10.1186/1471-2105-9-504
  40. 40
    Åberg, K. M.; Torgrip, R. J.; Kolmert, J.; Schuppe-Koistinen, I.; Lindberg, J. Feature detection and alignment of hyphenated chromatographic-mass spectrometric data. Extraction of pure ion chromatograms using Kalman tracking. J. Chromatogr. A 2008, 1192, 139146,  DOI: 10.1016/j.chroma.2008.03.033
  41. 41
    Gavard, R.; Rossell, D.; Spencer, S. E. F.; Barrow, M. P. Themis: Batch Preprocessing for Ultrahigh-Resolution Mass Spectra of Complex Mixtures. Anal. Chem. 2017, 89, 1138311390,  DOI: 10.1021/acs.analchem.7b02345
  42. 42
    Hughey, C. A.; Rodgers, R. P.; Marshall, A. G.; Qian, K.; Robbins, W. K. Identification of acidic NSO compounds in crude oils of different geochemical origins by negative ion electrospray Fourier transform ion cyclotron resonance mass spectrometry. Org. Geochem. 2002, 33, 743759,  DOI: 10.1016/S0146-6380(02)00038-4
  43. 43
    Stanford, L. A.; Kim, S.; Rodgers, R. P.; Marshall, A. G. Characterization of compositional changes in vacuum gas oil distillation cuts by electrospray ionization Fourier transform- ion cyclotron resonance (FT- ICR) mass spectrometry. Energy Fuels 2006, 20, 16641673,  DOI: 10.1021/ef060104g
  44. 44
    Barrow, M. P.; Headley, J. V.; Peru, K. M.; Derrick, P. J. Data visualization for the characterization of naphthenic acids within petroleum samples. Energy Fuels 2009, 23, 25922599,  DOI: 10.1021/ef800985z
  45. 45
    Van Krevelen, D. Graphical-statistical method for the study of structure and reaction processes of coal. Fuel 1950, 29, 269284
  46. 46
    Kim, S.; Kramer, R. W.; Hatcher, P. G. Graphical Method for Analysis of Ultrahigh-Resolution Broadband Mass Spectra of Natural Organic Matter, the Van Krevelen Diagram. Anal. Chem. 2003, 75, 53365344,  DOI: 10.1021/ac034415p
  47. 47
    Green, N. W.; Perdue, E. M.; Aiken, G. R.; Butler, K. D.; Chen, H.; Dittmar, T.; Niggemann, J.; Stubbins, A. An intercomparison of three methods for the large-scale isolation of oceanic dissolved organic matter. Mar. Chem. 2014, 161, 1419,  DOI: 10.1016/j.marchem.2014.01.012
  48. 48
    Hawkes, J. A.; Hansen, C. T.; Goldhammer, T.; Bach, W.; Dittmar, T. Molecular alteration of marine dissolved organic matter under experimental hydrothermal conditions. Geochim. Cosmochim. Acta 2016, 175, 6885,  DOI: 10.1016/j.gca.2015.11.025
  49. 49
    Palacio Lozano, D. C.; Ramírez, C. X.; Sarmiento Chaparro, J. A.; Thomas, M. J.; Gavard, R.; Jones, H. E.; Cabanzo Hernández, R.; Mejia-Ospino, E.; Barrow, M. P. Characterization of bio-crude components derived from pyrolysis of soft wood and its esterified product by ultrahigh resolution mass spectrometry and spectroscopic techniques. Fuel 2020, 259, 116085,  DOI: 10.1016/j.fuel.2019.116085
  50. 50
    Wickham, H. Tidyverse: Easily install and load ‘Tidyverse’ packages. R package version 2017, 1.
  51. 51
    Chang, W.; Cheng, J.; Allaire, J.; Xie, Y.; McPherson, J. shiny: Web application framework for R [Computer software] , R package version 1.0.0; 2017. https://cran.r-project.org/web/packages/shiny/index.html.
  52. 52
    Zhurov, K. O.; Kozhinov, A. N.; Fornelli, L.; Tsybin, Y. O. Distinguishing analyte from noise components in mass spectra of complex samples: Where to cut the noise?. Anal. Chem. 2014, 86, 33083316,  DOI: 10.1021/ac403278t
  53. 53
    Francl, T. J.; Sherman, M. G.; Hunter, R. L.; Locke, M. J.; Bowers, W. D.; McIver, R. T. Experimental determination of the effects of space charge on ion cyclotron resonance frequencies. Int. J. Mass Spectrom. Ion Processes 1983, 54, 189199,  DOI: 10.1016/0168-1176(83)85017-4
  54. 54
    Barry, J. A.; Robichaud, G.; Muddiman, D. C. Mass recalibration of FT-ICR mass spectrometry imaging data using the average frequency shift of ambient ions. J. Am. Soc. Mass Spectrom. 2013, 24, 11371145,  DOI: 10.1007/s13361-013-0659-0
  55. 55
    Cleveland, W. S.; Grosse, E.; Shyu, W. Local regression models. Statistical models in S; Chambers, J. M., Hastie, T. J.; 1992, pp 309376.
  56. 56
    Wickham, H.; Francois, R.; Müller, K. Tibble: Simple Data Frames. 2018. https://cran.r-project.org/web/packages/tibble/index.html

Cited By

ARTICLE SECTIONS
Jump To

This article is cited by 22 publications.

  1. Maxime Sueur, Julien F. Maillard, Oscar Lacroix-Andrivet, Christopher P. Rüger, Pierre Giusti, Hélène Lavanant, Carlos Afonso. PyC2MC: An Open-Source Software Solution for Visualization and Treatment of High-Resolution Mass Spectrometry Data. Journal of the American Society for Mass Spectrometry 2023, 34 (4) , 617-626. https://doi.org/10.1021/jasms.2c00323
  2. Shinjong Lee, Diana Catalina Palacio Lozano, Hugh E. Jones, Kyongsik Shin, Mark P. Barrow. Characterization of Mineral and Synthetic Base Oils by Gas Chromatography–Mass Spectrometry and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Energy & Fuels 2022, 36 (22) , 13518-13525. https://doi.org/10.1021/acs.energyfuels.2c02437
  3. Latifa K. Alostad, Diana Catalina Palacio Lozano, Benedict Gannon, Rory P. Downham, Hugh E. Jones, Mark P. Barrow. Investigating the Influence of n-Heptane versus n-Nonane upon the Extraction of Asphaltenes. Energy & Fuels 2022, 36 (16) , 8663-8673. https://doi.org/10.1021/acs.energyfuels.2c01168
  4. Diana Catalina Palacio Lozano, Hugh E. Jones, Remy Gavard, Mary J. Thomas, Claudia X. Ramírez, Christopher A. Wootton, José Aristóbulo Sarmiento Chaparro, Peter B. O’Connor, Simon E. F. Spencer, David Rossell, Enrique Mejia-Ospino, Matthias Witt, Mark P. Barrow. Revealing the Reactivity of Individual Chemical Entities in Complex Mixtures: the Chemistry Behind Bio-Oil Upgrading. Analytical Chemistry 2022, 94 (21) , 7536-7544. https://doi.org/10.1021/acs.analchem.2c00261
  5. XiaYinglinResearch Associate ProfessorSunJunProfessor. An Integrated Analysis of Microbiomes and Metabolomics. 2022https://doi.org/10.1021/acsinfocus.7e5003
  6. Max K. Edney, Anna M. Kotowska, Matteo Spanu, Gustavo F. Trindade, Edward Wilmot, Jacqueline Reid, Jim Barker, Jonathan W. Aylott, Alexander G. Shard, Morgan R. Alexander, Colin E. Snape, David J. Scurr. Molecular Formula Prediction for Chemical Filtering of 3D OrbiSIMS Datasets. Analytical Chemistry 2022, 94 (11) , 4703-4711. https://doi.org/10.1021/acs.analchem.1c04898
  7. Martha L. Chacón-Patiño, Murray R. Gray, Christopher Rüger, Donald F. Smith, Taylor J. Glattke, Sydney F. Niles, Anika Neumann, Chad R. Weisbrod, Andrew Yen, Amy M. McKenna, Pierre Giusti, Brice Bouyssiere, Caroline Barrère-Mangote, Harvey Yarranton, Christopher L. Hendrickson, Alan G. Marshall, Ryan P. Rodgers. Lessons Learned from a Decade-Long Assessment of Asphaltenes by Ultrahigh-Resolution Mass Spectrometry and Implications for Complex Mixture Analysis. Energy & Fuels 2021, 35 (20) , 16335-16376. https://doi.org/10.1021/acs.energyfuels.1c02107
  8. Steven M. Rowland, Donald F. Smith, Gregory T. Blakney, Yuri E. Corilo, Christopher L. Hendrickson, Ryan P. Rodgers. Online Coupling of Liquid Chromatography with Fourier Transform Ion Cyclotron Resonance Mass Spectrometry at 21 T Provides Fast and Unique Insight into Crude Oil Composition. Analytical Chemistry 2021, 93 (41) , 13749-13754. https://doi.org/10.1021/acs.analchem.1c01169
  9. Hugh E. Jones, Diana Catalina Palacio Lozano, Chris Huener, Mary J. Thomas, David J. Aaserud, J. Corinne DeMuth, Mathew P. Robin, Mark P. Barrow. Influence of Biodiesel on Base Oil Oxidation as Measured by FTICR Mass Spectrometry. Energy & Fuels 2021, 35 (15) , 11896-11908. https://doi.org/10.1021/acs.energyfuels.1c01240
  10. Limei Han, Jan Kaesler, Chang Peng, Thorsten Reemtsma, Oliver J. Lechtenfeld. Online Counter Gradient LC-FT-ICR-MS Enables Detection of Highly Polar Natural Organic Matter Fractions. Analytical Chemistry 2021, 93 (3) , 1740-1748. https://doi.org/10.1021/acs.analchem.0c04426
  11. Evan Terrell, Manuel Garcia-Perez. Novel Strategy To Analyze Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Data of Biomass Pyrolysis Oil for Oligomeric Structure Assignment. Energy & Fuels 2020, 34 (7) , 8466-8481. https://doi.org/10.1021/acs.energyfuels.0c01687
  12. Rory P. Downham, Benedict Gannon, Diana Catalina Palacio Lozano, Hugh E. Jones, Christopher H. Vane, Mark P. Barrow. Tracking the history of polycyclic aromatic compounds in London through a River Thames sediment core and ultrahigh resolution mass spectrometry. Journal of Hazardous Materials 2024, 473 , 134605. https://doi.org/10.1016/j.jhazmat.2024.134605
  13. Patricia Aguilar-Alarcón, Susana V. Gonzalez, Øyvind Mikkelsen, Alexandros G. Asimakopoulos. Molecular formula assignment of dissolved organic matter by ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry using two non-targeted data processing approaches: A case study from recirculating aquaculture systems. Analytica Chimica Acta 2024, 1288 , 342128. https://doi.org/10.1016/j.aca.2023.342128
  14. Diana Catalina Palacio Lozano, Hugh E. Jones, Mark P. Barrow, Martin Wills. Chemoselective derivatisation and ultrahigh resolution mass spectrometry for the determination of hydroxyl functional groups within complex bio-oils. RSC Advances 2023, 13 (26) , 17727-17741. https://doi.org/10.1039/D3RA02779A
  15. Clarisse Gosset-Erard, Frédéric Aubriet, Emmanuelle Leize-Wagner, Yannis-Nicolas François, Patrick Chaimbault. Hyphenation of Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) with separation methods: The art of compromises and the possible - A review. Talanta 2023, 257 , 124324. https://doi.org/10.1016/j.talanta.2023.124324
  16. Alina T. Roman-Hubers, Alexandra C. Cordova, Mark P. Barrow, Ivan Rusyn. Analytical chemistry solutions to hazard evaluation of petroleum refining products. Regulatory Toxicology and Pharmacology 2023, 137 , 105310. https://doi.org/10.1016/j.yrtph.2022.105310
  17. Yueyi Xia, Xiaoxiao Wang, Chenfei Ma, Xinxin Wang, Chunxia Zhao, Xinjie Zhao, Zhanquan Zhang, Yinglong Yu, Xiaohui Lin, Xin Lu, Guowang Xu. A data processing pipeline for petroleomics based on liquid chromatography-high resolution mass spectrometry. Journal of Chromatography A 2022, 1673 , 463194. https://doi.org/10.1016/j.chroma.2022.463194
  18. Diana Catalina Palacio Lozano, Hugh E. Jones, Tomas Ramirez Reina, Roberto Volpe, Mark P. Barrow. Unlocking the potential of biofuels via reaction pathways in van Krevelen diagrams. Green Chemistry 2021, 23 (22) , 8949-8963. https://doi.org/10.1039/D1GC01796A
  19. Mary J. Thomas, Hugh E. Jones, Diana Catalina Palacio Lozano, Rémy Gavard, Shaun Carney, Mark P. Barrow. Comprehensive analysis of multiple asphaltene fractions combining statistical analyses and novel visualization tools. Fuel 2021, 291 , 120132. https://doi.org/10.1016/j.fuel.2021.120132
  20. Beibei Ma, Tianyu Lou, Tingting Wang, Ruiji Li, Jinhui Liu, Shangyue Yu, Yudong Guo, Zhibin Wang, Jing Wang. Comprehensive metabolism study of swertiamarin in rats using ultra high-performance liquid chromatography coupled with Quadrupole-Exactive Orbitrap mass spectrometry. Xenobiotica 2021, 51 (4) , 455-466. https://doi.org/10.1080/00498254.2020.1869856
  21. Bebiana C. Sousa, Michael J.O. Wakelam, Andrea F. Lopez-Clavijo. Methods of lipid analysis. 2021, 53-83. https://doi.org/10.1016/B978-0-12-824048-9.00010-9
  22. Claudia X. Ramírez, Diana Catalina Palacio Lozano, Hugh E. Jones, Rafael Cabanzo Hernández, Mark P. Barrow, Enrique Mejia-Ospino. The fingerprint of essential bio-oils by Fourier transform ion cyclotron resonance mass spectrometry. The Analyst 2020, 145 (9) , 3414-3423. https://doi.org/10.1039/C9AN02492A
  • Abstract

    Figure 1

    Figure 1. KairosMS workflow. In Step 1, a mass list containing an identification number, m/z, intensity, and retention time is extracted using the instrument manufacturer’s software. The data processing is then performed during Steps 2–6 in KairosMS. KairosMS generates a single data set for compositional assignments (Step 7). Finally, a single data set or multiple data sets can be opened in KairosMS for interactive visualization and further data analysis (Step 8). All data and graphics are exportable.

    Figure 2

    Figure 2. Average frequency shift in Hz within EICs for each retention time before recalibration (red) and after (black) for the OSPW sample.

    Figure 3

    Figure 3. Histogram of the mass error in ppm of each peak for each EIC with the assigned m/z.

    Figure 4

    Figure 4. Comparison of TICs before (top line, black) and after (bottom line, red) processing.

    Figure 5

    Figure 5. (A) Heteroatom class distribution for retention times 12.10–12 min and 31.10–32 min. (B) EICs of selected homologous series from the O2[H] class. (C) EICs of selected molecular composition for the OSPW and two groundwater samples (G1 and G2).

    Figure 6

    Figure 6. (A) Total ion chromatogram (TIC) with the mass spectra for retention times 10.72 min and 20.57 min. (B) EICs for the heteroatom classes O4[H] and O7[H] with associated AUC. (C) Heteroatom class distribution for retention times 10.72–10.80 min and 20.57–20.66 min for the SRFA and Marine DOM samples.

    Figure 7

    Figure 7. (A) van Krevelen diagram with EICs displayed for two different points on the plot. (B) DBE vs carbon number of the O2[H] and O5[H] heteroatom classes with associated EICs for two data points on the plot. (C) DBE vs carbon number of the O3[H] and O4[H] heteroatom classes with associated EICs for two data points. Note that each data point in a van Krevelen diagram represents the sum of multiple molecular formulas, while each data point in a DBE vs carbon number plot is a single molecular formula.

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 56 other publications.

    1. 1
      Hertkorn, N.; Ruecker, C.; Meringer, M.; Gugisch, R.; Frommberger, M.; Perdue, E. M.; Witt, M.; Schmitt-Kopplin, P. High-precision frequency measurements: Indispensable tools at the core of the molecular-level analysis of complex systems. Anal. Bioanal. Chem. 2007, 389, 13111327,  DOI: 10.1007/s00216-007-1577-4
    2. 2
      Barrow, M. P. Petroleomics: study of the old and the new. Biofuels 2010, 1, 651655,  DOI: 10.4155/bfs.10.55
    3. 3
      Mullins, O. C.; Sheu, E. Y.; Hammami, A.; Marshall, A. G.; Eds. Asphaltenes, Heavy Oils, and Petroleomics; Springer: New York, NY, 2007.
    4. 4
      Marshall, A. G.; Rodgers, R. P. Petroleomics: Chemistry of the underworld. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 1809018095,  DOI: 10.1073/pnas.0805069105
    5. 5
      Hsu, C. S.; Hendrickson, C. L.; Rodgers, R. P.; McKenna, A. M.; Marshall, A. G. Petroleomics: Advanced molecular probe for petroleum heavy ends. J. Mass Spectrom. 2011, 46, 337343,  DOI: 10.1002/jms.1893
    6. 6
      Kellerman, A. M.; Dittmar, T.; Kothawala, D. N.; Tranvik, L. J. Chemodiversity of dissolved organic matter in lakes driven by climate and hydrology. Nat. Commun. 2014, 5. DOI: 10.1038/ncomms4804
    7. 7
      Stubbins, A.; Dittmar, T. Illuminating the deep: Molecular signatures of photochemical alteration of dissolved organic matter from North Atlantic Deep Water. Mar. Chem. 2015, 177, 318324,  DOI: 10.1016/j.marchem.2015.06.020
    8. 8
      Headley, J. V.; Peru, K. M.; Barrow, M. P. Advances in mass spectrometric characterization of naphthenic acids fraction compounds in oil sands environmental samples and crude oil - A review. Mass Spectrom. Rev. 2016, 35, 311328,  DOI: 10.1002/mas.21472
    9. 9
      Comisarow, M. B.; Marshall, A. G. Fourier transform ion cyclotron resonance spectroscopy. Chem. Phys. Lett. 1974, 25, 282283,  DOI: 10.1016/0009-2614(74)89137-2
    10. 10
      Comisarow, M. B.; Marshall, A. G. Selective-phase ion cyclotron resonance spectroscopy. Can. J. Chem. 1974, 52, 19971999,  DOI: 10.1139/v74-288
    11. 11
      Comisarow, M. B.; Marshall, A. G. Frequency-sweep fourier transform ion cyclotron resonance spectroscopy. Chem. Phys. Lett. 1974, 26, 489490,  DOI: 10.1016/0009-2614(74)80397-0
    12. 12
      Amster, I. J. Fourier transform mass spectrometry. J. Mass Spectrom. 1996, 31, 13251337,  DOI: 10.1002/(SICI)1096-9888(199612)31:12<1325::AID-JMS453>3.0.CO;2-W
    13. 13
      Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Fourier Transform Ion Cyclotron Resonance Mass Spectromeyry: A Primer. Mass Spectrom. Rev. 1998, 17, 135,  DOI: 10.1002/(SICI)1098-2787(1998)17:1<1::AID-MAS1>3.0.CO;2-K
    14. 14
      Barrow, M. P.; Burkitt, W. I.; Derrick, P. J. Principles of Fourier transform ion cyclotron resonance mass spectrometry and its application in structural biology. Analyst 2005, 130, 18,  DOI: 10.1039/b403880k
    15. 15
      Koch, B. P.; Ludwichowski, K. U.; Kattner, G.; Dittmar, T.; Witt, M. Advanced characterization of marine dissolved organic matter by combining reversed-phase liquid chromatography and FT-ICR-MS. Mar. Chem. 2008, 111, 233241,  DOI: 10.1016/j.marchem.2008.05.008
    16. 16
      Dittmar, T.; Stubbins, A. Treatise Geochemistry, 2nd ed.; Elsevier Ltd., 2013; Vol. 12; pp 125156.
    17. 17
      Headley, J. V.; Peru, K. M.; Janfada, A.; Fahlman, B.; Gu, C.; Hassan, S. Characterization of oil sands acids in plant tissue using Orbitrap ultra-high resolution mass spectrometry with electrospray ionization. Rapid Commun. Mass Spectrom. 2011, 25, 459462,  DOI: 10.1002/rcm.4877
    18. 18
      Zhurov, K. O.; Kozhinov, A. N.; Tsybin, Y. O. Evaluation of high-field orbitrap fourier transform mass spectrometer for petroleomics. Energy Fuels 2013, 27, 29742983,  DOI: 10.1021/ef400203g
    19. 19
      Hawkes, J. A.; Dittmar, T.; Patriarca, C.; Tranvik, L.; Bergquist, J. Evaluation of the Orbitrap Mass Spectrometer for the Molecular Fingerprinting Analysis of Natural Dissolved Organic Matter. Anal. Chem. 2016, 88, 76987704,  DOI: 10.1021/acs.analchem.6b01624
    20. 20
      Cho, E.; Witt, M.; Hur, M.; Jung, M. J.; Kim, S. Application of FT-ICR MS Equipped with Quadrupole Detection for Analysis of Crude Oil. Anal. Chem. 2017, 89, 1210112107,  DOI: 10.1021/acs.analchem.7b02644
    21. 21
      Krajewski, L. C.; Rodgers, R. P.; Marshall, A. G. Assigned Chemical Formulas from an Atmospheric Pressure Photoionization 9.4 T Fourier Transform Positive Ion Cyclotron Resonance Mass Spectrum. Anal. Chem. 2017, 89, 11318,  DOI: 10.1021/acs.analchem.7b02004
    22. 22
      Smith, D. F.; Podgorski, D. C.; Rodgers, R. P.; Blakney, G. T.; Hendrickson, C. L. 21 T FT-ICR Mass Spectrometer for Ultrahigh-Resolution Analysis of Complex Organic Mixtures. Anal. Chem. 2018, 90, 20412047,  DOI: 10.1021/acs.analchem.7b04159
    23. 23
      Palacio Lozano, D. C.; Gavard, R.; Arenas-Diaz, J. P.; Thomas, M. J.; Stranz, D. D.; Mejía-Ospino, E.; Guzman, A.; Spencer, S. E. F.; Rossell, D.; Barrow, M. P. Pushing the analytical limits: new insights into complex mixtures using mass spectra segments of constant ultrahigh resolving power. Chem. Sci. 2019, 10, 69666978,  DOI: 10.1039/C9SC02903F
    24. 24
      Barrow, M. P.; Peru, K. M.; Headley, J. V. An added dimension: GC atmospheric pressure chemical ionization FTICR MS and the athabasca oil sands. Anal. Chem. 2014, 86, 82818288,  DOI: 10.1021/ac501710y
    25. 25
      Patriarca, C.; Bergquist, J.; Sjoberg, P. J. R.; Tranvik, L.; Hawkes, J. A. Online HPLC-ESI-HRMS Method for the Analysis and Comparison of Different Dissolved Organic Matter Samples. Environ. Sci. Technol. 2018, 52, 20912099, PMID: 29241333.  DOI: 10.1021/acs.est.7b04508
    26. 26
      Kim, S.; Kim, D.; Kim, S.; Son, S.; Jung, M. J. Application of Online Liquid Chromatography 7 T FT-ICR Mass Spectrometer Equipped with Quadrupolar Detection for Analysis of Natural Organic Matter. Anal. Chem. 2019, 91, 76907697,  DOI: 10.1021/acs.analchem.9b00689
    27. 27
      Benigni, P.; Thompson, C. J.; Ridgeway, M. E.; Park, M. A.; Fernandez-Lima, F. Targeted high-resolution ion mobility separation coupled to ultrahigh-resolution mass spectrometry of endocrine disruptors in complex mixtures. Anal. Chem. 2015, 87, 43214325,  DOI: 10.1021/ac504866v
    28. 28
      Wenig, P.; Odermatt, J. OpenChrom: a cross-platform open source software for the mass spectrometric analysis of chromatographic data. BMC Bioinf. 2010, 11. DOI: 10.1186/1471-2105-11-405
    29. 29
      Tsugawa, H.; Cajka, T.; Kind, T.; Ma, Y.; Higgins, B.; Ikeda, K.; Kanazawa, M.; VanderGheynst, J.; Fiehn, O.; Arita, M. MS-DIAL: data-independent MS/MS deconvolution for comprehensive metabolome analysis. Nat. Methods 2015, 12, 523,  DOI: 10.1038/nmeth.3393
    30. 30
      Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. XCMS: Processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Anal. Chem. 2006, 78, 779787,  DOI: 10.1021/ac051437y
    31. 31
      Katajamaa, M.; Miettinen, J.; Orešič, M. MZmine: Toolbox for processing and visualization of mass spectrometry based molecular profile data. Bioinformatics 2006, 22, 634636,  DOI: 10.1093/bioinformatics/btk039
    32. 32
      Tolstikov, V. V.; Lommen, A.; Nakanishi, K.; Tanaka, N.; Fiehn, O. Monolithic Silica-Based Capillary Reversed-Phase Liquid Chromatography/Electrospray Mass Spectrometry for Plant Metabolomics. Anal. Chem. 2003, 75, 67376740,  DOI: 10.1021/ac034716z
    33. 33
      Baran, R.; Kochi, H.; Saito, N.; Suematsu, M.; Soga, T.; Nishioka, T.; Robert, M.; Tomita, M. MathDAMP: A package for differential analysis of metabolite profiles. BMC Bioinf. 2006, 7, 530,  DOI: 10.1186/1471-2105-7-530
    34. 34
      Idborg-Björkman, H.; Edlund, P. O.; Kvalheim, O. M.; Schuppe-Koistinen, I.; Jacobsson, S. P. Screening of biomarkers in rat urine using LC/electrospray ionization-MS and two-way data analysis. Anal. Chem. 2003, 75, 47844792,  DOI: 10.1021/ac0341618
    35. 35
      Marshall, A. G.; Rodgers, R. P. Petroleomics: The Next Grand Challenge for Chemical Analysis. Acc. Chem. Res. 2004, 37, 5359,  DOI: 10.1021/ar020177t
    36. 36
      Hur, M.; Oh, H. B.; Kim, S. Optimized automatic noise level calculations for broadband FT-ICR mass spectra of petroleum give more reliable and faster peak picking results. Bull. Korean Chem. Soc. 2009, 30, 26652668,  DOI: 10.5012/bkcs.2009.30.11.2665
    37. 37
      Rüger, C. P.; Schwemer, T.; Sklorz, M.; O’Connor, P. B.; Barrow, M. P.; Zimmermann, R. Comprehensive chemical comparison of fuel composition and aerosol particles emitted from a ship diesel engine by gas chromatography atmospheric pressure chemical ionisation ultra-high resolution mass spectrometry with improved data processing routines. Eur. J. Mass Spectrom. 2017, 23, 2839,  DOI: 10.1177/1469066717694286
    38. 38
      Schwemer, T.; Rüger, C. P.; Sklorz, M.; Zimmermann, R. Gas Chromatography Coupled to Atmospheric Pressure Chemical Ionization FT-ICR Mass Spectrometry for Improvement of Data Reliability. Anal. Chem. 2015, 87, 1195711961,  DOI: 10.1021/acs.analchem.5b02114
    39. 39
      Tautenhahn, R.; Böttcher, C.; Neumann, S. Highly sensitive feature detection for high resolution LC/MS. BMC Bioinf. 2008,. DOI: 10.1186/1471-2105-9-504
    40. 40
      Åberg, K. M.; Torgrip, R. J.; Kolmert, J.; Schuppe-Koistinen, I.; Lindberg, J. Feature detection and alignment of hyphenated chromatographic-mass spectrometric data. Extraction of pure ion chromatograms using Kalman tracking. J. Chromatogr. A 2008, 1192, 139146,  DOI: 10.1016/j.chroma.2008.03.033
    41. 41
      Gavard, R.; Rossell, D.; Spencer, S. E. F.; Barrow, M. P. Themis: Batch Preprocessing for Ultrahigh-Resolution Mass Spectra of Complex Mixtures. Anal. Chem. 2017, 89, 1138311390,  DOI: 10.1021/acs.analchem.7b02345
    42. 42
      Hughey, C. A.; Rodgers, R. P.; Marshall, A. G.; Qian, K.; Robbins, W. K. Identification of acidic NSO compounds in crude oils of different geochemical origins by negative ion electrospray Fourier transform ion cyclotron resonance mass spectrometry. Org. Geochem. 2002, 33, 743759,  DOI: 10.1016/S0146-6380(02)00038-4
    43. 43
      Stanford, L. A.; Kim, S.; Rodgers, R. P.; Marshall, A. G. Characterization of compositional changes in vacuum gas oil distillation cuts by electrospray ionization Fourier transform- ion cyclotron resonance (FT- ICR) mass spectrometry. Energy Fuels 2006, 20, 16641673,  DOI: 10.1021/ef060104g
    44. 44
      Barrow, M. P.; Headley, J. V.; Peru, K. M.; Derrick, P. J. Data visualization for the characterization of naphthenic acids within petroleum samples. Energy Fuels 2009, 23, 25922599,  DOI: 10.1021/ef800985z
    45. 45
      Van Krevelen, D. Graphical-statistical method for the study of structure and reaction processes of coal. Fuel 1950, 29, 269284
    46. 46
      Kim, S.; Kramer, R. W.; Hatcher, P. G. Graphical Method for Analysis of Ultrahigh-Resolution Broadband Mass Spectra of Natural Organic Matter, the Van Krevelen Diagram. Anal. Chem. 2003, 75, 53365344,  DOI: 10.1021/ac034415p
    47. 47
      Green, N. W.; Perdue, E. M.; Aiken, G. R.; Butler, K. D.; Chen, H.; Dittmar, T.; Niggemann, J.; Stubbins, A. An intercomparison of three methods for the large-scale isolation of oceanic dissolved organic matter. Mar. Chem. 2014, 161, 1419,  DOI: 10.1016/j.marchem.2014.01.012
    48. 48
      Hawkes, J. A.; Hansen, C. T.; Goldhammer, T.; Bach, W.; Dittmar, T. Molecular alteration of marine dissolved organic matter under experimental hydrothermal conditions. Geochim. Cosmochim. Acta 2016, 175, 6885,  DOI: 10.1016/j.gca.2015.11.025
    49. 49
      Palacio Lozano, D. C.; Ramírez, C. X.; Sarmiento Chaparro, J. A.; Thomas, M. J.; Gavard, R.; Jones, H. E.; Cabanzo Hernández, R.; Mejia-Ospino, E.; Barrow, M. P. Characterization of bio-crude components derived from pyrolysis of soft wood and its esterified product by ultrahigh resolution mass spectrometry and spectroscopic techniques. Fuel 2020, 259, 116085,  DOI: 10.1016/j.fuel.2019.116085
    50. 50
      Wickham, H. Tidyverse: Easily install and load ‘Tidyverse’ packages. R package version 2017, 1.
    51. 51
      Chang, W.; Cheng, J.; Allaire, J.; Xie, Y.; McPherson, J. shiny: Web application framework for R [Computer software] , R package version 1.0.0; 2017. https://cran.r-project.org/web/packages/shiny/index.html.
    52. 52
      Zhurov, K. O.; Kozhinov, A. N.; Fornelli, L.; Tsybin, Y. O. Distinguishing analyte from noise components in mass spectra of complex samples: Where to cut the noise?. Anal. Chem. 2014, 86, 33083316,  DOI: 10.1021/ac403278t
    53. 53
      Francl, T. J.; Sherman, M. G.; Hunter, R. L.; Locke, M. J.; Bowers, W. D.; McIver, R. T. Experimental determination of the effects of space charge on ion cyclotron resonance frequencies. Int. J. Mass Spectrom. Ion Processes 1983, 54, 189199,  DOI: 10.1016/0168-1176(83)85017-4
    54. 54
      Barry, J. A.; Robichaud, G.; Muddiman, D. C. Mass recalibration of FT-ICR mass spectrometry imaging data using the average frequency shift of ambient ions. J. Am. Soc. Mass Spectrom. 2013, 24, 11371145,  DOI: 10.1007/s13361-013-0659-0
    55. 55
      Cleveland, W. S.; Grosse, E.; Shyu, W. Local regression models. Statistical models in S; Chambers, J. M., Hastie, T. J.; 1992, pp 309376.
    56. 56
      Wickham, H.; Francois, R.; Müller, K. Tibble: Simple Data Frames. 2018. https://cran.r-project.org/web/packages/tibble/index.html
  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.9b05113.

    • Figure S1: Screenshot presenting KairosMS interface; Figure S2: Mass spectrum created using the EICs extracted to be used for molecular assignments; Figure S3: Comparison of the elution of the O2[H] class contribution between an OSPW and two groundwater samples using a scan by scan resolution; Figure S4: EICs for the monoisotopic form and isotopologues of C16H26O2 [H] for the G1, G2, and OSPW samples; Figure S5: Elution of DBE series (homologous series) comprising the O2[H] class; Figure S6: Percentage of contribution to the total signal for all of the classes identified in the SRFA and Marine DOM samples; Figure S7: Elemental contributions for the samples SRFA and Marine DOM based on all of the assigned EICs; Figure S8: van Krevelen diagram of the H/C ratio vs O/C ratio for the Marine DOM and SRFA samples; Figure S9: Comparison of the EIC of the same molecular assignment as seen in DA and KairosMS after peak picking at S/N 1; Figure S10: EIC from a peptide digest of ubiquitin analyzed by LC-FTICR MS (PDF)

    • Video showing interactive data visualization using KairosMS (MP4)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.