ACS Publications. Most Trusted. Most Cited. Most Read
Breaking Down Structural Diversity for Comprehensive Prediction of Ion-Neutral Collision Cross Sections
More

OR SEARCH CITATIONS

My Activity
Publications

Figure 1Loading Img
Download Hi-Res ImageDownload to MS-PowerPointCite This:Anal. Chem. 2020, 92, 6, 4548-4557
  • Open Access
Article

Breaking Down Structural Diversity for Comprehensive Prediction of Ion-Neutral Collision Cross Sections
Click to copy article linkArticle link copied!

  • Dylan H. Ross
    Dylan H. Ross
    Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195, United States
    More by Dylan H. Ross
  • Jang Ho Cho
    Jang Ho Cho
    Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195, United States
    More by Jang Ho Cho
  • Libin Xu*
    Libin Xu
    Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195, United States
    *Email: [email protected]. Tel: (206) 543-1080. Fax: (206) 685-3252.
    More by Libin Xu
    Orcidhttp://orcid.org/0000-0003-1021-5200
Open PDFSupporting Information (1)

Analytical Chemistry

Cite this: Anal. Chem. 2020, 92, 6, 4548–4557
Click to copy citationCitation copied!
https://doi.org/10.1021/acs.analchem.9b05772
Published February 25, 2020

Copyright © 2020 American Chemical Society. This publication is licensed under these Terms of Use.

Request reuse permissions

Abstract

Click to copy section linkSection link copied!
High Resolution Image
Download MS PowerPoint Slide

Identification of unknowns is a bottleneck for large-scale untargeted analyses like metabolomics or drug metabolite identification. Ion mobility-mass spectrometry (IM-MS) provides rapid two-dimensional separation of ions based on their mobility through a neutral buffer gas. The mobility of an ion is related to its collision cross section (CCS) with the buffer gas, a physical property that is determined by the size and shape of the ion. This structural dependency makes CCS a promising characteristic for compound identification, but this utility is limited by the availability of high-quality reference CCS values. CCS prediction using machine learning (ML) has recently shown promise in the field, but accurate and broadly applicable models are still lacking. Here we present a novel ML approach that employs a comprehensive collection of CCS values covering a wide range of chemical space. Using this diverse database, we identified the structural characteristics, represented by molecular quantum numbers (MQNs), that contribute to variance in CCS and assessed the performance of a variety of ML algorithms in predicting CCS. We found that by breaking down the chemical structural diversity using unsupervised clustering based on the MQNs, specific and accurate prediction models for each cluster can be trained, which showed superior performance than a single model trained with all data. Using this approach, we have robustly trained and characterized a CCS prediction model with high accuracy on diverse chemical structures. An all-in-one web interface (https://CCSbase.net) was built for querying the CCS database and accessing the predictive model to support unknown compound identifications.

Copyright © 2020 American Chemical Society

Subjects

what are subjects
Identification of unknown analytes is challenging in LC-MS-based metabolomics or drug metabolite identification experiments as identification is often dependent upon the availability of reference MS/MS spectra, and even with MS/MS spectral matching, the structural information garnered is often inconclusive because of uncharacteristic MS/MS fragmentations. Definitive compound identifications can be achieved using nuclear magnetic resonance (NMR) spectroscopy, but the large amount of material required makes this method costly and time-consuming. Therefore, there is a demand for increased confidence in identification of unknowns without sacrificing analytical throughput. (1,2)
Ion mobility spectrometry (IMS) is a rapid gas-phase separation technique based on the size and shape of analyte ions in the gas phase. (3−7) In IM separation, ions are driven through a neutral buffer gas under the influence of either a static electric field (as in drift tube IM, DTIM) or a dynamic electric field (as in traveling wave IM, TWIM), (8,9) where they are differentially slowed because of their interactions with the drift gas molecules. The mobility of an ion in IMS is related to its collision cross section (CCS), a unique physical property determined by the ion’s size, shape, and degree of interactions with the drift gas in the gas-phase. (10) When IMS is coupled with mass spectrometry (IM-MS), a powerful two-dimensional separation is achieved on the basis of CCS-to-charge and mass-to-charge ratio (m/z). In some cases, such separation is sufficient to justify the shortening or omission of chromatographic separation (e.g., liquid chromatography) entirely, which dramatically increases the analytical throughput while simultaneously providing additional structural information. In addition to its inherent relationship to gas-phase structure, CCS has the benefit of being highly reproducible across different measurements and instruments. A recent study comparing the reproducibility of CCS values measured on DTIM and TWIM instruments found that 93% of the compounds tested showed absolute errors ≤2% for protonated species, and 87% showed absolute errors ≤2% for sodiated adducts; however, a few entries displayed errors >3% (<4% of total comparable entries). (11) This reproducibility becomes even greater when considering CCS values measured on a single type of instrument. (12)
The structural dependency of CCS and the high reproducibility of its measurement make CCS an excellent property to be used for compound identification. This utility, however, is limited by the availability of reference CCS values to compare against. Many large collections of CCS values have been produced that cover various chemical classes, but because of the breadth of small molecule chemical space and limitation in available standards, it is infeasible, if not impossible, to cover all of the unknowns encountered in metabolomics or related studies. CCS prediction can be used to address this limitation by providing CCS values for compounds that do not have reference values. Current methods of CCS prediction fall into two categories: theory-driven (10,13−21) and data-driven approaches, (22−27) each of which has unique benefits and drawbacks. Theory-driven approaches generally use molecular modeling to produce an approximation of the 3D and electronic structure of a molecule and then compute CCS by simulating the interactions between the drift gas and analyte ion (both at varying levels of detail). These methods have the benefit of being rooted in solid theoretical principles, but depending on the level of detail, the computations can become very time-consuming and laborious to set up. Furthermore, depending on the assumptions made by the theory of a chosen approach, systematic errors may also be introduced and can be difficult to correct without introducing bias. Recently, a higher-throughput computational workflow, ISiCLE, was reported, but this method still requires large amounts of computational resources, multiple steps of computational setup (conformer generation and optimization and CCS calculation for each conformer), and thus, the throughput is still not ideal. (21) Furthermore, the computation results are good but still not superb (averaged error = 3.2% and R2 = 0.94–0.96 in correlation with experimental values). Additionally, data-driven approaches leverage trends within collections of reference values and generate predictions for similar compounds using their common characteristics, and they typically employ machine learning (ML) to do so. Data-driven approaches have the benefit that once a predictive model has been trained, predictions can be carried out almost instantaneously. These models can produce high quality predictions on unseen data, with the caveat that the quality of predicted CCS values is tied directly to the quality and relevance of the training data.
CCS prediction using ML has gained traction in recent years, (22−27) and a range of approaches have been adopted by different groups. These approaches share a similar general workflow: compilation and/or generation of a suitable set of training data, selection and optimization of a numerical representation for compounds (featurization), partitioning of data into training and testing/validation sets, selection of a ML model, training and optimizing the selected model, and finally testing and validating the trained model’s performance. The majority of published predictive models are trained on relatively specialized, single collections of compounds like small molecule metabolites, (22) lipids, (23,25) pesticides, (24) and drug-like compounds, (26) and therefore, the accurate prediction is limited to the types of molecules used for training. A recent study used data sets compiled from existing collections (mostly DTIM values, but some TWIM values) in order to train predictive models that can handle a wider variety of chemical features, (27) but there is limited coverage of drug and drug-like compounds (i.e., small molecules). Most existing models use some form of molecular descriptors (MDs) computed from the chemical structures as the features for performing ML. (22−26) These MDs often correspond to a summary property of a compound, such as polarity, LogP, and number of heavy atoms, and the combination of a variety of MDs provides a fingerprint of a compound’s chemical characteristics, but these MDs do not always reflect a compound’s structural features. The SMILES structure has also recently been used as input in a recent work using a convolutional neural network (CNN). (27) A variety of schemes have been used to partition data sets for training and validation, but generally, all examples include some form of performance validation using data not seen by the model during training. Another limitation of existing models is that they use either support vector machines (SVMs) (22,23) or artificial neural networks (ANNs), (24,26,27) which operate as “black boxes” and thus offer no interpretation of results beyond the predictions themselves.
To overcome the current limitations in CCS prediction, we first rigorously curated a large CCS database with nearly 7700 entries covering a diverse range of chemical space and identified structural characteristics, represented by molecular quantum numbers (MQNs), contributing to the variance in mass-CCS space. We then developed a novel and high-performance approach of building CCS prediction models by breaking down the chemical structural diversity using unsupervised clustering based on structural features, followed by training of individual prediction models on each cluster using ML. Using this approach, we robustly trained and characterized a comprehensive CCS prediction model with high accuracy on diverse chemical classes. Lastly, we built an all-in-one web interface (https://CCSbase.net) for querying the CCS database and accessing the predictive model to support unknown compound identifications.

Experimental Section

Click to copy section linkSection link copied!

Assembly of a Comprehensive CCS Database

A comprehensive CCS database was assembled from a variety of individual collections of CCS values available in the literature, (22−24,26,28−40) representing broad coverage of current measurement techniques and chemical classes. For lipid CCS values measured on TWIM instruments, only those calibrated with lipids were included into the database. (41) The source data sets were each manually examined for any errors, and relevant data (i.e., CCS, m/z, mass, SMILES if present, etc.) from each entry was converted into a JSON format, yielding consistently formatted cleaned data with separate files for each individual data set. The combined CCS database was constructed from the individual cleaned data sets using a series of build scripts developed in-house in order to be able to reproducibly rebuild the database when new data sets are added or when database organization is changed. A SQLite3 relational database was initialized with a table to hold relevant CCS measurement data (including MS adduct, m/z, and charge state) and metadata [including source data set and CCS measurement platforms and methods (DTIM: single field or stepped field vs TWIM: calibrants)]. Data from each source data set was added to the database, then entries with missing SMILES structures were attempted to be filled first by searching PubChem or LIPID MAPS databases by compound name and if not found in the existing databases, SMILES were obtained from manual search in the database or hand-drawn structures in ChemDraw, in combination with the online SMILES translator (https://cactus.nci.nih.gov/translate/). Next, a table was added to the database containing columns for each of the 42 molecular quantum numbers (MQNs) used as part of the features for machine learning (see below). Finally, rough chemical classifications (carbohydrates, lipids, peptides, small molecules) were assigned to each entry of the database on the basis of compound name and data source.

Feature Set for Machine Learning

The full set of features used for prediction of CCS (n = 50) includes the m/z of the observed MS adduct, one-hot encoded MS adduct (OHEA, n = 7), and a set of molecular descriptors that capture information about the size, composition, and topology of each chemical structure (42 MQNs). The m/z and MS adduct were already present in the source data sets, but the MS adduct had to be encoded into a numeric form in order to be used for CCS prediction. One-hot encoding was used to convert the MS adduct into a binary representation, with unique labels for each adduct type that had ≥100 examples in the database and the rest of adduct types represented under “other adducts” (total of 7 features, Table S1). The 42 MQNs were computed for all database entries containing SMILES structures using the RDKit library (https://www.rdkit.org) and stored in the database. This feature set reflects a variety of compound characteristics, ranging from size to composition and to structure topology (see complete list in Supporting Information, Table S2). Specifically, MQNs are molecular descriptors obtained from analyzing compounds as a molecular graph: i.e. collections of nodes (atoms) and edges (bonds). (42) The descriptors are properties of these graphs, consisting primarily of counts of various atom types, bond orders, connectivity, etc. A benefit of using graph properties as molecular descriptors is that they are invariant with respect to the software used to compute them (unlike many empirical properties, such as cLogP), facilitating the broad application of predictive models and promoting reproducibility.

Analysis of Structural Features Contributing to Variation in the Mass-CCS Space

Dimensionality reduction analyses were used to explore the important sources of variance within the combined CCS database and to determine what chemical characteristics contribute most strongly to variance in CCS values. Specifically, principal component analysis (PCA) and partial least-squares regression analysis (PLS-RA) were used to analyze the MQNs of all entries in the combined CCS database in an untargeted and targeted fashion, respectively. Both analyses work by finding multidimensional axes in the input data that explains as much variance as possible, and then subsequent orthogonal axes are chosen that explain as much of the remaining variance as possible. PLS-RA differs from PCA in that the first axis chosen is the one that explains the maximal variance in a target variable, in this case, CCS, rather than the input data, making it a targeted analysis. Both analyses are implemented in Scikit-Learn, (43) a free and open-source machine learning library for Python (sklearn.decomposition.PCA and sklearn.decomposition.PLSRegression).

CCS Prediction Using Machine Learning

It can be difficult to determine a priori what type of ML algorithm will have optimal performance characteristics for a given task, so a variety of ML models representing a range of algorithmic complexity were examined for use in predicting CCS. All ML models used in this work are implemented in Scikit-Learn. Mean squared error of predictions vs reference values (MSE) was used as the optimization target for model training. ML models tested vary from simple to complex, including multivariate linear regression (sklearn.linear_model.LinearRegression), a linear regression model employing L1 regularization (lasso, sklearn.linear_model.Lasso), a support vector regression model with a radial basis function kernel (svr, sklearn.svm.SVR), and a stochastically assembled ensemble of decision tree models (random forest, sklearn.ensemble.RandomForestRegressor). Description of these chosen models are discussed in the SI.
Before model training, the data set was processed in a stepwise fashion (Scheme S1). First, the data set was split into training and test sets (at proportions of 80% and 20%, respectively), and the test set was put aside until after model training. This data splitting was performed in a stochastic fashion (seeded for deterministic results), with rough stratification on the basis of CCS to ensure a somewhat similar distribution of CCS values between the training and test sets. The training data were then centered and scaled such that each feature would have a mean of 0 and a standard deviation of 1. Such normalization is necessary to ensure numerical stability in training certain predictive models and to avoid arbitrarily overemphasizing features on the basis of their magnitudes. When necessitated by the predictive model being tested (i.e., lasso, svr, random forest), hyperparameter optimization was performed using a grid search with 5-fold cross validation (sklearn.optimize.GridSearchCV) on the training data. Using the optimal hyperparameters, a predictive model was then trained on the training data, and performance metrics (see below) were computed from the training data. Finally, performance metrics were computed using the trained model to make predictions from the test set data. The entire process was repeated multiple times, using different pRNG (pseudorandom number generator) seeds to get different results, and the metrics (for both training and test data) were averaged together from each trial to get an idea of the average predictive model performance under a given set of conditions. Repeated trials were only used for bulk performance comparisons between models or feature sets, but not to actually produce usable individual CCS predictions.

K-Means Clustering for Untargeted Classification of Chemical Structures

Means clustering is a multivariate technique in which data is partitioned into clusters, such that the similarity between samples that are partitioned together is maximized. Briefly, in K-Means clustering, centroids with the same dimensions as the input data are chosen for each cluster, and each sample is assigned to the cluster with the nearest centroid. The centroid positions for the clusters are adjusted (and data partitioning repeated) such that the inertia (sum of squares of Euclidean distances of each sample from their assigned cluster centroid) is minimized. This is an unsupervised classification technique because the class assignment (i.e., partitioning into clusters) is done on the basis of similarity between subgroups within a data set, without using predetermined labels, and therefore offers a means of classifying chemical compounds without the bias inherent to traditional manual assignment. K-Means clustering is implemented in Scikit-Learn (sklearn.cluster.KMeans).

CCS Prediction Performance Metrics

The performance of predictive ML models trained on the combined CCS data set was assessed using an array of metrics, intended to offer a complete representation of model performance. Specifically, R2, mean and median absolute error (MAE and MDAE, respectively), root mean-squared error (RMSE), and cumulative distribution of prediction errors below 1, 3, 5, and 10% (CE135A) were used. Each metric was computed from predicted and reference CCS values (y′ and y, respectively) as follows. R2 is calculated by comparing the residual sum of squares with the variance in the reference values. MAE and MDAE are computed as the mean and median, respectively, of the absolute errors of model predictions. RMSE is computed as the square root of the mean-squared error. CE135A was computed as the proportions of predictions with relative absolute below 1, 3, 5, and 10%.
Although all metrics reflect the accuracy of the regression model in predicting CCS value, each metric specifically highlights a different aspect of model performance. R2 is a good metric for assessing goodness of fit for regression models, but it is only suitable for comparing between models with the same number of parameters. MAE gives a good indication of the average magnitude of errors made by the predictive model, but MDAE may be a better indicator of this magnitude when the distribution of prediction errors is significantly skewed from normal. A drawback of MAE and MDAE is that they can contribute to overfitting by allowing for small numbers of large errors to be easily balanced out by large numbers of small errors. In contrast, RMSE gives an indication of the magnitude of errors but does so in a way that more heavily penalizes large errors (due to the squaring of errors), favoring generalizable models with similar prediction errors across all samples. MAE, MDAE, and RMSE all share the benefit of being computed in the units of the target variable (in the case of CCS prediction, Å2), which can aid in their interpretation with respect to real world performance.

Results and Discussion

Click to copy section linkSection link copied!

Selection of Data Sets for Combined CCS Database

Individual collections of CCS values measured in nitrogen gas included in the combined CCS database were chosen on the basis of size, diversity, and quality of measurements. The combination of multiple data sets, measured in different laboratories and on different instruments, into a single database for training CCS prediction models is desirable from the standpoint of generalizability because any systematic errors/biases present in a single CCS data set can be averaged out by the presence of similar data from other data sets. CCS values measured on both DTIM and TWIM instruments were included in the database in order to maximize the size, breadth, and diversity of the CCS collection. Lipid CCS values measured on TWIM instruments were included only if the CCS was calibrated with lipid standards, which was found to give CCS values predominantly within 2% of DTIM values. (28−30,41) Comparing the CCS values of metabolites and drug and drug-like compounds measured on DTIM and TWIM, only a small percentage of the entries (4 out of 45 in Hines et al. (32) and 7 out of 194 in Hinnenkamp et al. (11)) display errors that are >3%, which justifies the inclusion of values for such compounds measured on both platforms. (22,24,26,32−38) Furthermore, during assembly of the database, we analyzed overlapping values to assess their degree of agreement. Figure 1 shows the number and proportions of overlapping values from the combined CCS database that fall within 1, 3, and 5% difference of one another. In the distribution of all overlapping values (all, n = 695), 46.0, 84.5, and 94.5% fall within 1, 3, and 5% of one another, respectively. This indicates generally good agreement between overlapping values in the database, regardless of the instrument types. Looking specifically at the variation between overlapping DTIM CCS values from different data sets (DT, n = 409), 43.3, 84.4, and 94.1% fall within 1, 3, and 5%, respectively, similar to that of the database as a whole. The variation between overlapping TWIM CCS values (TW, n = 163) is smaller than the overall data set with 65.0, 87.7, and 93.3% falling within 1, 3, and 5%, respectively. Finally, looking specifically at the overlapping values that contain both DTIM and TWIM CCS measurements (DT vs TW, n = 238), 36.6, 75.6, and 93.3% fall within 1, 3, and 5%, respectively, which displays the most variation but still similar to that of the other groups. These results indicate that using both DTIM and TWIM CCS values is unlikely to add additional uncertainty to CCS predictions made by models trained on this combined database relative to models trained on data sets containing only one type of CCS values. Furthermore, the use of the combined database would greatly increase the generalizability of the prediction model because of the broad coverage of the structural diversity. Overall, a total of 7405 CCS entries with SMILES structures (out of 7678 entries in the database at this time) are used for this study, which covers 3526 small molecules (5012 CCS values), 1041 lipids (2345 values), 91 peptides (112 values), and 84 carbohydrates (200 values), many of which have multiple observations from different MS adducts. (22−24,26,28−40)

Figure 1

Figure 1. (A) Counts and (B) comparison of agreement between measurements present in multiple sources. Agreement for all overlapping CCS values in blue. Agreement between DTIM CCS values in red. Agreement between TWIM CCS values in purple. Agreement between DTIM and TWIM CCS values in gold.

High Resolution Image
Download MS PowerPoint Slide

Structural Characterization of the Combined CCS Database

A three-component PCA was performed on all the structures in the combined CCS database using the complete feature set (m/z, OHEA, MQNs) in order to examine the primary contributors to the variance in the data set as a whole in an unsupervised fashion (i.e., entirely ignoring CCS for the time being). This feature set contains molecular descriptors that reflect various structural characteristics of the compounds in the database, thus computing a PCA on these features provides an indication of the breadth of chemical space spanned by this collection, as well as the most important chemical features that differ among them. The three principal axes captured 20.9, 14.4, and 5.7% of the total variance in the data set, respectively. Together, only ∼41% of the total variance was captured in this analysis, indicating that there are many sources of variance in this data set and many of these contributing factors are at least partially orthogonal to one another. Figure 2A–D depicts the projections of the full data set onto all three principal axes, with individual data points colored either by source data set (Figure 2A,B) or rough compound classification (Figure 2C,D). First, examining the projections colored by data set, there is no significant separation between data sets along all principal components. This indicates that the source data set is not strongly associated with the primary attributes of the data set that contribute most significantly to the overall variance; that is, there is overlapping structural diversity among the various data sets. Compounds in this collection were then assigned one of the following rough chemical classes: small molecule, lipid, peptide, or carbohydrate-based on the names of the compounds and information provided in the original publications. Examining PCA projections colored by these class labels, there is some separation along PC1 between small molecules and all other classes, the latter of which only separate from one another along PC2 and PC3. The separation between compound classes along the three principal axes coincides with the separation within each source data set, supporting the notion that the variance observed between different data sets is attributable primarily to the different chemical classes represented in each data set.

Figure 2

Figure 2. PCA projections of full CCS database onto principal axes 1, 2, and 3, colored by data set (A,B) or chemical classification (C,D). Correlation of the top three molecular descriptors contributing to separation along PC1 (E–G) and PC2 (H–J). hac = heavy atom count; m/z = mass to charge ratio; ao = acyclic oxygen count; hbd = H-bond donor atoms; ctv = cyclic trivalent nodes; r6 = 6-membered ring count.

High Resolution Image
Download MS PowerPoint Slide
We next sought to investigate the general chemical trends driving separation along the principal axes, by examining the individual feature loadings. Figures 2E–G depict the values of the top three features contributing to separation along PC1 vs their PC1 projections. The top three features are heavy atom count (count of non-hydrogen atoms, hac), mass-to-charge ratio (m/z), and acyclic oxygen count (ao), all of which are related to overall compound mass. Figures 2H–J depict the values of the corresponding features for PC2: H-bond donor atoms (hbd), cyclic trivalent node count (ctv, related to branching in the chemical structure involving cyclic systems), and 6-membered ring count (r6). The features driving separation along PC2 can be generally described as being related to the composition and topology of a chemical structure (full list of the features can be found in the Supporting Information, Table S2). The corresponding top three features contributing to separation along PC3 can be found in the SI (Figure S1). Taken together, these results indicate that the most significant source of variance in the data set as a whole is related to compound mass, but composition and topology of compound structures are also important contributors.
We next examined what features contribute the most strongly to the distribution of CCS values in the data set as a whole. PLS-RA was performed on all molecules using the complete feature set with CCS as the target variable. Unlike the unsupervised PCA, the first axis in PLS-RA (scores[0]) is chosen such that the most variance in a target variable (CCS) is explained, making it a supervised analysis. The second axis (scores[1]) is orthogonal to the first and explains the most of the remaining variance in the data set. Figure 3A,B show the projections of the full data set onto these axes, with points colored by data set source and rough chemical classification, respectively. Just as with the PCA projections, there does not appear to be a high degree of separation between data sets along scores[0] (Figure 3A), indicating that differences between data sets do not contribute strongly to the variance in CCS values. However, there does appear to be a similar pattern of separation between small molecules and all other classes along scores[0] (Figure 3B), indicating that the chemical characteristics that differ between these chemical classes contribute strongly to variance in CCS values. Given the similarity between the overall patterns observed between source data sets and chemical classes in both PCA and PLS-RA, it appears that the most significant sources of variance within the feature set have strong associations with variance in the target variable, i.e., CCS. Indeed, plotting the projections of the full data set along the most significant axes from PLS-RA and PCA (scores[0] vs PC1, Figure 3C) shows a high degree of correlation, supporting this notion.

Figure 3

Figure 3. PLS-RA projections of full CCS database onto axes 1, 2, colored by data set (A) or chemical classification (B). Correlation between PLS-RA projections along axis 1 and PCA projections along PC1 (C). Correlation between molecular descriptors and PLS-RA projections along axis 1 (blue) or CCS (red) for all compounds (D–K). hac = heavy atom count; m/z = mass to charge ratio; hbam = H-bond acceptor sites; c = carbon atom count; ao = acyclic oxygen count; asb = acyclic single bonds; asv = acyclic single valent nodes; adb = acyclic double bonds.

High Resolution Image
Download MS PowerPoint Slide
As with PCA, the contribution of individual features to separation along a given axis in PLS-RA can be investigated by examining the feature loadings. Figure 3D–K depicts the values of the top eight features that contribute to separation along the primary axis from PLS-RA vs scores[0] (shown in blue) or CCS (shown in red). The most significant contributors to variance in CCS are hac and m/z (Figure 3D,E, respectively), both of which are mass-related molecular descriptors. The remainder of the top features (Figure 3F–K) that show good correlations with scores[0] and CCS include features that are related to compound size (H-bond acceptor sites, hbam; carbon count, c; ao) as well as structure topology (acyclic single bonds, asb; acyclic monovalent nodes, asv; acyclic double bonds, adb). Much like the results from PCA, PLS-RA indicates that the most important features in the data set that contribute to variance in the CCS are related to compound mass and size, also with significant contributions from features describing molecular topology.

Model Specialization through Unsupervised Classification

Current examples of CCS prediction models were built using collections of measurements made for specific chemical classes, e.g. lipids or small molecules (i.e., for metabolomics). (22−26) Such model specialization is justifiable from the standpoint that the chemical characteristics that contribute to a compound’s CCS are likely to differ between chemical classes, and therefore, models trained to recognize the important characteristics of one chemical class is likely to perform poorly in producing predictions for other classes. Indeed, performing CCS predictions with LipidCCS Predictor on all compounds from the combined database tagged as lipids produced an RMSE of 5.5 Å2, while predictions on compounds tagged as peptides and carbohydrates produced RMSE scores of 46.9 and 56.3 Å2, respectively (Supporting Figure S2). Employing such specialized models is an excellent approach for attaining high prediction accuracy, as long as the delineation between chemical classes is clear. However, assigning chemical class labels in an unbiased way can become difficult when considering the diverse small molecule space, which encompasses complex compounds containing substructures with chemical characteristics resembling multiple conventional chemical classes. Furthermore, most ML algorithms act as “black boxes” with respect to the underlying characteristics of the training data that ultimately contribute to predictions. Even for ML algorithms that provide some insight into which features contribute most strongly (e.g., lasso, forest), there is still no way to tell whether and how subgroups within the training data influence this process. To address these limitations, rather than relying on manually assigned chemical classes, we performed unbiased and unsupervised classification using the K-Means clustering ML algorithm on all molecules with SMILES structures in the database to determine the most prominent groupings of compounds with respect to structural characteristics.
Using the full set of MDs (m/z, OHEA, and all MQNs) as features, K-Means clustering was used to group compounds on the basis of their structural similarity. During initial testing, fitting with four clusters was found to be optimal for capturing the major groups within the data set while maintaining sufficient numbers in each cluster for model training (at least 100 molecules per cluster). Figure 4A,B show the PCA projections of the full data set colored by rough chemical classes, and Figure 4C,D show the same PCA projection colored by clusters. The clustering analysis revealed similar groupings to the chemical class labels: most of the small molecules were assigned to cluster 1 and 2 (purple and blue, respectively), lipids to cluster 3 (magenta), and carbohydrates and peptides were assigned to cluster 4 (gold). However, a significant number of molecules in the original rough chemical classification were reassigned to new clusters, suggesting that manual assignment of chemical classes could result in classification error. As seen in the PCA plot (Figure 4C), the two small molecule clusters separate from the lipid and carbohydrate/peptide clusters along PC1, and separation within these groups occurs primarily along PC2. With the separation along PC1 being mostly related to compound mass, it appears that the separation between the small molecule clusters and the lipids and peptides/carbohydrate clusters is primarily mass-driven. Separation along PC2 is related more to topological differences in chemical structure, meaning that other factors such as branching or presence of rings drives the separation between the two small molecule clusters, and between the lipid and peptide/carbohydrate clusters, respectively. Interestingly, the assigned clusters conform to familiar trends in the IM-MS conformational space, although overlapping between clusters still exists (Figure 4E). Broadly speaking, the lipid cluster occupies the high-CCS and high-mass region, while the peptide/carbohydrate cluster occupies the low-CCS high-mass region. The small molecule clusters occupy significant overlapping space in the low-mass region, each covering the broad range of CCS values associated with this region. The most central structures within each cluster (minimum distance to the cluster centers in feature space) are presented in Figure 4F. PE(36:3) is a prototypical lipid, while CMP-N-acetylneruaminic acid has chemical characteristics related to both carbohydrates and peptides (large, containing many heteratoms). The small molecules etodolac and 3-methoxytyrosine represent the central structures within the two small molecule clusters and, as expected from the PCA, they seem to differ mostly on a topological basis (e.g., rings and branching) rather than by size or composition. Collectively, these results show that use of an unsupervised clustering approach recapitulates many of the classifications attainable through a manual approach, while also providing class separation based on more nuanced chemical characteristics. Importantly, such an approach allows for unbiased assignment of chemical class when considering new molecules, especially those that do not fall cleanly into a single conventional chemical class.

Figure 4

Figure 4. PCA projections of full CCS database onto principal axes 1, 2, and 3, colored by chemical class label (A, B), or by cluster (C, D). (E) Plot of CCS vs m/z for full CCS database, colored by cluster. (F) Central structures within each cluster. (G–I) Average predictive performance of models (lasso, forest, svr, respectively) by MDAE and RMSE from five independent trials, trained on the full CCS database (training set = blue, test set = red) or on individual cluster data sets (training set = purple, test set = gold).

High Resolution Image
Download MS PowerPoint Slide
Within each cluster, individual ML models were trained using the complete feature set, and the average performance of this ensemble of models was compared to corresponding individual models trained on the complete data set (Figure 4G–I). Feature selection is commonly performed in ML projects as it promotes generalizability of a trained model; however, we did not observe significant differences in performance between using the complete feature set and using selected features following a previous practice (see Figure S3 and associated discussion in SI). (23) This is likely due to the already small number of structure-related MQNs as MDs (compare with the hundreds of MDs typically used in the literature) and the intrinsic feature selection process when training with models like Lasso, random forest, and svr. Training individual models specific to each cluster led to a marked performance increase by MDAE and RMSE for the lasso models and to a lesser degree for the svr models (panels G and I, respectively, of Figure 4).
The performance by MDAE did not increase with the addition of clustering for the forest models. However, the RMSE did decrease, indicating that the use of clustering reduced the proportion of higher-magnitude errors for these models (Figure 4H). Across all models (with or without clustering), the average performance by both metrics for the test set predictions was at parity with the training set performance, indicating good generalizability with respect to making predictions on new data. These results demonstrate that the inclusion of untargeted clustering followed by building individual predictive CCS models using each cluster can increase the model performance and that such an approach is well suited for application to large compound collections covering diverse chemical space. Additionally, this approach offers the benefit of interpretability at the classification level: the assignment of compounds to individual clusters provides information on the common chemical characteristics that define chemical classes in an unbiased fashion.

Training and Performance Characteristics of the Final Optimized Prediction Model

On the basis of the insights gained from the above work, we built a final deployable predictive CCS model using K-Means clustering and four individual svr models with radial basis function kernels trained on each of the fitted clusters. Figure 5 describes the final workflow for building and training this model, and Figure 6 summarizes all of the performance characteristics of the model on both the training and test set data. The R2 scores for training and test set data were 0.994 and 0.991, respectively (Figure 6A), indicating excellent generalizability. Judging from additional metrics, this model was able to achieve high performance on the training data, with MAE, MDAE, and RMSE scores of 2.92, 1.70, and 5.48 Å2, respectively (Figure 6B, blue). The generalizability was also excellent based on the corresponding test set scores of 3.83, 2.37, and 6.46 Å2, respectively, which show no significant performance lapse relative to the training data (Figure 6B, red). The cumulative error distributions of CCS predictions give a more detailed indication of the error structure of the CCS predictions, with 56.1 and 86.4% of CCS predictions for the training data falling within 1 and 3% of the reference values, respectively (Figure 6C, blue). The model achieves similar performance on the test set data, with 44.7 and 81.2% of CCS predictions falling within 1 and 3% of the reference values, respectively (Figure 6C, red). Taken together, these performance metrics indicate that this final model is capable of performing CCS prediction with high accuracy on diverse compounds and that this performance is robust when applied to unseen data.

Figure 5

Figure 5. Workflow describing the process for training and validating the final prediction model. First, MQNs are generated from the compound SMILES structure in addition to the m/z and MS adduct, and this data is stored in a database. The complete data set is randomly partitioned into a training set and test set, preserving the approximate distribution of CCS values between the two sets. The training set is then fit using K-Means clustering to find the dominant groupings within the data set in terms of chemical similarity. The data from each assigned cluster is then used to train an individual predictive model that is specialized for that group of compounds. Finally, the overall CCS prediction performance of this set of models is validated using the test set data by first assigning each sample to one of the fitted clusters then predicting CCS using the corresponding predictive model.

High Resolution Image
Download MS PowerPoint Slide

Figure 6

Figure 6. (A–C) Complete performance metrics for final predictive model on training (blue) and test (red) data. (A) R2 (B) mean/median absolute error and root mean squared error (C) proportion of predictions falling within 1, 3, 5, and 10% of reference values. (D–F) Comparison of CCS prediction performance between final model (purple) and DeepCCS (gold) on all data sets used for training DeepCCS. (D) R2 (E) mean/median absolute error and root mean squared error (F) mean/median relative error (MRE and MDRE).

High Resolution Image
Download MS PowerPoint Slide
We next compared our final model with DeepCCS, the other comprehensive CCS prediction model using a CNN trained directly on SMILES structures, albeit with limited coverage of small molecules. A comparison data set was assembled from the data sets used in the training and testing of DeepCCS (2298 compounds after selection of valid SMILES and MS adducts). Using this comparison data set, 1960 CCS values were predicted by DeepCCS, and the accuracy of these predictions was compared with our model by a variety of metrics (Figure 6D–F). The final prediction model presented here outperformed DeepCCS by all performance metrics, despite using far fewer parameters to fit the data. This higher CCS prediction performance is likely attributable to two reasons. First, the larger and more diverse collection of data used to train the model enables greater accuracy in predictions by learning more robust and generalizable trends from the training data. Second, the use of untargeted clustering to partition the data on the basis of common structural features allows this model to learn specific trends for different classes of chemicals, thus increasing the overall accuracy of CCS predictions through model specialization.

Building an All-in-One Web Interface for Querying the CCS Database and Accessing the Prediction Model

To increase the accessibility of the database and the prediction model to the field, we have assembled the combined CCS database and the final prediction model into a convenient web interface (https://CCSbase.net). This interface allows users to query the database for reference CCS values with fine filtering control, including name, mass with accuracy, CCS with accuracy, polarity, SMILES structures, adduct types, charge state, and fuzzy search. This easily accessible database offers complementary and broader coverage (7678 entries) in comparison with the existing CCS databases, (35,38,44,45) including the carefully assembled CCS compendium on DTIM CCS values (3833 values). (44) This interface also allows rapid prediction of CCS values (with confident prediction for six different MS adducts) directly from SMILES structures using the final cluster-based prediction model discussed above. Batch query and prediction can be achieved with a simple single CSV file input. All results are directly viewable on the interface in the form of a table and in an interactive CCS-mass plot with the main trendline of the entire database being the background. The results are also downloadable as a CSV file. This platform can be easily built into existing metabolomics workflows and thus serve as a useful tool in the identification of unknowns from large-scale untargeted analyses.
The primary utility of the CCS database and predictive model are to support unknown compound identification by CCS. The metabolomics standards initiative (MSI) (46) defines 4 annotation levels reflecting the rigor of compound identification. Level 1 annotations are obtained from matching at least two orthogonal properties, such as m/z, CCS, or MS/MS, against values determined experimentally from authentic standards. Level 2 annotations are the same, with the exception that the reference measured values are taken from a secondary source (like the literature). Level 3 annotations are obtained when reference measurements are not directly available for a compound, but can be inferred based on existing measurements from similar compounds. Level 4 annotations correspond to compounds that are unidentified, but can be differentiated from other signals. Identifications based on measured values from the CCS database would therefore constitute level 2 annotations, i.e., using measured m/z and CCS as orthogonal properties, while those made using CCS values generated using the predictive model would constitute level 3 annotations (i.e., using m/z and predicted CCS values).

Conclusions

Click to copy section linkSection link copied!
The ability to predict high-quality CCS values for unknowns is a key step toward using CCS as a broadly applicable identifier for metabolites. Several major advances in ML-based CCS prediction are achieved in this work. First, by assembling the largest CCS database to date, a broad coverage of chemical structural diversity is achieved. Second, by performing statistical analysis of the structural features of all compounds, we have identified structural features that display high correlation with CCS values (Figure 3). Such correlation has not been systemically examined previously. Third, the use of structure-related MQNs as MDs is more relevant to the structure-dependent CCS than MDs reflecting the physical properties that were often used in the literature. Fourth, by breaking down the structural diversity using structure-based unsupervised clustering in combination with individual prediction models, the integrated model displays greatly improved performance and generalizability than without clustering. Importantly, this model also provides interpretable results based on the cluster that the unknown is assigned to, unlike previous work mostly using “black box” prediction models. Finally, we have built an easily accessible all-in-one web interface for efficient querying of the database and the prediction model. We anticipate continuous growth of this database and improvement of the prediction model as molecules with additional structural features are added.

Supporting Information

Click to copy section linkSection link copied!

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.9b05772.

  • All code used to assemble the CCS database and train predictive models is available on GitHub (https://github.com/dylanhross/c3sdb); SI includes Experimental Section and schema; MS adduct encodings, molecular quantum numbers (MQNs), top three features contributing to separation along PC3, CCS prediction accuracy of LipidCCS on different chemical classes, and feature selection trials; and additional Results and Discussion (PDF)

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

Click to copy section linkSection link copied!
  • Corresponding Author
  • Authors
    • Dylan H. Ross - Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195, United States
    • Jang Ho Cho - Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195, United States
  • Author Contributions

    The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

  • Notes
    The authors declare no competing financial interest.

Acknowledgments

Click to copy section linkSection link copied!

This work was supported by the Drug Metabolism Transport and Pharmacogenetics Research Fund of the School of Pharmacy at the University of Washington (UW), UW CoMotion Innovation Gap Fund, and startup funds from the Department of Medicinal Chemistry to L.X.

References

Click to copy section linkSection link copied!

This article references 46 other publications.

  1. 1
    Prakash, C.; Shaffer, C. L.; Nedderman, A. Mass Spectrom. Rev. 2007, 26, 340– 369,  DOI: 10.1002/mas.20128
    Google Scholar
    1
    Analytical strategies for identifying drug metabolites
    Prakash, Chandra; Shaffer, Christopher L.; Nedderman, Angus
    Mass Spectrometry Reviews (2007), 26 (3), 340-369CODEN: MSRVD3; ISSN:0277-7037. (John Wiley & Sons, Inc.)
    A review. With the dramatic increase in the no. of new chem. entities (NCEs) arising from combinatorial chem. and modern high-throughput bioassays, novel bioanal. techniques are required for the rapid detn. of the metabolic stability and metabolites of these NCEs. Knowledge of the metabolic site(s) of the NCEs in early drug discovery is essential for selecting compds. with favorable pharmacokinetic credentials and aiding medicinal chemists in modifying metabolic "soft spots". In development, elucidation of biotransformation pathways of a drug candidate by identifying its circulatory and excretory metabolites is vitally important to understand its physiol. effects. Mass spectrometry (MS) and NMR have played an invaluable role in the structural characterization and quantification of drug metabolites. Indeed, liq. chromatog. (LC) coupled with atm. pressure ionization (API) MS has now become the most powerful tool for the rapid detection, structure elucidation, and quantification of drug-derived material within various biol. fluids. Often, however, MS alone is insufficient to identify the exact position of oxidn., to differentiate isomers, or to provide the precise structure of unusual and/or unstable metabolites. In addn., an excess of endogenous material in biol. samples often suppress the ionization of drug-related material complicating metabolite identification by MS. In these cases, multiple anal. and wet chem. techniques, such as LC-NMR, enzymic hydrolysis, chem. derivatization, and hydrogen/deuterium-exchange (H/D-exchange) combined with MS are used to characterize the novel and isomeric metabolites of drug candidates. This review describes sample prepn. and introduction strategies to minimize ion suppression by biol. matrixes for metabolite identification studies, the application of various LC-tandem MS (LC-MS/MS) techniques for the rapid quantification and identification of drug metabolites, and future trends in this field.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXlvFCgu7Y%253D&md5=0bad5d31fb5d2252e9492253c9a0d172
  2. 2
    Blazenovic, I.; Kind, T.; Ji, J.; Fiehn, O. Metabolites 2018, 8, 31,  DOI: 10.3390/metabo8020031
    Google Scholar
    2
    Software tools and approaches for compound identification of LC-MS/MS data in metabolomics
    Blazenovic, Ivana; Kind, Tobias; Ji, Jian; Fiehn, Oliver
    Metabolites (2018), 8 (2), 31/1-31/23CODEN: METALU; ISSN:2218-1989. (MDPI AG)
    The annotation of small mols. remains a major challenge in untargeted mass spectrometry-based metabolomics. We here critically discuss structured elucidation approaches and software that are designed to help during the annotation of unknown compds. Only by elucidating unknown metabolites first is it possible to biol. interpret complex systems, to map compds. to pathways and to create reliable predictive metabolic models for translational and clin. research. These strategies include the construction and quality of tandem mass spectral databases such as the coalition of MassBank repositories and investigations of MS/MS matching confidence. We present in silico fragmentation tools such as MS-FINDER, CFM-ID, MetFrag, ChemDistiller and CSI:FingerID that can annotate compds. from existing structure databases and that have been used in the CASMI (crit. assessment of small mol. identification) contests. Furthermore, the use of retention time models from liq. chromatog. and the utility of collision cross-section modeling from ion mobility expts. are covered. Workflows and published examples of successfully annotated unknown compds. are included.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhvV2lt73I&md5=371fd01734c223fb65560768722c830c
  3. 3
    Clemmer, D. E.; Hudgins, R. R.; Jarrold, M. F. J. Am. Chem. Soc. 1995, 117, 10141– 10142,  DOI: 10.1021/ja00145a037
    Google Scholar
    3
    Naked Protein Conformations: Cytochrome c in the Gas Phase
    Clemmer, David E.; Hudgins, Robert R.; Jarrold, Martin F.
    Journal of the American Chemical Society (1995), 117 (40), 10141-2CODEN: JACSAT; ISSN:0002-7863. (American Chemical Society)
    The conformers present for naked cytochrome c ions in the gas phase were examd. using electrospray/ion mobility measurements. A no. of resolved features that correspond to distinct partially folded conformations were obsd. The abundance of the native conformer was negligible under the conditions employed. The obsd. conformers may be intermediates in the folding process.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXotVyqsL8%253D&md5=6f11c3d8b73925c5c6ff0ad475bd8a86
  4. 4
    von Helden, G.; Wyttenbach, T.; Bowers, M. T. Science 1995, 267, 1483– 1485,  DOI: 10.1126/science.267.5203.1483
    Google Scholar
    4
    Conformation of macromolecules in the gas phase: use of matrix-assisted laser desorption methods in ion chromatography
    Von Helden, Gert; Wyttenbach, Thomas; Bowers, Michael T.
    Science (Washington, D. C.) (1995), 267 (5203), 1483-5CODEN: SCIEAS; ISSN:0036-8075. (American Association for the Advancement of Science)
    Conformational data of macromols. in the gas phase were obtained by the coupling of matrix-assisted laser desorption ion source to an ion chromatograph. A series of polyethylene glycols (PEG), cationized by sodium ions were studied, in comparison with protonated bradykinin. Detailed modeling of Na+PEG9 with mol. mechanics methods indicates that the lowest energy structure has the Na+ ion "solvated" by the polymer chain with seven oxygen atoms as nearest neighbors. The agreement between the model and expt. is within 1% for Na+PEG9, Na+PEG13, and Na+PEG17, which strongly supported both the method and the deduced structures.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXktlCltL4%253D&md5=6f8d9b91958b623e95eb2a121b7deeb9
  5. 5
    McLean, J. A.; Ruotolo, B. T.; Gillig, K. J.; Russell, D. H. Int. J. Mass Spectrom. 2005, 240, 301– 315,  DOI: 10.1016/j.ijms.2004.10.003
    Google Scholar
    5
    Ion mobility-mass spectrometry: a new paradigm for proteomics
    McLean, John A.; Ruotolo, Brandon T.; Gillig, Kent J.; Russell, David H.
    International Journal of Mass Spectrometry (2005), 240 (3), 301-315CODEN: IMSPF8; ISSN:1387-3806. (Elsevier B.V.)
    A review. Matrix-assisted laser desorption/ionization (MALDI) coupled with ion mobility-mass spectrometry (IM-MS) provides a rapid (μs-ms) means for the 2-dimensional (2D) sepn. of complex biol. samples (e.g., peptides, oligonucleotides, glycoconjugates, lipids, etc.), elucidation of solvent-free secondary structural elements (e.g., helixes, β-hairpins, random coils, etc.), rapid identification of post-translational modifications (e.g., phosphorylation, glycosylation, etc.) or ligation of small mols., and simultaneous and comprehensive sequencing information of biopolymers. In IM-MS, protein-identification information is complemented by structural characterization data, which is difficult to obtain using conventional proteomic techniques. New avenues for enhancing the figures of merit (e.g., sensitivity, limits of detection, dynamic range, and analyte selectivity) and optimizing IM-MS exptl. parameters are described in the context of deriving new information at the forefront of proteomics research.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXktF2rsA%253D%253D&md5=2d6eb1cf9e7418b1203a61ab094adc98
  6. 6
    Kanu, B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H. H., Jr. J. Mass Spectrom. 2008, 43, 1– 22,  DOI: 10.1002/jms.1383
    Google Scholar
    6
    Ion mobility-mass spectrometry
    Kanu, Abu B.; Dwivedi, Prabha; Tam, Maggie; Matz, Laura; Hill, Herbert H., Jr.
    Journal of Mass Spectrometry (2008), 43 (1), 1-22CODEN: JMSPFJ; ISSN:1076-5174. (John Wiley & Sons Ltd.)
    This review article compares and contrasts various types of ion mobility-mass spectrometers available today and describes their advantages for application to a wide range of analytes. Ion mobility spectrometry (IMS), when coupled with mass spectrometry, offers value-added data not possible from mass spectra alone. Sepn. of isomers, isobars, and conformers; redn. of chem. noise; and measurement of ion size are possible with the addn. of ion mobility cells to mass spectrometers. In addn., structurally similar ions and ions of the same charge state can be sepd. into families of ions which appear along a unique mass-mobility correlation line. This review describes the four methods of ion mobility sepn. currently used with mass spectrometry. They are (1) drift-time ion mobility spectrometry (DTIMS), (2) aspiration ion mobility spectrometry (AIMS), (3) differential-mobility spectrometry (DMS) which is also called field-asym. waveform ion mobility spectrometry (FAIMS) and (4) traveling-wave ion mobility spectrometry (TWIMS). DTIMS provides the highest IMS resolving power and is the only IMS method which can directly measure collision cross-sections. AIMS is a low resoln. mobility sepn. method but can monitor ions in a continuous manner. DMS and FAIMS offer continuous-ion monitoring capability as well as orthogonal ion mobility sepn. in which high-sepn. selectivity can be achieved. TWIMS is a novel method of IMS with a low resolving power but has good sensitivity and is well integrated into a com. mass spectrometer. One hundred and sixty refs. on ion mobility-mass spectrometry (IMMS) are provided.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXisFKntbk%253D&md5=44236351655e694207187c94eeca82e6
  7. 7
    Fenn, L. S.; Kliman, M.; Mahsut, A.; Zhao, S. R.; McLean, J. A. Anal. Bioanal. Chem. 2009, 394, 235– 244,  DOI: 10.1007/s00216-009-2666-3
    Google Scholar
    7
    Characterizing ion mobility-mass spectrometry conformation space for the analysis of complex biological samples
    Fenn, Larissa S.; Kliman, Michal; Mahsut, Ablatt; Zhao, Sophie R.; McLean, John A.
    Analytical and Bioanalytical Chemistry (2009), 394 (1), 235-244CODEN: ABCNBP; ISSN:1618-2642. (Springer)
    The conformation space occupied by different classes of biomols. measured by ion mobility-mass spectrometry (IM-MS) is described for utility in the characterization of complex biol. samples. Although the qual. sepn. of different classes of biomols. on the basis of structure or collision cross section is known, there is relatively little quant. cross-section information available for species apart from peptides. In this report, collision cross sections are measured for a large suite of biol. salient species, including oligonucleotides (n = 96), carbohydrates (n = 192), and lipids (n = 53), which are compared to reported values for peptides (n = 610). In general, signals for each class are highly correlated, and at a given mass, these correlations result in predicted collision cross sections that increase in the order oligonucleotides < carbohydrates < peptides < lipids. The specific correlations are described by logarithmic regressions, which best approx. the theor. trend of increasing collision cross section as a function of increasing mass. A statistical treatment of the signals obsd. within each mol. class suggests that the breadth of conformation space occupied by each class increases in the order lipids < oligonucleotides < peptides < carbohydrates. The utility of conformation space anal. in the direct anal. of complex biol. samples is described, both in the context of qual. mol. class identification and in fine structure examn. within a class. The latter is demonstrated in IM-MS sepns. of isobaric oligonucleotides, which are interpreted by mol. dynamics simulations.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXisVWqsro%253D&md5=febdb07ff7e0cfb10c0e9b5f22d5fa58
  8. 8
    Pringle, S. D.; Giles, K.; Wildgoose, J. L.; Williams, J. P.; Slade, S. E.; Thalassinos, K.; Bateman, R. H.; Bowers, M. T.; Scrivens, J. H. Int. J. Mass Spectrom. 2007, 261, 1– 12,  DOI: 10.1016/j.ijms.2006.07.021
    Google Scholar
    8
    An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/travelling wave IMS/oa-ToF instrument
    Pringle, Steven D.; Giles, Kevin; Wildgoose, Jason L.; Williams, Jonathan P.; Slade, Susan E.; Thalassinos, Konstantinos; Bateman, Robert H.; Bowers, Michael T.; Scrivens, James H.
    International Journal of Mass Spectrometry (2007), 261 (1), 1-12CODEN: IMSPF8; ISSN:1387-3806. (Elsevier B.V.)
    Ion mobility coupled with mass spectrometry has evolved into a powerful anal. technique for investigating the gas-phase structures of bio-mols. Here the authors present the mobility sepn. of some peptide and protein ions using a new hybrid quadrupole/travelling wave ion mobility separator/orthogonal acceleration time-of-flight instrument. Comparison of the mobility data obtained from the relatively new travelling wave sepn. device with data obtained using various other mobility separators demonstrate that while the mobility characteristics are similar, the new hybrid instrument geometry provides mobility sepn. without compromising the base sensitivity of the mass spectrometer. This capability facilitates mobility studies of samples at anal. significant levels.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhsFCru7c%253D&md5=1bcf7eba693142d9534f16ec8e978ec7
  9. 9
    May, J. C.; McLean, J. A. Anal. Chem. 2015, 87, 1422– 1436,  DOI: 10.1021/ac504720m
    Google Scholar
    9
    Ion Mobility-Mass Spectrometry: Time-Dispersive Instrumentation
    May, Jody C.; McLean, John A.
    Analytical Chemistry (Washington, DC, United States) (2015), 87 (3), 1422-1436CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    A review. The field of ion mobility-mass spectrometry (IM-MS) has grown with significant momentum in recent years in both fundamental advances and pioneering applications. A search of the terms "ion mobility" and "mass spectrometry" returns more than 2,000 papers, with over half of these being published in the past four years. This increased interest has been motivated in large part by improved technologies which have enabled contemporary IM-MS to be amendable to a variety of samples in biol. and medicine with high sensitivity, resolving power, and sample throughput.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXitFehs73P&md5=43635f295a34d972fde1f3fbe0cf4b83
  10. 10
    Mesleh, M. F.; Hunter, J. M.; Shvartsburg, A. A.; Schatz, G. C.; Jarrold, M. F. J. Phys. Chem. 1996, 100, 16082– 16086,  DOI: 10.1021/jp961623v
    Google Scholar
    10
    Structural Information from Ion Mobility Measurements: Effects of the Long-Range Potential
    Mesleh, M. F.; Hunter, J. M.; Shvartsburg, A. A.; Schatz, G. C.; Jarrold, M. F.
    Journal of Physical Chemistry (1996), 100 (40), 16082-16086CODEN: JPCHAX; ISSN:0022-3654. (American Chemical Society)
    In a no. of recent studies, information about the structure of large polyat. ions has been deduced from gas-phase ion mobility measurements by comparing mobilities measured in helium to those estd. for assumed geometries using a hard-sphere projection approxn. To examine the validity of this approach, we have compared mobilities calcd. using the hard-sphere projection approxn. for a range of fullerenes (C20-C240) to those detd. from trajectory calcns. with a more realistic He-fullerene potential. The He-fullerene potential we have employed, a sum of two-body 6-12 interactions plus a sum of ion-induced dipole interactions, was calibrated using the measured mobility of C60+ in helium over an 80-380 K temp. range. For the systems studied, the long-range interactions between the ion and buffer gas have a small, less than 10%, effect on the calcd. mobility at room temp. However, the effects are not insignificant, and in many cases it will be necessary to consider the long-range interactions if the correct structural assignments are to be made from measured ion mobilities.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XlsFCqu7w%253D&md5=d161824e4f762fb750c196d6270476d4
  11. 11
    Hinnenkamp, V.; Klein, J.; Meckelmann, S. W.; Balsaa, P.; Schmidt, T. C.; Schmitz, O. J. Anal. Chem. 2018, 90, 12042– 12050,  DOI: 10.1021/acs.analchem.8b02711
    Google Scholar
    11
    Comparison of CCS Values Determined by Traveling Wave Ion Mobility Mass Spectrometry and Drift Tube Ion Mobility Mass Spectrometry
    Hinnenkamp, Vanessa; Klein, Julia; Meckelmann, Sven W.; Balsaa, Peter; Schmidt, Torsten C.; Schmitz, Oliver J.
    Analytical Chemistry (Washington, DC, United States) (2018), 90 (20), 12042-12050CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Collision cross section (CCS, Ω) values detd. by ion mobility mass spectrometry (IM-MS) provide the study of ion shape in the gas phase and use of these as further identification criteria in anal. approaches. Databases of CCS values for a variety of mols. detd. by different instrument types are available. The comparability of CCS values detd. by a drift tube ion mobility mass spectrometer (DTIM-MS) and a traveling wave ion mobility mass spectrometer (TWIM-MS) was studied to test if a common database could be used across IM techniques. A total of 124 substances were measured with both systems and CCS values of [M + H]+ and [M + Na]+ adducts were compared. Deviations <1% were found for most substances, but some compds. show deviations up to 6.2%, which indicate that CCS databases cannot be used without care independently from the instrument type. Addnl., for several mols. [2M + Na]+ ions were formed during electrospray ionization, whereas a part of them disintegrates to [M + Na]+ ions after passing through the drift tube and before reaching the TOF region, resulting in 2 signals in their drift spectrum for the [M + Na]+ adduct. Finally, the impact of different LC-IM-MS settings (solvent compn., solvent flow rate, desolvation temp., and desolvation gas flow rate) were studied to test whether they have an influence on the CCS values or not. These conditions have no significant impact. Only for karbutilate changes in the drift spectrum could be obsd. with different solvent types and flow rates using the DTIM-MS system, which could be caused by the protonation at different sites in the mol.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhslaitbzN&md5=0f957ee087e93729d70c72d4df1f3347
  12. 12
    Stow, S. M.; Causon, T. J.; Zheng, X.; Kurulugama, R. T.; Mairinger, T.; May, J. C.; Rennie, E. E.; Baker, E. S.; Smith, R. D.; McLean, J. A.; Hann, S.; Fjeldsted, J. C. Anal. Chem. 2017, 89, 9048– 9055,  DOI: 10.1021/acs.analchem.7b01729
    Google Scholar
    12
    An Interlaboratory Evaluation of Drift Tube Ion Mobility-Mass Spectrometry Collision Cross Section Measurements
    Stow, Sarah M.; Causon, Tim J.; Zheng, Xueyun; Kurulugama, Ruwan T.; Mairinger, Teresa; May, Jody C.; Rennie, Emma E.; Baker, Erin S.; Smith, Richard D.; McLean, John A.; Hann, Stephan; Fjeldsted, John C.
    Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9048-9055CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Collision cross section (CCS) measurements resulting from ion mobility-mass spectrometry (IM-MS) expts. provide a promising orthogonal dimension of structural information in MS-based anal. sepns. As with any mol. identifier, interlab. standardization must precede broad range integration into anal. workflows. The authors present a ref. drift tube ion mobility mass spectrometer (DTIM-MS) where improvements on the measurement accuracy of exptl. parameters influencing IM sepns. provide standardized drift tube, nitrogen CCS values (DTCCSN2) for over 120 unique ion species with the lowest measurement uncertainty to date. The reproducibility of these DTCCSN2 values are evaluated across three addnl. labs. on a com. available DTIM-MS instrument. The traditional stepped field CCS method performs with a relative std. deviation of 0.29% for all ion species across the three addnl. labs. The calibrated single field CCS method, which is compatible with a wide range of chromatog. inlet systems, performs with an av., abs. bias of 0.54% to the standardized stepped field DTCCSN2 values on the ref. system. The low relative std. deviation and biases obsd. in this interlab. study illustrate the potential of DTIM-MS for providing a mol. identifier for a broad range of discovery based analyses.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhtbnL&md5=9081cdf22f75275dcb5f209741e49f9f
  13. 13
    Shvartsburg, A.; Jarrold, M. F. Chem. Phys. Lett. 1996, 261, 86– 91,  DOI: 10.1016/0009-2614(96)00941-4
    Google Scholar
    13
    An exact hard-spheres scattering model for the mobilities of polyatomic ions
    Shvartsburg, Alexandre A.; Jarrold, Martin F.
    Chemical Physics Letters (1996), 261 (1,2), 86-91CODEN: CHPLBC; ISSN:0009-2614. (Elsevier)
    We describe an exact hard-spheres scattering model for calcg. the gas-phase mobilities of polyat. ions. Ion mobility measurements have recently been used to deduce structural information for clusters and biomols. in the gas phase. In virtually all of the previous ion mobility studies, mobilities were evaluated for comparison with the exptl. data using a projection approxn. Comparison of the collision integrals calcd. using the exact hard-spheres scattering model with those estd. using the projection approxn. shows that large deviations, over 20%, occur for some geometries with grossly concave surfaces.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XlvFKnsbk%253D&md5=ee35ea2bf473e9e83188f21ada9d37ae
  14. 14
    Kim, H. I.; Kim, H.; Pang, E. S.; Ryu, E. K.; Beegle, L. W.; Loo, J. A.; Goddard, W. A.; Kanik, I. Anal. Chem. 2009, 81, 8289– 8297,  DOI: 10.1021/ac900672a
    Google Scholar
    14
    Structural Characterization of Unsaturated Phosphatidylcholines Using Traveling Wave Ion Mobility Spectrometry
    Kim, Hugh I.; Kim, Hyungjun; Pang, Eric S.; Ryu, Ernest K.; Beegle, Luther W.; Loo, Joseph A.; Goddard, William A.; Kanik, Isik
    Analytical Chemistry (Washington, DC, United States) (2009), 81 (20), 8289-8297CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    A no. of phosphatidylcholine (PC) cations spanning a mass range of 400-1000 Da are investigated using electrospray ionization mass spectrometry coupled with traveling wave ion mobility spectrometry (TWIMS). A high correlation between mass and mobility is demonstrated with satd. phosphatidylcholine cations in N2. A significant deviation from this mass-mobility correlation line is obsd. for the unsatd. PC cation. The authors found that the double bond in the acyl chain causes a 5% redn. in drift time. The drift time is reduced at a rate of ∼1% for each addnl. double bond. Theor. collision cross sections of PC cations exhibit good agreement with exptl. evaluated values. Collision cross sections are detd. using the recently derived relation between mobility and drift time in TWIMS stacked ring ion guide (SRIG) and compared to estd. collision cross sections using an empiric calibration method. Computational anal. was performed using the modified trajectory (TJ) method with nonspherical N2 mols. as the drift gas. The difference between estd. collision cross sections and theor. collision cross sections of PC cations is related to the sensitivity of the PC cation collision cross sections to the details of the ion-neutral interactions. The origin of the obsd. correlation and deviation between mass and mobility of PC cations is discussed in terms of the structural rigidity of these mols. using mol. dynamic simulations.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXhtFGlsLrL&md5=79ffad12c1e053d5450955fbccca9fdf
  15. 15
    Kim, H.; Kim, H. I.; Johnson, P. V.; Beegle, L. W.; Beauchamp, J. L.; Goddard, W. A.; Kanik, I. Anal. Chem. 2008, 80, 1928,  DOI: 10.1021/ac701888e
    Google Scholar
    15
    Experimental and Theoretical Investigation into the Correlation between Mass and Ion Mobility for Choline and Other Ammonium Cations in N2
    Kim, Hyungjun; Kim, Hugh I.; Johnson, Paul V.; Beegle, Luther W.; Beauchamp, J. L.; Goddard, William A.; Kanik, Isik
    Analytical Chemistry (Washington, DC, United States) (2008), 80 (6), 1928-1936CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    A no. of tertiary amine and quaternary ammonium cations spanning a mass range of 60-146 amu (trimethylamine, tetramethylammonium, trimethylethylammonium, N,N-dimethylaminoethanol, choline, N,N-dimethylglycine, betaine, acetylcholine, (3-carboxypropyl)trimethylammonium) were investigated using electrospray ionization ion mobility spectrometry. Measured ion mobilities demonstrate a high correlation between mass and mobility in N2. In addn., identical mobilities within exptl. uncertainties are obsd. for structurally dissimilar ions with similar ion masses. For example, dimethylethylammonium (88 amu) cations and protonated N,N-dimethylaminoethanol cations (90 amu) show identical mobilities (1.93 cm2 V-1 s-1) though N,N-dimethylaminoethanol contains a hydroxyl functional group while dimethylethylammonium only contains alkyl groups. Computational anal. was performed using the modified trajectory (TJ) method with nonspherical N2 mols. as the drift gas. The sensitivity of the ammonium cation collision cross sections to the details of the ion-neutral interactions was investigated and compared to other classes of org. mols. (carboxylic acids and abiotic amino acids). The specific charge distribution of the mol. ions in the investigated mass range has an insignificant affect on the collision cross section.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhvFChsbY%253D&md5=f0354c65a2e2274c042129ccdba8f7ca
  16. 16
    Campuzano, I.; Bush, M. F.; Robinson, C. V.; Beaumont, C.; Richardson, K.; Kim, H.; Kim, H. I. Anal. Chem. 2012, 84, 1026– 1033,  DOI: 10.1021/ac202625t
    Google Scholar
    16
    Structural Characterization of Drug-like Compounds by Ion Mobility Mass Spectrometry: Comparison of Theoretical and Experimentally Derived Nitrogen Collision Cross Sections
    Campuzano, Iain; Bush, Matthew F.; Robinson, Carol V.; Beaumont, Claire; Richardson, Keith; Kim, Hyungjun; Kim, Hugh I.
    Analytical Chemistry (Washington, DC, United States) (2012), 84 (2), 1026-1033CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    We present the use of drug-like mols. as a traveling wave (T-wave) ion mobility (IM) calibration sample set, covering the m/z range of 122.1-609.3, the nitrogen collision cross-section (ΩN2) range of 124.5-254.3 Å2 and the helium collision cross-section (ΩHe) range of 63.0-178.8 Å2. Abs. ΩN2 and ΩHe values for the drug-like calibrants and two diastereomers were measured using a drift-tube instrument with radio frequency (RF) ion confinement. T-wave drift-times for the protonated diastereomers betamethasone and dexamethasone are reproducibly different. Calibration of these drift-times yields T-wave ΩN2 values of 189.4 and 190.4 Å2, resp. These results demonstrate the ability of T-wave IM spectrometry to differentiate diastereomers differing in ΩN2 value by only 1 Å2, even though the resoln. of these IM expts. were ∼40 (Ω/ΔΩ). Demonstrated through d. functional theory optimized geometries and ionic electrostatic surface potential anal., the small but measurable mobility difference between the two diastereomers is mainly due to short-range van der Waals interactions with the neutral buffer gas and not long-range charge-induced dipole interactions. The exptl. RF-confining drift-tube and T-wave ΩN2 values were also evaluated using a nitrogen based trajectory method, optimized for T-wave operating temp. and pressures, incorporating addnl. scaling factors to the Lennard-Jones potentials. Exptl. ΩHe values were also compared to the original and optimized helium based trajectory methods.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhsFOhtrvJ&md5=f5d0d911bc241a8c07aafbbe80139173
  17. 17
    Larriba, C.; Hogan, C. J., Jr. J. Phys. Chem. A 2013, 117, 3887– 3901,  DOI: 10.1021/jp312432z
    Google Scholar
    17
    Ion Mobilities in Diatomic Gases: Measurement versus Prediction with Non-Specular Scattering Models
    Larriba, Carlos; Hogan, Christopher J.
    Journal of Physical Chemistry A (2013), 117 (19), 3887-3901CODEN: JPCAFH; ISSN:1089-5639. (American Chemical Society)
    Ion/elec. mobility measurements of nanoparticles and polyat. ions are typically linked to particle/ion phys. properties through either application of the Stokes-Millikan relationship or comparison to mobilities predicted from polyat. models, which assume that gas mols. scatter specularly and elastically from rigid structural models. However, there is a discrepancy between these approaches; when specular, elastic scattering models (i.e., elastic-hard-sphere scattering, EHSS) are applied to polyat. models of nanometer-scale ions with finite-sized impinging gas mols., predictions are in substantial disagreement with the Stokes-Millikan equation. To rectify this discrepancy, a new approach is developed and tested for mobility calcns. using polyat. models in which non-specular (diffuse) and inelastic gas-mol. scattering is considered. Two distinct semiempirical models of gas-mol. scattering from particle surfaces were considered. In the first, which has been traditionally invoked in the study of aerosol nanoparticles, 91% of collisions are diffuse and thermally accommodating, and 9% are specular and elastic. In the second, all collisions are considered to be diffuse and accommodating, but the av. speed of the gas mols. reemitted from a particle surface is 8% lower than the mean thermal speed at the particle temp. Both scattering models attempt to mimic exchange between translational, vibrational, and rotational modes of energy during collision, as would be expected during collision between a nonmonoat. gas mol. and a nonfrozen particle surface. The mobility calcn. procedure was applied considering both hard-sphere potentials between gas mols. and the atoms within a particle and the long-range ion-induced dipole (polarization) potential. Predictions were compared to previous measurements in air near room temp. of multiply charged poly(ethylene glycol) (PEG) ions, which range in morphol. from compact to highly linear, and singly charged tetraalkylammonium cations. It was found that both non-specular, inelastic scattering rules lead to excellent agreement between predictions and exptl. mobility measurements (within 5% of each other) and that polarization potentials must be considered to make correct predictions for high-mobility particles/ions. Conversely, traditional specular, elastic scattering models were found to substantially overestimate the mobilities of both types of ions.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXjvFWhsb0%253D&md5=a40f1a9527fc4108f161648277010c9d
  18. 18
    Ewing, S. A.; Donor, M. T.; Wilson, J. W.; Prell, J. S. J. Am. Soc. Mass Spectrom. 2017, 28, 587– 596,  DOI: 10.1007/s13361-017-1594-2
    Google Scholar
    18
    Collidoscope: An Improved Tool for Computing Collisional Cross-Sections with the Trajectory Method
    Ewing, Simon A.; Donor, Micah T.; Wilson, Jesse W.; Prell, James S.
    Journal of the American Society for Mass Spectrometry (2017), 28 (4), 587-596CODEN: JAMSEF; ISSN:1044-0305. (Springer)
    Ion mobility-mass spectrometry (IM-MS) can be a powerful tool for detg. structural information about ions in the gas phase, from small covalent analytes to large, native-like or denatured proteins and complexes. For large biomol. ions, which may have a wide variety of possible gas-phase conformations and multiple charge sites, quant., phys. explicit modeling of collisional cross sections (CCSs) for comparison to IMS data can be challenging and time-consuming. We present a "trajectory method" (TM) based CCS calculator, named "Collidoscope," which utilizes parallel processing and optimized trajectory sampling, and implements both He and N2 as collision gas options. Also included is a charge-placement algorithm for detg. probable charge site configurations for protonated protein ions given an input geometry in PDB file format. Results from Collidoscope are compared with those from the current state-of-the-art CCS simulation suite, IMoS. Collidoscope CCSs are within 4% of IMoS values for ions with masses from ∼18 Da to ∼800 kDa. Collidoscope CCSs using X-ray crystal geometries are typically within a few percent of IM-MS exptl. values for ions with mass up to ∼3.5 kDa (melittin), and discrepancies for larger ions up to ∼800 kDa (GroEL) are attributed in large part to changes in ion structure during and after the electrospray process. Due to its phys. explicit modeling of scattering, computational efficiency, and accuracy, Collidoscope can be a valuable tool for IM-MS research, esp. for large biomol. ions.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXisF2mt7k%253D&md5=1f840c45579e5371f1601aadd1dc0e60
  19. 19
    Lee, J. W.; Lee, H. H. L.; Davidson, K. L.; Bush, M. F.; Kim, H. I. Analyst 2018, 143, 1786– 1796,  DOI: 10.1039/C8AN00270C
    Google Scholar
    19
    Structural characterization of small molecular ions by ion mobility mass spectrometry in nitrogen drift gas: improving the accuracy of trajectory method calculations
    Lee, Jong Wha; Lee, Hyun Hee L.; Davidson, Kimberly L.; Bush, Matthew F.; Kim, Hugh I.
    Analyst (Cambridge, United Kingdom) (2018), 143 (8), 1786-1796CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
    The investigation of ion structures based on a combination of ion mobility mass spectrometry (IM-MS) expts. and theor. collision cross section (CCS) calcns. has become important to many fields of research. However, the accuracy of current CCS calcns. for ions in nitrogen drift gas limits the information content of many expts. In particular, few studies have evaluated and attempted to improve the theor. tools for CCS calcn. in nitrogen drift gas. In this study, based on high-quality exptl. measurements and theor. modeling, a comprehensive evaluation of various aspects of CCS calcns. in nitrogen drift gas is performed. It is shown that the modification of the ion-nitrogen van der Waals (vdW) interaction potential enables accurate CCS predictions of 29 small ions with ca. 3% max. relative error. The present method exhibits no apparent systematic bias with respect to ion CCS (size) and dipole moment, suggesting that the method adequately describes the long-range interactions between the ions and the buffer gas. However, the method shows limitations in reproducing exptl. CCS at low temps. ( < 150 K) and for macromol. ions, and calcns. for these cases should be complemented by CCS calcn. methods in helium drift gas. This study presents an accurate and well-characterized CCS calcn. method for ions in nitrogen drift gas that is expected to become an important tool for ion structural characterization and mol. identification. The exptl. values reported here also provide a foundation for future studies aiming at developing more efficient computational tools.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXkslamtbY%253D&md5=906f8f6bc85485ff55f1a0bd26257c0f
  20. 20
    Zanotto, L.; Heerdt, G.; Souza, P. C. T.; Araujo, G.; Skaf, M. S. J. Comput. Chem. 2018, 39, 1675– 1681,  DOI: 10.1002/jcc.25199
    Google Scholar
    20
    High performance collision cross section calculation-HPCCS
    Zanotto, Leandro; Heerdt, Gabriel; Souza, Paulo C. T.; Araujo, Guido; Skaf, Munir S.
    Journal of Computational Chemistry (2018), 39 (21), 1675-1681CODEN: JCCHDD; ISSN:0192-8651. (John Wiley & Sons, Inc.)
    Since the com. introduction of Ion Mobility coupled with Mass Spectrometry (IM-MS) devices in 2003, a large no. of research labs. have embraced the technique. IM-MS is a fairly rapid expt. used as a mol. sepn. tool and to obtain structural information. The interpretation of IM-MS data is still challenging and relies heavily on theor. calcns. of the mol.'s collision cross section (CCS) against a buffer gas. Here, a new software (HPCCS) is presented, which performs CCS calcns. using high performance computing techniques. Based on the trajectory method, HPCCS can accurately calc. CCS for a great variety of mols., ranging from small org. mols. to large protein complexes, using helium or nitrogen as buffer gas with considerable gains in computer time compared to publicly available codes under the same level of theory. HPCCS is available as free software under the Academic Use License at . © 2018 Wiley Periodicals, Inc.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjs1Cltb8%253D&md5=8e74ea412b89c3f305f223fac23acf8c
  21. 21
    Colby, S. M.; Thomas, D. G.; Nunez, J. R.; Baxter, D. J.; Glaesemann, K. R.; Brown, J. M.; Pirrung, M. A.; Govind, N.; Teeguarden, J. G.; Metz, T. O.; Renslow, R. S. Anal. Chem. 2019, 91, 4346– 4356,  DOI: 10.1021/acs.analchem.8b04567
    Google Scholar
    21
    ISiCLE: A Quantum Chemistry Pipeline for Establishing in Silico Collision Cross Section Libraries
    Colby, Sean M.; Thomas, Dennis G.; Nunez, Jamie R.; Baxter, Douglas J.; Glaesemann, Kurt R.; Brown, Joseph M.; Pirrung, Meg A.; Govind, Niranjan; Teeguarden, Justin G.; Metz, Thomas O.; Renslow, Ryan S.
    Analytical Chemistry (Washington, DC, United States) (2019), 91 (7), 4346-4356CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    High-throughput, comprehensive, and confident identifications of metabolites and other chems. in biol. and environmental samples will revolutionize the understanding of the role these chem. diverse mols. play in biol. systems. Despite recent technol. advances, metabolomics studies still result in the detection of a disproportionate no. of features that cannot be confidently assigned to a chem. structure. This inadequacy is driven by the single most significant limitation in metabolomics, the reliance on ref. libraries constructed by anal. of authentic ref. materials with limited com. availability. To this end, the authors have developed the in silico chem. library engine (ISiCLE), a high-performance computing-friendly cheminformatics workflow for generating libraries of chem. properties. In the instantiation described here, the authors predict probable three-dimensional mol. conformers (i.e., conformational isomers) using chem. identifiers as input, from which collision cross sections (CCS) are derived. The approach employs first-principles simulation, distinguished by the use of mol. dynamics, quantum chem., and ion mobility calcns., to generate structures and chem. property libraries, all without training data. Importantly, optimization of ISiCLE included a refactoring of the popular MOBCAL code for trajectory-based mobility calcns., improving its computational efficiency by over 2 orders of magnitude. Calcd. CCS values were validated against 1983 exptl. measured CCS values and compared to previously reported CCS calcn. approaches. Av. calcd. CCS error for the validation set is 3.2% using std. parameters, outperforming other d. functional theory (DFT)-based methods and machine learning methods (e.g., MetCCS). An online database is introduced for sharing both calcd. and exptl. CCS values (metabolomics.pnnl.gov), initially including a CCS library with over 1 million entries. Finally, three successful applications of mol. characterization using calcd. CCS are described, including providing evidence for the presence of an environmental degrdn. product, the sepn. of mol. isomers, and an initial characterization of complex blinded mixts. of exposure chems. This work represents a method to address the limitations of small mol. identification and offers an alternative to generating chem. identification libraries exptl. by analyzing authentic ref. materials. All code is available at github.com/pnnl.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXivVOgsb0%253D&md5=786b8a6e8d764d48a9a4debfc31e4d59
  22. 22
    Zhou, Z.; Shen, X.; Tu, J.; Zhu, Z. J. Anal. Chem. 2016, 88, 11084– 11091,  DOI: 10.1021/acs.analchem.6b03091
    Google Scholar
    22
    Large-Scale Prediction of Collision Cross-Section Values for Metabolites in Ion Mobility-Mass Spectrometry
    Zhou, Zhiwei; Shen, Xiaotao; Tu, Jia; Zhu, Zheng-Jiang
    Analytical Chemistry (Washington, DC, United States) (2016), 88 (22), 11084-11091CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    The rapid development of metabolomics has significantly advanced health and disease related research. However, metabolite identification remains a major anal. challenge for untargeted metabolomics. While the use of collision cross-section (CCS) values obtained in ion mobility - mass spectrometry (IM-MS) effectively increases identification confidence of metabolites, it is restricted by the limited no. of available CCS values for metabolites. Here, the authors demonstrated the use of a machine-learning algorithm called support vector regression (SVR) to develop a prediction method that utilized 14 common mol. descriptors to predict CCS values for metabolites. In this work, the authors first exptl. measured CCS values (ΩN2) of ∼400 metabolites in nitrogen buffer gas, and used these values as training data to optimize the prediction method. The high prediction precision of this method was externally validated using an independent set of metabolites with a median relative error (MRE) of ∼3%, better than conventional theor. calcn. Using the SVR based prediction method, a large-scale predicted CCS database was generated for 35,203 metabolites in the Human Metabolome Database (HMDB). For each metabolite, five different ion adducts in pos. and neg. modes were predicted, accounting for 176,015 CCS values in total. Finally, improved metabolite identification accuracy was demonstrated using real biol. samples. Conclusively, the results proved that the SVR based prediction method can accurately predict nitrogen CCS values (ΩN2) of metabolites from mol. descriptors, and effectively improve identification accuracy and efficiency in untargeted metabolomics. The predicted CCS database, namely, MetCCS, is freely available on the Internet.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhslWjsbbM&md5=2e93305e7cce5f48dae1ed5fe2e861a9
  23. 23
    Zhou, Z.; Tu, J.; Xiong, X.; Shen, X.; Zhu, Z. J. Anal. Chem. 2017, 89, 9559– 9566,  DOI: 10.1021/acs.analchem.7b02625
    Google Scholar
    23
    LipidCCS: Prediction of Collision Cross-Section Values for Lipids with High Precision To Support Ion Mobility-Mass Spectrometry-Based Lipidomics
    Zhou, Zhiwei; Tu, Jia; Xiong, Xin; Shen, Xiaotao; Zhu, Zheng-Jiang
    Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9559-9566CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    The use of collision cross-section (CCS) values derived from ion mobility-mass spectrometry (IM-MS) has been proven to facilitate lipid identifications. Its utility is restricted by the limited availability of CCS values. Recently, the machine-learning algorithm-based prediction (e.g., MetCCS) is reported to generate CCS values in a large-scale. However, the prediction precision is not sufficient to differentiate lipids due to their high structural similarities and subtle differences on CCS values. To address this challenge, we developed a new approach, namely, LipidCCS, to precisely predict lipid CCS values. In LipidCCS, a set of mol. descriptors were optimized using bioinformatic approaches to comprehensively describe the subtle structure differences for lipids. The use of optimized mol. descriptors together with a large set of std. CCS values for lipids (458 in total) to build the prediction model significantly improved the precision. The prediction precision of LipidCCS was externally validated with median relative errors (MRE) of ∼1% using independent data sets across different instruments (Agilent DTIM-MS and Waters TWIM-MS) and labs. We also demonstrated that the improved precision in the predicted LipidCCS database (15,646 lipids and 63,434 CCS values in total) could effectively reduce false-pos. identifications of lipids. Common users can freely access our LipidCCS web server for the following: (1) the prediction of lipid CCS values directly from SMILES structure; (2) database search; and (3) lipid match and identification. We believe LipidCCS will be a valuable tool to support IM-MS-based lipidomics.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhurbN&md5=147694b1cc6fdc83de29883511f8185c
  24. 24
    Bijlsma, L.; Bade, R.; Celma, A.; Mullin, L.; Cleland, G.; Stead, S.; Hernandez, F.; Sancho, J. V. Anal. Chem. 2017, 89, 6583– 6589,  DOI: 10.1021/acs.analchem.7b00741
    Google Scholar
    24
    Prediction of Collision Cross-Section Values for Small Molecules: Application to Pesticide Residue Analysis
    Bijlsma, Lubertus; Bade, Richard; Celma, Alberto; Mullin, Lauren; Cleland, Gareth; Stead, Sara; Hernandez, Felix; Sancho, Juan V.
    Analytical Chemistry (Washington, DC, United States) (2017), 89 (12), 6583-6589CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    The use of collision cross-section (CCS) values obtained by ion mobility high-resoln. mass spectrometry has added a third dimension (alongside retention time and exact mass) to aid in the identification of compds. However, its utility is limited by the no. of exptl. CCS values currently available. This work demonstrates the potential of artificial neural networks (ANNs) for the prediction of CCS values of pesticides. The predictor, based on eight software-chosen mol. descriptors, was optimized using CCS values of 205 small mols. and validated using a set of 131 pesticides. The relative error was within 6% for 95% of all CCS values for protonated mols., resulting in a median relative error less than 2%. In order to demonstrate the potential of CCS prediction, the strategy was applied to spinach samples. It notably improved the confidence in the tentative identification of suspect and nontarget pesticides.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXosVOqsr0%253D&md5=a3328a65ce4adc8c838b7aee5aa6c2d4
  25. 25
    Soper-Hopper, M. T.; Petrov, A. S.; Howard, J. N.; Yu, S. S.; Forsythe, J. G.; Grover, M. A.; Fernandez, F. M. Chem. Commun. (Cambridge, U. K.) 2017, 53, 7624– 7627,  DOI: 10.1039/C7CC04257D
    Google Scholar
    25
    Collision cross section predictions using 2-dimensional molecular descriptors
    Soper-Hopper, M. T.; Petrov, A. S.; Howard, J. N.; Yu, S.-S.; Forsythe, J. G.; Grover, M. A.; Fernandez, F. M.
    Chemical Communications (Cambridge, United Kingdom) (2017), 53 (54), 7624-7627CODEN: CHCOFS; ISSN:1359-7345. (Royal Society of Chemistry)
    Traditional methods for deriving computationally-generated collision cross sections for comparisons with ion mobility-mass spectrometry data require 3-dimensional energy-minimized structures and are often time consuming, preventing high throughput implementation. Here, we introduce a method to predict ion mobility collision cross sections of lipids and peptide analogs important in prebiotic chem. and other fields. Using less than 100 2-D mol. descriptors this approach resulted in prediction errors of less than 2%.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhtVCltbjP&md5=d5d2efa7af0ed7e4c83680770fb8ba1b
  26. 26
    Mollerup, C. B.; Mardal, M.; Dalsgaard, P. W.; Linnet, K.; Barron, L. P. J. Chromatogr., A 2018, 1542, 82– 88,  DOI: 10.1016/j.chroma.2018.02.025
    Google Scholar
    26
    Prediction of collision cross section and retention time for broad scope screening in gradient reversed-phase liquid chromatography-ion mobility-high resolution accurate mass spectrometry
    Mollerup, Christian Brinch; Mardal, Marie; Dalsgaard, Petur Weihe; Linnet, Kristian; Barron, Leon Patrick
    Journal of Chromatography A (2018), 1542 (), 82-88CODEN: JCRAEY; ISSN:0021-9673. (Elsevier B.V.)
    Exact mass, retention time (RT), and collision cross section (CCS) are used as identification parameters in liq. chromatog. coupled to ion mobility high resoln. accurate mass spectrometry (LC-IM-HRMS). Targeted screening analyses are now more flexible and can be expanded for suspect and nontargeted screening. These allow for tentative identification of new compds., and in-silico predicted ref. values are used for improving confidence and filtering false-pos. identifications. In this work, predictions of both RT and CCS values are performed with machine learning using artificial neural networks (ANNs). Prediction was based on mol. descriptors, 827 RTs, and 357 CCS values from pharmaceuticals, drugs of abuse, and their metabolites. ANN models for the prediction of RT or CCS sep. were examd., and the potential to predict both from a single model was investigated for the first time. The optimized combined RT-CCS model was a four-layered multilayer perceptron ANN, and the 95th prediction error percentiles were within 2 min RT error and 5% relative CCS error for the external validation set (n = 36) and the full RT-CCS dataset (n = 357). 88.6% (n = 733) of predicted RTs were within 2 min error for the full dataset. Overall, when using 2 min RT error and 5% relative CCS error, 91.9% (n = 328) of compds. were retained, while 99.4% (n = 355) were retained when using at least one of these thresholds. This combined prediction approach can therefore be useful for rapid suspect/nontargeted screening involving HRMS, and will support current workflows.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjtFamsbw%253D&md5=4d52e6dad7704a31434f4ff47e50923a
  27. 27
    Plante, P. L.; Francovic-Fontaine, E.; May, J. C.; McLean, J. A.; Baker, E. S.; Laviolette, F.; Marchand, M.; Corbeil, J. Anal. Chem. 2019, 91, 5191– 5199,  DOI: 10.1021/acs.analchem.8b05821
    Google Scholar
    27
    Predicting Ion Mobility Collision Cross-Sections Using a Deep Neural Network: DeepCCS
    Plante, Pier-Luc; Francovic-Fontaine, Elina; May, Jody C.; McLean, John A.; Baker, Erin S.; Laviolette, Francois; Marchand, Mario; Corbeil, Jacques
    Analytical Chemistry (Washington, DC, United States) (2019), 91 (8), 5191-5199CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Untargeted metabolomic measurements using mass spectrometry are a powerful tool for uncovering new small mols. with environmental and biol. importance. The small mol. identification step, however, still remains an enormous challenge due to fragmentation difficulties or unspecific fragment ion information. Current methods to address this challenge are often dependent on databases or require the use of NMR, which have their own difficulties. The use of the gas-phase collision cross section (CCS) values obtained from ion mobility spectrometry (IMS) measurements were recently demonstrated to reduce the no. of false pos. metabolite identifications. While promising, the amt. of empirical CCS information currently available is limited, thus predictive CCS methods need to be developed. In this article, the authors expand upon current exptl. IMS capabilities by predicting the CCS values using a deep learning algorithm. The authors successfully developed and trained a prediction model for CCS values requiring only information about a compd.'s SMILES notation and ion type. The use of data from five different labs. using different instruments allowed the algorithm to be trained and tested on more than 2400 mols. The resulting CCS predictions were found to achieve a coeff. of detn. of 0.97 and median relative error of 2.7% for a wide range of mols. Furthermore, the method requires only a small amt. of processing power to predict CCS values. Considering the performance, time, and resources necessary, as well as its applicability to a variety of mols., this model was able to outperform all currently available CCS prediction algorithms.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmtFOisLY%253D&md5=d735e5f80e99301113280f7c878314a5
  28. 28
    Hines, K.; Herron, J.; Xu, L. J. Lipid Res. 2017, 58, 809– 819,  DOI: 10.1194/jlr.D074724
    Google Scholar
    28
    Assessment of altered lipid homeostasis by HILIC-ion mobility-mass spectrometry-based lipidomics
    Hines, Kelly M.; Herron, Josi; Xu, Libin
    Journal of Lipid Research (2017), 58 (4), 809-819CODEN: JLPRAW; ISSN:0022-2275. (American Society for Biochemistry and Molecular Biology)
    Ion mobility-mass spectrometry (IM-MS) has proven to be a highly informative technique for the characterization of lipids from cells and tissues. We report the combination of hydrophilic-interaction liq. chromatog. (HILIC) with traveling-wave IM-MS (TWIM-MS) for comprehensive lipidomics anal. Main lipid categories such as glycerolipids, sphingolipids, and glycerophospholipids are sepd. on the basis of their lipid backbones in the IM dimension, whereas subclasses of each category are mostly sepd. on the basis of their headgroups in the HILIC dimension, demonstrating the orthogonality of HILIC and IM sepns. Using our previously established lipid calibrants for collision cross-section (CCS) measurements in TWIM, we measured over 250 CCS values covering 12 lipid classes in pos. and neg. modes. The coverage of the HILIC-IM-MS method is demonstrated in the anal. of Neuro2a neuroblastoma cells exposed to benzalkonium chlorides (BACs) with C10 or C16 alkyl chains, which we have previously shown to affect gene expression related to cholesterol and lipid homeostasis. We found that BAC exposure resulted in significant changes to several lipid classes, including glycerides, sphingomyelins, phosphatidylcholines, and phosphatidylethanolamines. Our results indicate that BAC exposure modifies lipid homeostasis in a manner that is dependent upon the length of the BAC alkyl chain.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXlsFSjtLg%253D&md5=b204c490ee0b8fe634cb08a50accae69
  29. 29
    Hines, K. M.; Waalkes, A.; Penewit, K.; Holmes, E. A.; Salipante, S. J.; Werth, B. J.; Xu, L. mSphere 2017, 2, e00492-17  DOI: 10.1128/mSphere.00492-17
    Google Scholar
    There is no corresponding record for this reference.
  30. 30
    Hines, K. M.; Xu, L. Chem. Phys. Lipids 2019, 219, 15– 22,  DOI: 10.1016/j.chemphyslip.2019.01.007
    Google Scholar
    30
    Lipidomic consequences of phospholipid synthesis defects in Escherichia coli revealed by HILIC-ion mobility-mass spectrometry
    Hines, Kelly M.; Xu, Libin
    Chemistry and Physics of Lipids (2019), 219 (), 15-22CODEN: CPLIA4; ISSN:0009-3084. (Elsevier Ireland Ltd.)
    Our understanding of phospholipid biosynthesis in Gram-pos. and Gram-neg. bacteria is derived from the prototypical Gram-neg. organism Escherichia coli. The inner and outer membranes of E. coli are largely composed of phosphatidylethanolamine (PE), minor amts. of phosphatidylglycerol (PG) and cardiolipin (CL). We report here the utility of hydrophilic interaction liq. chromatog. (HILIC) paired with ion mobility-mass spectrometry (IM-MS) for the comprehensive anal. of the E. coli lipidome. Using strains with chromosomal deletions in the PG and CL synthesis genes pgsA and clsABC, resp., we show that defective phospholipid biosynthesis in E. coli results in fatty-acid specific changes in select lipid classes and the presence of the minor triacylated phospholipids, acylphosphatidyl glycerol (acylPG) and N-acylphosphatidylethanolamine (N-acylPE). Notably, acylPGs were accumulated in the clsABC-KO strain, but were absent in other mutant strains. The sepn. of 1-lyso and 2-lyso-phosphatidylethanolamines (lysoPEs) is demonstrated in both the HILIC and IM dimensions. Using our previously validated calibration method, collision cross section values of nearly 200 phospholipids found in E. coli were detd. on a traveling wave IM-MS platform, including newly reported values for cardiolipins, positional isomers of lysoPEs, acylPGs and N-acylPEs.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXhvFOhurY%253D&md5=d2b158f0b956c6c56a794f80fe5d7133
  31. 31
    Groessl, M.; Graf, S.; Knochenmuss, R. Analyst 2015, 140, 6904– 6911,  DOI: 10.1039/C5AN00838G
    Google Scholar
    31
    High resolution ion mobility-mass spectrometry for separation and identification of isomeric lipids
    Groessl, M.; Graf, S.; Knochenmuss, R.
    Analyst (Cambridge, United Kingdom) (2015), 140 (20), 6904-6911CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
    Lipidomics is a particularly difficult anal. challenge due to the no. and importance of isomeric species that are known or postulated in biol. samples. Current sepn. and identification techniques are too often insufficiently powerful, slow or ambiguous. High resoln., low field ion mobility coupled to mass spectrometry is shown here to have sufficient performance to represent a new alternative for lipidomics. For the first time, drift-tube ion mobility sepn. of lipid isomers that differ only in position of the acyl chain, position of the double bond or double bond geometry is demonstrated. Differences in collision cross sections of <1% are sufficient for baseline sepn. The same level of performance is maintained in complex biol. mixts. More than 130 high-precision reduced mobility and collision cross section values were also detd. for a range of lipids. Such data can be the basis of a new lipidomics workflow, as the appropriate libraries are developed.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtlKntbvP&md5=dd7628f57b198a94f5520191a96f9435
  32. 32
    Hines, K. M.; Ross, D. H.; Davidson, K. L.; Bush, M. F.; Xu, L. Anal. Chem. 2017, 89, 9023– 9030,  DOI: 10.1021/acs.analchem.7b01709
    Google Scholar
    32
    Large-Scale Structural Characterization of Drug and Drug-Like Compounds by High-Throughput Ion Mobility-Mass Spectrometry
    Hines, Kelly M.; Ross, Dylan H.; Davidson, Kimberly L.; Bush, Matthew F.; Xu, Libin
    Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9023-9030CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Ion mobility-mass spectrometry (IM-MS) can provide orthogonal information, i.e. m/z and collision cross section (CCS), for the identification of drugs and drug metabolites. However, only a small no. of CCS values are available for drugs, which limits the use of CCS as an identification parameter and the assessment of structure-function relationships of drugs using IM-MS. Here the authors report the development of a rapid workflow for the measurement of CCS values of a large no. of drug or drug-like mols. in nitrogen on the widely available traveling wave IM-MS (TWIM-MS) platform. Using a combination of small mol. and polypeptide CCS calibrants, the authors successfully detd. the nitrogen CCS values of 1425 drug or drug-like mols. in the MicroSource Discovery Systems' Spectrum Collection using flow injection anal. of 384-well plates. Software was developed to streamline data extn., processing, and calibration. The authors found that the overall drug collection covers a wide CCS range for the same masses, suggesting a large structural diversity of these drugs. However, individual drug classes appear to occupy a narrow and unique space in the CCS-mass 2D spectrum, suggesting a tight structure-function relationship for each class of drugs with a specific target. The authors obsd. bimodal distributions for several antibiotic species due to multiple protomers, including the known fluoroquinolone protomers and the newer finding of cephalosporin protomers. Lastly, the authors demonstrated the utility of the high-throughput method and drug CCS database by quickly and confidently confirming the active component in a pharmaceutical product.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhtbfE&md5=6dac3abb0c1050b8370a8e79066a7d34
  33. 33
    May, J. C.; Goodwin, C. R.; Lareau, N. M.; Leaptrot, K. L.; Morris, C. B.; Kurulugama, R. T.; Mordehai, A.; Klein, C.; Barry, W.; Darland, E.; Overney, G.; Imatani, K.; Stafford, G. C.; Fjeldsted, J. C.; McLean, J. A. Anal. Chem. 2014, 86, 2107– 2116,  DOI: 10.1021/ac4038448
    Google Scholar
    33
    Conformational Ordering of Biomolecules in the Gas Phase: Nitrogen Collision Cross Sections Measured on a Prototype High Resolution Drift Tube Ion Mobility-Mass Spectrometer
    May, Jody C.; Goodwin, Cody R.; Lareau, Nichole M.; Leaptrot, Katrina L.; Morris, Caleb B.; Kurulugama, Ruwan T.; Mordehai, Alex; Klein, Christian; Barry, William; Darland, Ed; Overney, Gregor; Imatani, Kenneth; Stafford, George C.; Fjeldsted, John C.; McLean, John A.
    Analytical Chemistry (Washington, DC, United States) (2014), 86 (4), 2107-2116CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Ion mobility-mass spectrometry measurements which describe the gas-phase scaling of mol. size and mass are of both fundamental and pragmatic utility. Fundamentally, such measurements expand the authors' understanding of intrinsic intramol. folding forces in the absence of solvent. Practically, reproducible transport properties, such as gas-phase collision cross-section (CCS), are anal. useful metrics for identification and characterization purposes. Here, the authors report 594 CCS values obtained in nitrogen drift gas on an electrostatic drift tube ion mobility-mass spectrometry (IM-MS) instrument. The instrument platform is a newly developed prototype incorporating a uniform-field drift tube bracketed by electrodynamic ion funnels and coupled to a high resoln. quadrupole time-of-flight mass spectrometer. The CCS values reported here are of high exptl. precision (±0.5% or better) and represent four chem. distinct classes of mols. (quaternary ammonium salts, lipids, peptides, and carbohydrates), which enables structural comparisons to be made between mols. of different chem. compns. for the rapid "omni-omic" characterization of complex biol. samples. Comparisons made between helium and nitrogen-derived CCS measurements demonstrate that nitrogen CCS values are systematically larger than helium values; however, general sepn. trends between chem. classes are retained regardless of the drift gas. These results underscore that, for the highest CCS accuracy, care must be exercised when using helium-derived CCS values to calibrate measurements obtained in nitrogen, as is the common practice in the field.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXpslehtQ%253D%253D&md5=9e420afba857551e3d389ad7e038187a
  34. 34
    Paglia, G.; Williams, J. P.; Menikarachchi, L.; Thompson, J. W.; Tyldesley-Worster, R.; Halldorsson, S.; Rolfsson, O.; Moseley, A.; Grant, D.; Langridge, J.; Palsson, B. O.; Astarita, G. Anal. Chem. 2014, 86, 3985– 3993,  DOI: 10.1021/ac500405x
    Google Scholar
    34
    Ion Mobility Derived Collision Cross Sections to Support Metabolomics Applications
    Paglia, Giuseppe; Williams, Jonathan P.; Menikarachchi, Lochana; Thompson, J. Will; Tyldesley-Worster, Richard; Halldorsson, Skarphedinn; Rolfsson, Ottar; Moseley, Arthur; Grant, David; Langridge, James; Palsson, Bernhard O.; Astarita, Giuseppe
    Analytical Chemistry (Washington, DC, United States) (2014), 86 (8), 3985-3993CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Metabolomics is a rapidly evolving anal. approach in life and health sciences. The structural elucidation of the metabolites of interest remains a major anal. challenge in the metabolomics workflow. Here, we investigate the use of ion mobility as a tool to aid metabolite identification. Ion mobility allows for the measurement of the rotationally averaged collision cross-section (CCS), which gives information about the ionic shape of a mol. in the gas phase. We measured the CCSs of 125 common metabolites using traveling-wave ion mobility-mass spectrometry (TW-IM-MS). CCS measurements were highly reproducible on instruments located in three independent labs. (RSD < 5% for 99%). We also detd. the reproducibility of CCS measurements in various biol. matrixes including urine, plasma, platelets, and red blood cells using ultra performance liq. chromatog. (UPLC) coupled with TW-IM-MS. The mean RSD was < 2% for 97% of the CCS values, compared to 80% of retention times. Finally, as proof of concept, we used UPLC-TW-IM-MS to compare the cellular metabolome of epithelial and mesenchymal cells, an in vitro model used to study cancer development. Exptl. detd. and computationally derived CCS values were used as orthogonal anal. parameters in combination with retention time and accurate mass information to confirm the identity of key metabolites potentially involved in cancer. Thus, our results indicate that adding CCS data to searchable databases and to routine metabolomics workflows will increase the identification confidence compared to traditional anal. approaches.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXksVWisbs%253D&md5=361ab03df449c68121984ec0f1cb4241
  35. 35
    Zheng, X.; Aly, N. A.; Zhou, Y.; Dupuis, K. T.; Bilbao, A.; Paurus, V. L.; Orton, D. J.; Wilson, R.; Payne, S. H.; Smith, R. D.; Baker, E. S. Chem. Sci. 2017, 8, 7724– 7736,  DOI: 10.1039/C7SC03464D
    Google Scholar
    35
    A structural examination and collision cross section database for over 500 metabolites and xenobiotics using drift tube ion mobility spectrometry
    Zheng, Xueyun; Aly, Noor A.; Zhou, Yuxuan; Dupuis, Kevin T.; Bilbao, Aivett; Paurus, Vanessa L.; Orton, Daniel J.; Wilson, Ryan; Payne, Samuel H.; Smith, Richard D.; Baker, Erin S.
    Chemical Science (2017), 8 (11), 7724-7736CODEN: CSHCCN; ISSN:2041-6520. (Royal Society of Chemistry)
    The confident identification of metabolites and xenobiotics in biol. and environmental studies is an anal. challenge due to their immense dynamic range, vast chem. space and structural diversity. Ion mobility spectrometry (IMS) is widely used for small mol. analyses since it can sep. isomeric species and be easily coupled with front end sepns. and mass spectrometry for multidimensional characterizations. However, to date IMS metabolomic and exposomic studies have been limited by an inadequate no. of accurate collision cross section (CCS) values for small mols., causing features to be detected but not confidently identified. In this work, we utilized drift tube IMS (DTIMS) to directly measure CCS values for over 500 small mols. including primary metabolites, secondary metabolites and xenobiotics. Since DTIMS measurements do not need calibrant ions or calibration like some other IMS techniques, they avoid calibration errors which can cause problems in distinguishing structurally similar mols. All measurements were performed in triplicate in both pos. and neg. polarities with nitrogen gas and seven different elec. fields, so that relative std. deviations (RSD) could be assessed for each mol. and structural differences studied. The primary metabolites analyzed to date have come from key metab. pathways such as glycolysis, the pentose phosphate pathway and the tricarboxylic acid cycle, while the secondary metabolites consisted of classes such as terpenes and flavonoids, and the xenobiotics represented a range of mols. from antibiotics to polycyclic arom. hydrocarbons. Different CCS trends were obsd. for several of the diverse small mol. classes and when urine features were matched to the database, the addn. of the IMS dimension greatly reduced the possible no. of candidate mols. This CCS database and structural information are freely available for download at http://panomics.pnnl.gov/metabolites/ with new mols. being added frequently.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhsFOlu7bE&md5=08bd6a4ed911bea19834302d123923b4
  36. 36
    Zhou, Z.; Xiong, X.; Zhu, Z. J. Bioinformatics 2017, 33, 2235– 2237,  DOI: 10.1093/bioinformatics/btx140
    Google Scholar
    36
    MetCCS predictor: a web server for predicting collision cross-section values of metabolites in ion mobility-mass spectrometry based metabolomics
    Zhou, Zhiwei; Xiong, Xin; Zhu, Zheng-Jiang
    Bioinformatics (2017), 33 (14), 2235-2237CODEN: BOINFP; ISSN:1460-2059. (Oxford University Press)
    In metabolomics, rigorous structural identification of metabolites presents a challenge for bioinformatics. The use of collision cross-section (CCS) values of metabolites derived from ion mobility-mass spectrometry effectively increases the confidence of metabolite identification, but this technique suffers from the limit no. of available CCS values. Currently, there is no software available for rapidly generating the metabolites' CCS values. Here, we developed the first web server, namely, MetCCS Predictor, for predicting CCS values. It can predict the CCS values of metabolites using mol. descriptors within a few seconds. Common users with limited background on bioinformatics can benefit from this software and effectively improve the metabolite identification in metabolomics.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhvV2nur7N&md5=5ca86ebe753f0fadca8cc4f26ba0e1d7
  37. 37
    Nichols, C. M.; May, J. C.; Sherrod, S. D.; McLean, J. A. Analyst 2018, 143, 1556– 1559,  DOI: 10.1039/C8AN00056E
    Google Scholar
    37
    Automated flow injection method for the high precision determination of drift tube ion mobility collision cross sections
    Nichols, Charles M.; May, Jody C.; Sherrod, Stacy D.; McLean, John A.
    Analyst (Cambridge, United Kingdom) (2018), 143 (7), 1556-1559CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
    The field of ion mobility-based omics studies requires high-quality collision cross section (CCS) libraries to effectively utilize CCS as a mol. descriptor. Abs. CCS values with the highest precision are obtained on drift tube instruments by measuring the drift time of ions at multiple drift voltages, commonly referred to as a 'stepped field' expt. However, generating large scale abs. CCS libraries from drift tube instruments is time consuming due to the current lack of high-throughput methods. This communication reports a fully automated stepped-field method to acquire abs. CCS on com. available equipment. Using a drift tube ion mobility-mass spectrometer (DTIM-MS) coupled to a minimally modified liq. chromatog. (LC) system, CCS values can be measured online with a carefully timed flow injection anal. (FIA) expt. Results demonstrate that the FIA stepped-field method yields CCS values which are of high anal. precision (<0.4% relative std. deviation, RSD) and accuracy (≤0.4% difference) comparable to CCS values obtained using traditional direct-infusion stepped-field expts. This high-throughput CCS method consumes very little sample vol. (20 μL) and will expedite the generation of large-scale CCS libraries to support mol. identification within global untargeted studies.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXks1Ois78%253D&md5=15f77a4e81c957c39c5019980edd1c13
  38. 38
    Righetti, L.; Bergmann, A.; Galaverna, G.; Rolfsson, O.; Paglia, G.; Dall’Asta, C. Anal. Chim. Acta 2018, 1014, 50– 57,  DOI: 10.1016/j.aca.2018.01.047
    Google Scholar
    38
    Ion mobility-derived collision cross section database: Application to mycotoxin analysis
    Righetti, Laura; Bergmann, Andreas; Galaverna, Gianni; Rolfsson, Ottar; Paglia, Giuseppe; Dall'Asta, Chiara
    Analytica Chimica Acta (2018), 1014 (), 50-57CODEN: ACACAM; ISSN:0003-2670. (Elsevier B.V.)
    The recent hyphenation of ion mobility spectrometry (IMS) with high resoln. mass spectrometry (HRMS) has risen as a powerful technique for both targeted and non-targeted screening, reducing background noise and allowing sepn. of isomeric and isobaric compds. Nevertheless, such an approach remains largely unexplored in food safety applications, such as mycotoxin anal. To implement ion mobility in routinely MS-based mycotoxin workflows, searchable databases with collusion cross section (CCS) values and accurate mass-values are required. This paper provides for the first time a traveling-wave IMS (TWIMS)-derived CCS database for mycotoxins, including more than 100 CCS values. The measurements showed high reproducibility (RSD<2%) across different instrumental conditions as well as several complex cereal matrixes, showing a mean inter-matrix precision of RSD <0.9%. As a proof of concept, the database was applied to the anal. of several spiked as well as naturally incurred cereal-based samples. In addn., the effect of adducts on the drift time was studied in a series of mycotoxins in order to understand potential deviations from expected drift time behaviors. Overall, our study confirmed that CCS values represent a physicochem. property that can be used alongside the traditional mol. identifiers of precursor ion accurate mass, fragment ions, isotopic pattern, and retention time.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXisFelu74%253D&md5=7258333f8021aa0c8868bb53aa0a7d3d
  39. 39
    Leaptrot, K. L.; May, J. C.; Dodds, J. N.; McLean, J. A. Nat. Commun. 2019, 10, 985,  DOI: 10.1038/s41467-019-08897-5
    Google Scholar
    39
    Ion mobility conformational lipid atlas for high confidence lipidomics
    Leaptrot Katrina L; May Jody C; Dodds James N; McLean John A
    Nature communications (2019), 10 (1), 985 ISSN:.
    Lipids are highly structurally diverse molecules involved in a wide variety of biological processes. Here, we use high precision ion mobility-mass spectrometry to compile a structural database of 456 mass-resolved collision cross sections (CCS) of sphingolipid and glycerophospholipid species. Our CCS database comprises sphingomyelin, cerebroside, ceramide, phosphatidylethanolamine, phosphatidylcholine, phosphatidylserine, and phosphatidic acid classes. Primary differences observed are between lipid categories, with sphingolipids exhibiting 2-6% larger CCSs than glycerophospholipids of similar mass, likely a result of the sphingosine backbone's restriction of the sn1 tail length, limiting gas-phase packing efficiency. Acyl tail length and degree of unsaturation are found to be the primary structural descriptors determining CCS magnitude, with degree of unsaturation being four times as influential per mass unit. The empirical CCS values and previously unmapped quantitative structural trends detailed in this work are expected to facilitate prediction of CCS in broadscale lipidomics research.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3cfos1GitA%253D%253D&md5=f4b60c23db2a9889cc167c49d40519e8
  40. 40
    Blaženović, I.; Shen, T.; Mehta, S. S.; Kind, T.; Ji, J.; Piparo, M.; Cacciola, F.; Mondello, L.; Fiehn, O. Anal. Chem. 2018, 90 (18), 10758– 10764,  DOI: 10.1021/acs.analchem.8b01527
    Google Scholar
    40
    Increasing Compound Identification Rates in Untargeted Lipidomics Research with Liquid Chromatography Drift Time-Ion Mobility Mass Spectrometry
    Blazenovic, Ivana; Shen, Tong; Mehta, Sajjan S.; Kind, Tobias; Ji, Jian; Piparo, Marco; Cacciola, Francesco; Mondello, Luigi; Fiehn, Oliver
    Analytical Chemistry (Washington, DC, United States) (2018), 90 (18), 10758-10764CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Unknown metabolites represent a bottleneck in untargeted metabolomics research. Ion mobility-mass spectrometry (IM-MS) facilitates lipid identification because it yields collision cross section (CCS) information that is independent from mass or lipophilicity. To date, only a few CCS values are publicly available for complex lipids such as phosphatidylcholines, sphingomyelins, or triacylglycerides. This scarcity of data limits the use of CCS values as an identification parameter that is orthogonal to mass, MS/MS, or retention time. A combination of lipid descriptors was used to train five different machine learning algorithms for automatic lipid annotations, combining accurate mass (m/z), retention time (RT), CCS values, carbon no., and unsatn. level. Using a training data set of 429 true pos. lipid annotations from four lipid classes, 92.7% correct annotations overall were achieved using internal cross-validation. The trained prediction model was applied to an unknown milk lipidomics data set and allowed for class 3 level annotations of most features detected in this application set according to Metabolomics Stds. Initiative (MSI) reporting guidelines.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhsVOltrvO&md5=9bb9b150edfe9444863e2a7274fa0c98
  41. 41
    Hines, K. M.; May, J. C.; McLean, J. A.; Xu, L. Anal. Chem. 2016, 88, 7329– 7336,  DOI: 10.1021/acs.analchem.6b01728
    Google Scholar
    41
    Evaluation of Collision Cross Section Calibrants for Structural Analysis of Lipids by Traveling Wave Ion Mobility-Mass Spectrometry
    Hines, Kelly M.; May, Jody C.; McLean, John A.; Xu, Libin
    Analytical Chemistry (Washington, DC, United States) (2016), 88 (14), 7329-7336CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Collision cross section (CCS) measurement of lipids using traveling wave ion mobility-mass spectrometry (TWIM-MS) is of high interest to the lipidomics field. However, currently available calibrants for CCS measurement using TWIM are predominantly peptides that display quite different phys. properties and gas-phase conformations from lipids, which could lead to large CCS calibration errors for lipids. Here we report the direct CCS measurement of a series of phosphatidylcholines (PCs) and phosphatidylethanolamines (PEs) in nitrogen using a drift tube ion mobility (DTIM) instrument and an evaluation of the accuracy and reproducibility of PCs and PEs as CCS calibrants for phospholipids against different classes of calibrants, including polyalanine (PolyAla), tetraalkylammonium salts (TAA), and hexakis(fluoroalkoxy)phosphazines (HFAP), in both pos. and neg. modes in TWIM-MS anal. We demonstrate that structurally mismatched calibrants lead to larger errors in calibrated CCS values while the structurally matched calibrants, PCs and PEs, gave highly accurate and reproducible CCS values at different traveling wave parameters. Using the lipid calibrants, the majority of the CCS values of several classes of phospholipids measured by TWIM are within 2% error of the CCS values measured by DTIM. The development of phospholipid CCS calibrants will enable high-accuracy structural studies of lipids and add an addnl. level of validation in the assignment of identifications in untargeted lipidomics expts.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhtVSltb7J&md5=6c37b06c11ffbdc24ac1bfc9961f9e40
  42. 42
    Reymond, J.-L.; Awale, M. ACS Chem. Neurosci. 2012, 3, 649– 657,  DOI: 10.1021/cn3000422
    Google Scholar
    42
    Exploring Chemical Space for Drug Discovery Using the Chemical Universe Database
    Reymond, Jean-Louis; Awale, Mahendra
    ACS Chemical Neuroscience (2012), 3 (9), 649-657CODEN: ACNCDM; ISSN:1948-7193. (American Chemical Society)
    Herein we review our recent efforts in searching for bioactive ligands by enumeration and virtual screening of the unknown chem. space of small mols. Enumeration from first principles shows that almost all small mols. (>99.9%) have never been synthesized and are still available to be prepd. and tested. We discuss open access sources of mols., the classification and representation of chem. space using mol. quantum nos. (MQN), its exhaustive enumeration in form of the chem. universe generated databases (GDB), and examples of using these databases for prospective drug discovery. MQN-searchable GDB, PubChem, and DrugBank are freely accessible at www.gdb.unibe.ch.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XmtValu70%253D&md5=bbbe3c0931328f3796ce999189374864
  43. 43
    Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V.; Vanderplas, J.; Passos, A.; Cournapeau, D.; Brucher, M.; Perrot, M.; Duchesnay, E. J. Mach Learn Res. 2011, 12, 2825– 2830
    Google Scholar
    There is no corresponding record for this reference.
  44. 44
    Picache, J. A.; Rose, B. S.; Balinski, A.; Leaptrot, K. L.; Sherrod, S. D.; May, J. C.; McLean, J. A. Chem. Sci. 2019, 10, 983– 993,  DOI: 10.1039/C8SC04396E
    Google Scholar
    44
    Collision cross section compendium to annotate and predict multi-omic compound identities
    Picache, Jaqueline A.; Rose, Bailey S.; Balinski, Andrzej; Leaptrot, Katrina L.; Sherrod, Stacy D.; May, Jody C.; McLean, John A.
    Chemical Science (2019), 10 (4), 983-993CODEN: CSHCCN; ISSN:2041-6520. (Royal Society of Chemistry)
    Ion mobility mass spectrometry (IM-MS) expands the analyte coverage of existing multi-omic workflows by providing an addnl. sepn. dimension as well as a parameter for characterization and identification of mols. - the collision cross section (CCS). This work presents a large, Unified CCS compendium of >3800 exptl. acquired CCS values obtained from traceable mol. stds. and measured with drift tube ion mobility-mass spectrometers. An interactive visualization of this compendium along with data analytic tools have been made openly accessible. Represented in the compendium are 14 structurally-based chem. super classes, consisting of a total of 80 classes and 157 subclasses. Using this large data set, regression fitting and predictive statistics have been performed to describe mass-CCS correlations specific to each chem. ontol. These structural trends provide a rapid and effective filtering method in the traditional untargeted workflow for identification of unknown biochem. species. The utility of the approach is illustrated by an application to metabolites in human serum, quantified trends of which were used to assess the probability of an unknown compd. belonging to a given class. CCS-based filtering narrowed the chem. search space by 60% while increasing the confidence in the remaining isomeric identifications from a single class, thus demonstrating the value of integrating predictive analyses into untargeted expts. to assist in identification workflows. The predictive abilities of this compendium will improve in specificity and expand to more chem. classes as addnl. data from the IM-MS community is contributed. Instructions for data submission to the compendium and criteria for inclusion are provided.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXitlSqtrbN&md5=177c56f0b3557ca849e5f224ffd5a224
  45. 45
    Hernandez-Mesa, M.; Le Bizec, B.; Monteau, F.; Garcia-Campana, A. M.; Dervilly-Pinel, G. Anal. Chem. 2018, 90, 4616– 4625,  DOI: 10.1021/acs.analchem.7b05117
    Google Scholar
    45
    Collision cross section (CCS) database: An additional measure to characterize steroids
    Hernandez-Mesa, Maykel; Le Bizec, Bruno; Monteau, Fabrice; Garcia-Campana, Ana M.; Dervilly-Pinel, Gaud
    Analytical Chemistry (Washington, DC, United States) (2018), 90 (7), 4616-4625CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Ion mobility spectrometry enhances the performance characteristics of liq. chromatog.-mass spectrometry workflows intended to steroid profiling by providing a new sepn. dimension and a novel characterization parameter, the so-called collision cross section (CCS). This work proposes the first CCS database for 300 steroids (i.e., endogenous, including phase I and phase II metabolites, and exogenous synthetic compds.), which involves 1080 ions and covers the CCS of 127 androgens, 84 estrogens, 50 corticosteroids, and 39 progestagens. This large database provides information related to all the ionized species identified for each steroid in pos. electrospray ionization mode as well as for estrogens in neg. ionization mode. CCS values have been measured using nitrogen as drift gas in the ion mobility cell. Generally, direct correlation exists between mass-to-charge ratio (m/z) and CCS because both are related parameters. However, several steroids mainly steroid glucuronides and steroid esters have been characterized as more compact or elongated mols. than expected. In such cases, CCS results in addnl. relevant information to retention time and mass spectral data for the identification of steroids. Moreover, several isomeric steroid pairs (e.g., 5β-androstane-3,17-dione and 5α-androstane-3,17-dione) have been sepd. based on their CCS differences. These results indicate that adding the CCS to databases in anal. workflows increases selectivity, thus improving the confidence in steroids anal. Consequences in terms of identification and quantification are discussed. Quality criteria and a construction of an interlab. reproducibility approach are also reported for the obtained CCS values. The CCS database described here is made publicly available.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXktlGisL8%253D&md5=b73301562f0f191833005d2a5f3b9fc2
  46. 46
    Sumner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W. M.; Fiehn, O.; Goodacre, R.; Griffin, J. L.; Hankemeier, T.; Hardy, N.; Harnly, J.; Higashi, R.; Kopka, J.; Lane, A. N.; Lindon, J. C.; Marriott, P.; Nicholls, A. W.; Reily, M. D.; Thaden, J. J.; Viant, M. R. Metabolomics 2007, 3 (3), 211– 221,  DOI: 10.1007/s11306-007-0082-2
    Google Scholar
    46
    Proposed minimum reporting standards for chemical analysis. Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI)
    Sumner, Lloyd W.; Amberg, Alexander; Barrett, Dave; Beale, Michael H.; Beger, Richard; Daykin, Clare A.; Fan, Teresa W.-M.; Fiehn, Oliver; Goodacre, Royston; Griffin, Julian L.; Hankemeier, Thomas; Hardy, Nigel; Harnly, James; Higashi, Richard; Kopka, Joachim; Lane, Andrew N.; Lindon, John C.; Marriott, Philip; Nicholls, Andrew W.; Reily, Michael D.; Thaden, John J.; Viant, Mark R.
    Metabolomics (2007), 3 (3), 211-221CODEN: METAHQ; ISSN:1573-3882. (Springer)
    There is a general consensus that supports the need for standardized reporting of metadata or information describing large-scale metabolomics and other functional genomics data sets. Reporting of std. metadata provides a biol. and empirical context for the data, facilitates exptl. replication, and enables the reinterrogation and comparison of data by others. Accordingly, the Metabolomics Stds. Initiative is building a general consensus concerning the min. reporting stds. for metabolomics expts. of which the Chem. Anal. Working Group (CAWG) is a member of this community effort. This article proposes the min. reporting stds. related to the chem. anal. aspects of metabolomics expts. including: sample prepn., exptl. anal., quality control, metabolite identification, and data pre-processing. These min. stds. currently focus mostly upon mass spectrometry and NMR spectroscopy due to the popularity of these techniques in metabolomics. However, addnl. input concerning other techniques is welcomed and can be provided via the CAWG online discussion forum at http://msi-workgroups.sourceforge.net/ or http://[email protected]. Further, community input related to this document can also be provided via this electronic forum.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhtFOntbvL&md5=e636dad276f9f1b2e756db169368eb53

Cited By

Click to copy section linkSection link copied!
Citation Statements
Explore this article's citation statements on scite.ai
powered by  Scite

This article is cited by 100 publications.

  1. Kaylie I. Kirkwood-Donelson, Prashant Rai, Lalith Perera, Michael B. Fessler, Alan K. Jarmusch. Bromine-Based Derivatization of Carboxyl-Containing Metabolites for Liquid Chromatography–Trapped Ion Mobility Spectrometry–Mass Spectrometry. Journal of the American Society for Mass Spectrometry 2025, Article ASAP.
  2. Anjana Elapavalore, Dylan H. Ross, Valentin Grouès, Dagny Aurich, Allison M. Krinsky, Sunghwan Kim, Paul A. Thiessen, Jian Zhang, James N. Dodds, Erin S. Baker, Evan E. Bolton, Libin Xu, Emma L. Schymanski. PubChemLite Plus Collision Cross Section (CCS) Values for Enhanced Interpretation of Nontarget Environmental Data. Environmental Science & Technology Letters 2025, 12 (2) , 166-174. https://doi.org/10.1021/acs.estlett.4c01003
  3. Xingliang He, Bin Wu, Xing Guo, Fulong Deng, Hong’en Sun, Zhihao He, Yixiang Duan, Zhongjun Zhao. Twisted Dipole Ion Guide (TDIG) for Flexible Ion Transfer in Atmospheric Pressure Ionization Mass Spectrometry. Analytical Chemistry 2025, 97 (2) , 1070-1077. https://doi.org/10.1021/acs.analchem.4c03255
  4. Thomas O. Metz, Christine H. Chang, Vasuk Gautam, Afia Anjum, Siyang Tian, Fei Wang, Sean M. Colby, Jamie R. Nunez, Madison R. Blumer, Arthur S. Edison, Oliver Fiehn, Dean P. Jones, Shuzhao Li, Edward T. Morgan, Gary J. Patti, Dylan H. Ross, Madelyn R. Shapiro, Antony J. Williams, David S. Wishart. Introducing “Identification Probability” for Automated and Transferable Assessment of Metabolite Identification Confidence in Metabolomics and Related Studies. Analytical Chemistry 2025, 97 (1) , 1-11. https://doi.org/10.1021/acs.analchem.4c04060
  5. Marcelino Varona, Daniel P. Dobson, José G. Napolitano, Rekha Thomas, Jessica L. Ochoa, David J. Russell, Christopher M. Crittenden. High Resolution Ion Mobility Enables the Structural Characterization of Atropisomers of GDC-6036, a KRAS G12C Covalent Inhibitor. Journal of the American Society for Mass Spectrometry 2024, 35 (11) , 2586-2595. https://doi.org/10.1021/jasms.4c00103
  6. Mithony Keng, Kenneth M Merz, Jr.. Eliminating the Deadwood: A Machine Learning Model for CCS Knowledge-Based Conformational Focusing for Lipids. Journal of Chemical Information and Modeling 2024, 64 (20) , 7864-7872. https://doi.org/10.1021/acs.jcim.4c01051
  7. Cheng Wang, Chuang Yuan, Yahui Wang, Yuying Shi, Tao Zhang, Gary J. Patti. Predicting Collision Cross-Section Values for Small Molecules through Chemical Class-Based Multimodal Graph Attention Network. Journal of Chemical Information and Modeling 2024, 64 (16) , 6305-6315. https://doi.org/10.1021/acs.jcim.3c01934
  8. Sara M. de Cripan, Trisha Arora, Adrià Olomí, Núria Canela, Gary Siuzdak, Xavier Domingo-Almenara. Predicting the Predicted: A Comparison of Machine Learning-Based Collision Cross-Section Prediction Models for Small Molecules. Analytical Chemistry 2024, 96 (22) , 9088-9096. https://doi.org/10.1021/acs.analchem.4c00630
  9. Ryan Nguyen, Ryan P. Seguin, Dylan H. Ross, Pengyu Chen, Sean Richardson, Jennifer Liem, Yvonne S. Lin, Libin Xu. Development and Application of a Multidimensional Database for the Detection of Quaternary Ammonium Compounds and Their Phase I Hepatic Metabolites in Humans. Environmental Science & Technology 2024, 58 (14) , 6236-6249. https://doi.org/10.1021/acs.est.3c10845
  10. Pattipong Wisanpitayakorn, Sitanan Sartyoungkul, Alongkorn Kurilung, Yongyut Sirivatanauksorn, Wonnop Visessanguan, Nuankanya Sathirapongsasuti, Sakda Khoomrung. Accurate Prediction of Ion Mobility Collision Cross-Section Using Ion’s Polarizability and Molecular Mass with Limited Data. Journal of Chemical Information and Modeling 2024, 64 (5) , 1533-1542. https://doi.org/10.1021/acs.jcim.3c01491
  11. Jana M. Carpenter, Hannah M. Hynds, Kingsley Bimpeh, Kelly M. Hines. HILIC-IM-MS for Simultaneous Lipid and Metabolite Profiling of Bacteria. ACS Measurement Science Au 2024, 4 (1) , 104-116. https://doi.org/10.1021/acsmeasuresciau.3c00051
  12. Hannah M. Hynds, Kelly M. Hines. MOCCal: A Multiomic CCS Calibrator for Traveling Wave Ion Mobility Mass Spectrometry. Analytical Chemistry 2024, 96 (3) , 1185-1194. https://doi.org/10.1021/acs.analchem.3c04290
  13. Xue-Chao Song, Elena Canellas, Nicola Dreolin, Jeff Goshawk, Meilin Lv, Guangbo Qu, Cristina Nerin, Guibin Jiang. Application of Ion Mobility Spectrometry and the Derived Collision Cross Section in the Analysis of Environmental Organic Micropollutants. Environmental Science & Technology 2023, 57 (51) , 21485-21502. https://doi.org/10.1021/acs.est.3c03686
  14. Haosong Zhang, Mingdu Luo, Hongmiao Wang, Fandong Ren, Yandong Yin, Zheng-Jiang Zhu. AllCCS2: Curation of Ion Mobility Collision Cross-Section Atlas for Small Molecules Using Comprehensive Molecular Representations. Analytical Chemistry 2023, 95 (37) , 13913-13921. https://doi.org/10.1021/acs.analchem.3c02267
  15. Noelle Reimers, Quynh Do, Rutan Zhang, Angela Guo, Ryan Ostrander, Alyson Shoji, Chau Vuong, Libin Xu. Tracking the Metabolic Fate of Exogenous Arachidonic Acid in Ferroptosis Using Dual-Isotope Labeling Lipidomics. Journal of the American Society for Mass Spectrometry 2023, 34 (9) , 2016-2024. https://doi.org/10.1021/jasms.3c00181
  16. Yulemni Morel, Jace W. Jones. Utilization of LC–MS/MS and Drift Tube Ion Mobility for Characterizing Intact Oxidized Arachidonate-Containing Glycerophosphatidylethanolamine. Journal of the American Society for Mass Spectrometry 2023, 34 (8) , 1609-1620. https://doi.org/10.1021/jasms.3c00083
  17. Maria Mar Aparicio-Muriana, Renato Bruni, Francisco J. Lara, Monsalud del Olmo-Iruela, Maykel Hernandez-Mesa, Ana M. García-Campaña, Chiara Dall’Asta, Laura Righetti. Implementing the Use of Collision Cross Section Database for Phycotoxin Screening Analysis. Journal of Agricultural and Food Chemistry 2023, 71 (26) , 10178-10189. https://doi.org/10.1021/acs.jafc.3c01060
  18. Dylan Ross, Aivett Bilbao, Joon-Yong Lee, Xueyun Zheng. mzapy: An Open-Source Python Library Enabling Efficient Extraction and Processing of Ion Mobility Spectrometry-Mass Spectrometry Data in the MZA File Format. Analytical Chemistry 2023, 95 (25) , 9428-9431. https://doi.org/10.1021/acs.analchem.3c01653
  19. Samuel Cajahuaringa, Daniel L. Z. Caetano, Leandro N. Zanotto, Guido Araujo, Munir S. Skaf. MassCCS: A High-Performance Collision Cross-Section Software for Large Macromolecular Assemblies. Journal of Chemical Information and Modeling 2023, 63 (11) , 3557-3566. https://doi.org/10.1021/acs.jcim.3c00405
  20. Carter K. Asef, Markace A. Rainey, Brianna M. Garcia, Goncalo J. Gouveia, Amanda O. Shaver, Franklin E. Leach, III, Alison M. Morse, Arthur S. Edison, Lauren M. McIntyre, Facundo M. Fernández. Unknown Metabolite Identification Using Machine Learning Collision Cross-Section Prediction and Tandem Mass Spectrometry. Analytical Chemistry 2023, 95 (2) , 1047-1056. https://doi.org/10.1021/acs.analchem.2c03749
  21. Longchan Liu, Ziying Wang, Qian Zhang, Yuqi Mei, Linnan Li, Huwei Liu, Zhengtao Wang, Li Yang. Ion Mobility Mass Spectrometry for the Separation and Characterization of Small Molecules. Analytical Chemistry 2023, 95 (1) , 134-151. https://doi.org/10.1021/acs.analchem.2c02866
  22. Alberto Celma, Richard Bade, Juan Vicente Sancho, Félix Hernandez, Melissa Humphries, Lubertus Bijlsma. Prediction of Retention Time and Collision Cross Section (CCSH+, CCSH–, and CCSNa+) of Emerging Contaminants Using Multiple Adaptive Regression Splines. Journal of Chemical Information and Modeling 2022, 62 (22) , 5425-5434. https://doi.org/10.1021/acs.jcim.2c00847
  23. Narumol Jariyasopit, Suphitcha Limjiasahapong, Alongkorn Kurilung, Sitanan Sartyoungkul, Pattipong Wisanpitayakorn, Narong Nuntasaen, Chutima Kuhakarn, Vichai Reutrakul, Prasat Kittakoop, Yongyut Sirivatanauksorn, Sakda Khoomrung. Traveling Wave Ion Mobility-Derived Collision Cross Section Database for Plant Specialized Metabolites: An Application to Ventilago harmandiana Pierre. Journal of Proteome Research 2022, 21 (10) , 2481-2492. https://doi.org/10.1021/acs.jproteome.2c00413
  24. Xue-Chao Song, Elena Canellas, Nicola Dreolin, Jeff Goshawk, Cristina Nerin. Identification of Nonvolatile Migrates from Food Contact Materials Using Ion Mobility–High-Resolution Mass Spectrometry and in Silico Prediction Tools. Journal of Agricultural and Food Chemistry 2022, 70 (30) , 9499-9508. https://doi.org/10.1021/acs.jafc.2c03615
  25. Bailey S. Rose, Jody C. May, Allison R. Reardon, John A. McLean. Collision Cross-Section Calibration Strategy for Lipid Measurements in SLIM-Based High-Resolution Ion Mobility. Journal of the American Society for Mass Spectrometry 2022, 33 (7) , 1229-1237. https://doi.org/10.1021/jasms.2c00067
  26. Xue-Chao Song, Nicola Dreolin, Elena Canellas, Jeff Goshawk, Cristina Nerin. Prediction of Collision Cross-Section Values for Extractables and Leachables from Plastic Products. Environmental Science & Technology 2022, 56 (13) , 9463-9473. https://doi.org/10.1021/acs.est.2c02853
  27. David Izquierdo-Sandoval, David Fabregat-Safont, Leticia Lacalle-Bergeron, Juan V. Sancho, Félix Hernández, Tania Portoles. Benefits of Ion Mobility Separation in GC-APCI-HRMS Screening: From the Construction of a CCS Library to the Application to Real-World Samples. Analytical Chemistry 2022, 94 (25) , 9040-9047. https://doi.org/10.1021/acs.analchem.2c01118
  28. MaKayla Foster, Markace Rainey, Chandler Watson, James N. Dodds, Kaylie I. Kirkwood, Facundo M. Fernández, Erin S. Baker. Uncovering PFAS and Other Xenobiotics in the Dark Metabolome Using Ion Mobility Spectrometry, Mass Defect Analysis, and Machine Learning. Environmental Science & Technology 2022, 56 (12) , 9133-9143. https://doi.org/10.1021/acs.est.2c00201
  29. Dylan H. Ross, Ryan P. Seguin, Allison M. Krinsky, Libin Xu. High-Throughput Measurement and Machine Learning-Based Prediction of Collision Cross Sections for Drugs and Drug Metabolites. Journal of the American Society for Mass Spectrometry 2022, 33 (6) , 1061-1072. https://doi.org/10.1021/jasms.2c00111
  30. Amber D. Rolland, James S. Prell. Approaches to Heterogeneity in Native Mass Spectrometry. Chemical Reviews 2022, 122 (8) , 7909-7951. https://doi.org/10.1021/acs.chemrev.1c00696
  31. Xue-Chao Song, Elena Canellas, Nicola Dreolin, Jeff Goshawk, Cristina Nerin. A Collision Cross Section Database for Extractables and Leachables from Food Contact Materials. Journal of Agricultural and Food Chemistry 2022, 70 (14) , 4457-4466. https://doi.org/10.1021/acs.jafc.2c00724
  32. Xue-Chao Song, Nicola Dreolin, Tito Damiani, Elena Canellas, Cristina Nerin. Prediction of Collision Cross Section Values: Application to Non-Intentionally Added Substance Identification in Food Contact Materials. Journal of Agricultural and Food Chemistry 2022, 70 (4) , 1272-1281. https://doi.org/10.1021/acs.jafc.1c06989
  33. John R. F. B. Connolly, Jordi Munoz-Muriedas, Cris Lapthorn, David Higton, Johannes P. C. Vissers, Alison Webb, Claire Beaumont, Gordon J. Dear. Investigation into Small Molecule Isomeric Glucuronide Metabolite Differentiation Using In Silico and Experimental Collision Cross-Section Values. Journal of the American Society for Mass Spectrometry 2021, 32 (8) , 1976-1986. https://doi.org/10.1021/jasms.0c00427
  34. Leicheng Zhang, Tengfei Xu, Jingtao Zhang, Stephen Choong Chee Wong, Mark Ritchie, Han Wei Hou, Yulan Wang. Single Cell Metabolite Detection Using Inertial Microfluidics-Assisted Ion Mobility Mass Spectrometry. Analytical Chemistry 2021, 93 (30) , 10462-10468. https://doi.org/10.1021/acs.analchem.1c00106
  35. Cameron N. Naylor, Brian H. Clowers. Reevaluating the Role of Polarizability in Ion Mobility Spectrometry. Journal of the American Society for Mass Spectrometry 2021, 32 (3) , 618-627. https://doi.org/10.1021/jasms.0c00338
  36. Corey D. Broeckling, Linxing Yao, Giorgis Isaac, Marisa Gioioso, Valentin Ianchis, Johannes P.C. Vissers. Application of Predicted Collisional Cross Section to Metabolome Databases to Probabilistically Describe the Current and Future Ion Mobility Mass Spectrometry. Journal of the American Society for Mass Spectrometry 2021, 32 (3) , 661-669. https://doi.org/10.1021/jasms.0c00375
  37. Evelyn Rampler, Yasin El Abiead, Harald Schoeny, Mate Rusz, Felina Hildebrand, Veronika Fitz, Gunda Koellensperger. Recurrent Topics in Mass Spectrometry-Based Metabolomics and Lipidomics—Standardization, Coverage, and Throughput. Analytical Chemistry 2021, 93 (1) , 519-545. https://doi.org/10.1021/acs.analchem.0c04698
  38. Daniela Mesa Sanchez, Steve Creger, Veerupaksh Singla, Ruwan T. Kurulugama, John Fjeldsted, Julia Laskin. Ion Mobility-Mass Spectrometry Imaging Workflow. Journal of the American Society for Mass Spectrometry 2020, 31 (12) , 2437-2442. https://doi.org/10.1021/jasms.0c00142
  39. Dylan H. Ross, Jang Ho Cho, Rutan Zhang, Kelly M. Hines, Libin Xu. LiPydomics: A Python Package for Comprehensive Prediction of Lipid Collision Cross Sections and Retention Times and Analysis of Ion Mobility-Mass Spectrometry-Based Lipidomics Data. Analytical Chemistry 2020, 92 (22) , 14967-14975. https://doi.org/10.1021/acs.analchem.0c02560
  40. Laura Righetti, Nicola Dreolin, Alberto Celma, Mike McCullagh, Gitte Barknowitz, Juan V. Sancho, Chiara Dall’Asta. Travelling Wave Ion Mobility-Derived Collision Cross Section for Mycotoxins: Investigating Interlaboratory and Interplatform Reproducibility. Journal of Agricultural and Food Chemistry 2020, 68 (39) , 10937-10943. https://doi.org/10.1021/acs.jafc.0c04498
  41. Jaqueline A. Picache, Jody C. May, John A. McLean. Chemical Class Prediction of Unknown Biomolecules Using Ion Mobility-Mass Spectrometry and Machine Learning: Supervised Inference of Feature Taxonomy from Ensemble Randomization. Analytical Chemistry 2020, 92 (15) , 10759-10767. https://doi.org/10.1021/acs.analchem.0c02137
  42. Manon Meunier, Martina Haack, Dania Awad, Thomas Brück, Khalijah Awang, Marc Litaudon, Frédéric Saubion, Marc Legeay, Dimitri Bréard, David Guilet, Séverine Derbré, Andreas Schinkovitz. Matrix free laser desorption ionization coupled to trapped ion mobility mass spectrometry: an innovative approach for isomer differentiation and molecular network visualization. Talanta 2025, 287 , 127626. https://doi.org/10.1016/j.talanta.2025.127626
  43. Henrik Hupatz, Ida Rahu, Wei-Chieh Wang, Pilleriin Peets, Emma H. Palm, Anneli Kruve. Critical review on in silico methods for structural annotation of chemicals detected with LC/HRMS non-targeted screening. Analytical and Bioanalytical Chemistry 2025, 417 (3) , 473-493. https://doi.org/10.1007/s00216-024-05471-x
  44. Soumyadeep Sarkar, Xueyun Zheng, Geremy C. Clair, Yu Mi Kwon, Youngki You, Adam C. Swensen, Bobbie-Jo M. Webb-Robertson, Ernesto S. Nakayasu, Wei-Jun Qian, Thomas O. Metz. Exploring new frontiers in type 1 diabetes through advanced mass-spectrometry-based molecular measurements. Trends in Molecular Medicine 2024, 30 (12) , 1137-1151. https://doi.org/10.1016/j.molmed.2024.07.009
  45. Chloe Engler Hart, António José Preto, Shaurya Chanana, David Healey, Tobias Kind, Daniel Domingo-Fernández. Evaluating the generalizability of graph neural networks for predicting collision cross section. Journal of Cheminformatics 2024, 16 (1) https://doi.org/10.1186/s13321-024-00899-w
  46. Dmitriy D. Matyushin, Ivan A. Burov, Anastasia Yu. Sholokhova. Uncertainty Quantification and Flagging of Unreliable Predictions in Predicting Mass Spectrometry-Related Properties of Small Molecules Using Machine Learning. International Journal of Molecular Sciences 2024, 25 (23) , 13077. https://doi.org/10.3390/ijms252313077
  47. Laura Schlüter, Kine Østnes Hansen, Johan Isaksson, Jeanette Hammer Andersen, Espen Holst Hansen, Jörn Kalinowski, Yannik Karl-Heinz Schneider. Discovery of thiazostatin D/E using UPLC-HR-MS2-based metabolomics and σ-factor engineering of Actinoplanes sp. SE50/110. Frontiers in Bioengineering and Biotechnology 2024, 12 https://doi.org/10.3389/fbioe.2024.1497138
  48. Mehmet Atakay, Hacı Mehmet Kayılı, Ülkü Güler, Bekir Salih. Ion Mobility-Mass Spectrometry for Macromolecule Analysis. 2024, 1-35. https://doi.org/10.2174/9789815050059122020003
  49. Jakob Koch, Lukas Neumann, Katharina Lackner, Monica L. Fernández-Quintero, Katrin Watschinger, Markus A. Keller. Trapped ion mobility spectrometry-guided molecular discrimination between plasmalogens and other ether lipids in lipidomics experiments. 2024https://doi.org/10.1101/2024.10.23.619801
  50. Ting Xie, Qiong Yang, Jinyu Sun, Hailiang Zhang, Yue Wang, Zhimin Zhang, Hongmei Lu. Large-scale prediction of collision cross-section with very deep graph convolutional network for small molecule identification. Chemometrics and Intelligent Laboratory Systems 2024, 252 , 105177. https://doi.org/10.1016/j.chemolab.2024.105177
  51. Thomas O. Metz, Christine H. Chang, Vasuk Gautam, Afia Anjum, Siyang Tian, Fei Wang, Sean M. Colby, Jamie R. Nunez, Madison R. Blumer, Arthur S. Edison, Oliver Fiehn, Dean P. Jones, Shuzhao Li, Edward T. Morgan, Gary J. Patti, Dylan H. Ross, Madelyn R. Shapiro, Antony J. Williams, David S. Wishart. Introducing ‘identification probability’ for automated and transferable assessment of metabolite identification confidence in metabolomics and related studies. 2024https://doi.org/10.1101/2024.07.30.605945
  52. Robbin Bouwmeester, Keith Richardson, Richard Denny, Ian D. Wilson, Sven Degroeve, Lennart Martens, Johannes P.C. Vissers. Predicting ion mobility collision cross sections and assessing prediction variation by combining conventional and data driven modeling. Talanta 2024, 274 , 125970. https://doi.org/10.1016/j.talanta.2024.125970
  53. Xingliang He, Xing Guo, Fulong Deng, Pengyu Zeng, Bin Wu, Hong'en Sun, Zhongjun Zhao, Yixiang Duan. A study of the transient gas flow affected ion transmission in atmospheric pressure interfaces based on large eddy simulation for electrospray ionization mass spectrometry. Talanta 2024, 274 , 125980. https://doi.org/10.1016/j.talanta.2024.125980
  54. Dylan H. Ross, Harsh Bhotika, Xueyun Zheng, Richard D. Smith, Kristin E. Burnum‐Johnson, Aivett Bilbao. Computational tools and algorithms for ion mobility spectrometry‐mass spectrometry. PROTEOMICS 2024, 24 (12-13) https://doi.org/10.1002/pmic.202200436
  55. Hongda Wang, Lin Zhang, Xiaohang Li, Mengxiao Sun, Meiting Jiang, Xiaojian Shi, Xiaoyan Xu, Mengxiang Ding, Boxue Chen, Heshui Yu, Zheng Li, Dean Guo, Wenzhi Yang. Machine learning prediction for constructing a universal multidimensional information library of Panax saponins (ginsenosides). Food Chemistry 2024, 439 , 138106. https://doi.org/10.1016/j.foodchem.2023.138106
  56. Alongkorn Kurilung, Suphitcha Limjiasahapong, Khwanta Kaewnarin, Pattipong Wisanpitayakorn, Narumol Jariyasopit, Kwanjeera Wanichthanarak, Sitanan Sartyoungkul, Stephen Choong Chee Wong, Nuankanya Sathirapongsasuti, Chagriya Kitiyakara, Yongyut Sirivatanauksorn, Sakda Khoomrung. Measurement of very low-molecular weight metabolites by traveling wave ion mobility and its use in human urine samples. Journal of Pharmaceutical Analysis 2024, 14 (5) , 100921. https://doi.org/10.1016/j.jpha.2023.12.011
  57. Francisco José Díaz-Galiano, María Murcia-Morales, Víctor Cutillas, Amadeo R. Fernández-Alba. Ions on the move: The combination of ion mobility and food metabolomics. Trends in Food Science & Technology 2024, 147 , 104446. https://doi.org/10.1016/j.tifs.2024.104446
  58. Ignacio Pérez-Victoria. Natural Products Dereplication: Databases and Analytical Methods. 2024, 1-56. https://doi.org/10.1007/978-3-031-59567-7_1
  59. Guillermo Ramajo, Constantino García, Alberto Gil, Abraham Otero. Training Deep Learning Neural Networks for Predicting CCS Using the METLIN-CCS Dataset. 2024, 225-236. https://doi.org/10.1007/978-3-031-64636-2_17
  60. Orobola E. Olajide, Kimberly Y. Kartowikromo, Ahmed M. Hamid. Ion Mobility Mass Spectrometry: Instrumentation and Applications. 2023https://doi.org/10.5772/intechopen.1002767
  61. Mingdu Luo, Yandong Yin, Zhiwei Zhou, Haosong Zhang, Xi Chen, Hongmiao Wang, Zheng-Jiang Zhu. A mass spectrum-oriented computational method for ion mobility-resolved untargeted metabolomics. Nature Communications 2023, 14 (1) https://doi.org/10.1038/s41467-023-37539-0
  62. Renfeng Guo, Youjia Zhang, Yuxuan Liao, Qiong Yang, Ting Xie, Xiaqiong Fan, Zhonglong Lin, Yi Chen, Hongmei Lu, Zhimin Zhang. Highly accurate and large-scale collision cross sections prediction with graph neural networks. Communications Chemistry 2023, 6 (1) https://doi.org/10.1038/s42004-023-00939-w
  63. Juliane Hollender, Emma L. Schymanski, Lutz Ahrens, Nikiforos Alygizakis, Frederic Béen, Lubertus Bijlsma, Andrea M. Brunner, Alberto Celma, Aurelie Fildier, Qiuguo Fu, Pablo Gago-Ferrero, Ruben Gil-Solsona, Peter Haglund, Martin Hansen, Sarit Kaserzon, Anneli Kruve, Marja Lamoree, Christelle Margoum, Jeroen Meijer, Sylvain Merel, Cassandra Rauert, Pawel Rostkowski, Saer Samanipour, Bastian Schulze, Tobias Schulze, Randolph R. Singh, Jaroslav Slobodnik, Teresa Steininger-Mairinger, Nikolaos S. Thomaidis, Anne Togola, Katrin Vorkamp, Emmanuelle Vulliet, Linyan Zhu, Martin Krauss. NORMAN guidance on suspect and non-target screening in environmental monitoring. Environmental Sciences Europe 2023, 35 (1) https://doi.org/10.1186/s12302-023-00779-4
  64. Amy Li, Libin Xu. MALDI-IM-MS Imaging of Brain Sterols and Lipids in a Mouse Model of Smith-Lemli-Opitz Syndrome. 2023https://doi.org/10.1101/2023.10.02.560415
  65. Rutan Zhang, Nate K. Ashford, Amy Li, Dylan H. Ross, Brian J. Werth, Libin Xu. High-throughput analysis of lipidomic phenotypes of methicillin-resistant Staphylococcus aureus by coupling in situ 96-well cultivation and HILIC-ion mobility-mass spectrometry. Analytical and Bioanalytical Chemistry 2023, 415 (25) , 6191-6199. https://doi.org/10.1007/s00216-023-04890-6
  66. Jia-Hui Wen, An-Qi Guo, Meng-Ning Li, Hua Yang. A structural similarity networking assisted collision cross-section prediction interval filtering strategy for multi-compound identification of complex matrix by ion-mobility mass spectrometry. Analytica Chimica Acta 2023, 1278 , 341720. https://doi.org/10.1016/j.aca.2023.341720
  67. Anne Miller, Elisa M. York, Sylwia A. Stopka, Juan Ramón Martínez-François, Md Amin Hossain, Gerard Baquer, Michael S. Regan, Nathalie Y. R. Agar, Gary Yellen. Spatially resolved metabolomics and isotope tracing reveal dynamic metabolic responses of dentate granule neurons with acute stimulation. Nature Metabolism 2023, 5 (10) , 1820-1835. https://doi.org/10.1038/s42255-023-00890-z
  68. Kimberly Y. Kartowikromo, Orobola E. Olajide, Ahmed M. Hamid. Collision cross section measurement and prediction methods in omics. Journal of Mass Spectrometry 2023, 58 (9) https://doi.org/10.1002/jms.4973
  69. Ian Ramtanon, Alexandra Berlioz-Barbier, Simon Remy, Jean-Hugues Renault, Agnès Le Masle. A combined liquid chromatography – trapped ion mobility – tandem high-resolution mass spectrometry and multivariate analysis approach for the determination of enzymatic reactivity descriptors in biomass hydrolysates. Journal of Chromatography A 2023, 1706 , 464277. https://doi.org/10.1016/j.chroma.2023.464277
  70. Randolph R. Singh, Yann Aminot, Karine Héas-Moisan, Hugues Preud'homme, Catherine Munschy. Cracked and shucked: GC-APCI-IMS-HRMS facilitates identification of unknown halogenated organic chemicals in French marine bivalves. Environment International 2023, 178 , 108094. https://doi.org/10.1016/j.envint.2023.108094
  71. Dylan H. Ross, Aivett Bilbao, Richard D. Smith, Xueyun Zheng. Ion Mobility Spectrometry‐Mass Spectrometry for High‐Throughput Analysis. 2023, 183-213. https://doi.org/10.1002/9781119678496.ch6
  72. Anne Miller, Elisa York, Sylwia Stopka, Juan Martínez-François, Md Amin Hossain, Gerard Baquer, Michael Regan, Nathalie Agar, Gary Yellen. Spatially resolved metabolomics and isotope tracing reveal dynamic metabolic responses of dentate granule neurons with acute stimulation. 2023https://doi.org/10.21203/rs.3.rs-2276903/v1
  73. He Xingliang, Guo Xing, Wu Mengfan, Deng Fulong, Zeng Pengyu, Zhao Zhongjun, Duan Yixiang. From droplets to ions: a comprehensive and consecutive ion formation modelling in atmosphere pressure interface of electrospray ionization mass spectrometry. The Analyst 2023, 148 (14) , 3174-3178. https://doi.org/10.1039/D3AN00607G
  74. Noelle Reimers, Quynh Do, Rutan Zhang, Angela Guo, Ryan Ostrander, Alyson Shoji, Chau Vuong, Libin Xu. Tracking the Metabolic Fate of Exogenous Arachidonic Acid in Ferroptosis Using Dual-Isotope Labeling Lipidomics. 2023https://doi.org/10.1101/2023.05.28.542640
  75. Elias Iturrospe, Rani Robeyns, Katyeny Manuela da Silva, Maria van de Lavoir, Joost Boeckmans, Tamara Vanhaecke, Alexander L. N. van Nuijs, Adrian Covaci. Metabolic signature of HepaRG cells exposed to ethanol and tumor necrosis factor alpha to study alcoholic steatohepatitis by LC–MS-based untargeted metabolomics. Archives of Toxicology 2023, 97 (5) , 1335-1353. https://doi.org/10.1007/s00204-023-03470-y
  76. Xiaohang Li, Hongda Wang, Meiting Jiang, Mengxiang Ding, Xiaoyan Xu, Bei Xu, Yadan Zou, Yuetong Yu, Wenzhi Yang. Collision Cross Section Prediction Based on Machine Learning. Molecules 2023, 28 (10) , 4050. https://doi.org/10.3390/molecules28104050
  77. Kaylie I. Kirkwood, Melanie T. Odenkirk, Erin S. Baker. Ion Mobility Spectrometry. 2023, 151-182. https://doi.org/10.1002/9783527836512.ch6
  78. Nikiforos Alygizakis, Francois Lestremau, Pablo Gago-Ferrero, Rubén Gil-Solsona, Katarzyna Arturi, Juliane Hollender, Emma L. Schymanski, Valeria Dulio, Jaroslav Slobodnik, Nikolaos S. Thomaidis. Towards a harmonized identification scoring system in LC-HRMS/MS based non-target screening (NTS) of emerging contaminants. TrAC Trends in Analytical Chemistry 2023, 159 , 116944. https://doi.org/10.1016/j.trac.2023.116944
  79. Marie Lenski, Saïd Maallem, Gianni Zarcone, Guillaume Garçon, Jean-Marc Lo-Guidice, Sébastien Anthérieu, Delphine Allorge. Prediction of a Large-Scale Database of Collision Cross-Section and Retention Time Using Machine Learning to Reduce False Positive Annotations in Untargeted Metabolomics. Metabolites 2023, 13 (2) , 282. https://doi.org/10.3390/metabo13020282
  80. Alberto Celma. The Use of Ion Mobility Separation as an Additional Dimension for the Screening of Organic Micropollutants in Environmental Samples. 2023https://doi.org/10.1007/698_2023_1055
  81. Katyeny Manuela da Silva, Maria van de Lavoir, Rani Robeyns, Elias Iturrospe, Lisa Verheggen, Adrian Covaci, Alexander L. N. van Nuijs. Guidelines and considerations for building multidimensional libraries for untargeted MS-based metabolomics. Metabolomics 2023, 19 (1) https://doi.org/10.1007/s11306-022-01965-w
  82. Yuping Cai, Zhiwei Zhou, Zheng-Jiang Zhu. Advanced analytical and informatic strategies for metabolite annotation in untargeted metabolomics. TrAC Trends in Analytical Chemistry 2023, 158 , 116903. https://doi.org/10.1016/j.trac.2022.116903
  83. Ting Xie, Qiong Yang, Jinyu Sun, Hailiang Zhang, Yue Wang, Zhimin zhang, Hongmei Lu. Large-Scale Prediction of Collision Cross-Section with Graph Convolutional Network for Compound Identification. 2023https://doi.org/10.2139/ssrn.4505380
  84. Robbin Bouwmeester, Keith Richardson, Richard Denny, Ian D. Wilson, Sven Degroeve, Lennart Martens, Johannes PC Vissers. Predicting Ion Mobility Collision Cross Sections and Assessing Prediction Variation by Combining Conventional and Data Driven Modeling. 2023https://doi.org/10.2139/ssrn.4646924
  85. Frank Menger, Alberto Celma, Emma L. Schymanski, Foon Yin Lai, Lubertus Bijlsma, Karin Wiberg, Félix Hernández, Juan V. Sancho, Lutz Ahrens. Enhancing spectral quality in complex environmental matrices: Supporting suspect and non-target screening in zebra mussels with ion mobility. Environment International 2022, 170 , 107585. https://doi.org/10.1016/j.envint.2022.107585
  86. Hiba Mohammed Taha, Reza Aalizadeh, Nikiforos Alygizakis, Jean-Philippe Antignac, Hans Peter H. Arp, Richard Bade, Nancy Baker, Lidia Belova, Lubertus Bijlsma, Evan E. Bolton, Werner Brack, Alberto Celma, Wen-Ling Chen, Tiejun Cheng, Parviel Chirsir, Ľuboš Čirka, Lisa A. D’Agostino, Yannick Djoumbou Feunang, Valeria Dulio, Stellan Fischer, Pablo Gago-Ferrero, Aikaterini Galani, Birgit Geueke, Natalia Głowacka, Juliane Glüge, Ksenia Groh, Sylvia Grosse, Peter Haglund, Pertti J. Hakkinen, Sarah E. Hale, Felix Hernandez, Elisabeth M.-L. Janssen, Tim Jonkers, Karin Kiefer, Michal Kirchner, Jan Koschorreck, Martin Krauss, Jessy Krier, Marja H. Lamoree, Marion Letzel, Thomas Letzel, Qingliang Li, James Little, Yanna Liu, David M. Lunderberg, Jonathan W. Martin, Andrew D. McEachran, John A. McLean, Christiane Meier, Jeroen Meijer, Frank Menger, Carla Merino, Jane Muncke, Matthias Muschket, Michael Neumann, Vanessa Neveu, Kelsey Ng, Herbert Oberacher, Jake O’Brien, Peter Oswald, Martina Oswaldova, Jaqueline A. Picache, Cristina Postigo, Noelia Ramirez, Thorsten Reemtsma, Justin Renaud, Pawel Rostkowski, Heinz Rüdel, Reza M. Salek, Saer Samanipour, Martin Scheringer, Ivo Schliebner, Wolfgang Schulz, Tobias Schulze, Manfred Sengl, Benjamin A. Shoemaker, Kerry Sims, Heinz Singer, Randolph R. Singh, Mark Sumarah, Paul A. Thiessen, Kevin V. Thomas, Sonia Torres, Xenia Trier, Annemarie P. van Wezel, Roel C. H. Vermeulen, Jelle J. Vlaanderen, Peter C. von der Ohe, Zhanyun Wang, Antony J. Williams, Egon L. Willighagen, David S. Wishart, Jian Zhang, Nikolaos S. Thomaidis, Juliane Hollender, Jaroslav Slobodnik, Emma L. Schymanski. The NORMAN Suspect List Exchange (NORMAN-SLE): facilitating European and worldwide collaboration on suspect screening in high resolution mass spectrometry. Environmental Sciences Europe 2022, 34 (1) https://doi.org/10.1186/s12302-022-00680-6
  87. Lidia Belova, Alberto Celma, Glenn Van Haesendonck, Filip Lemière, Juan Vicente Sancho, Adrian Covaci, Alexander L.N. van Nuijs, Lubertus Bijlsma. Revealing the differences in collision cross section values of small organic molecules acquired by different instrumental designs and prediction models. Analytica Chimica Acta 2022, 1229 , 340361. https://doi.org/10.1016/j.aca.2022.340361
  88. Fan Yang, Denice van Herwerden, Hugues Preud’homme, Saer Samanipour. Collision Cross Section Prediction with Molecular Fingerprint Using Machine Learning. Molecules 2022, 27 (19) , 6424. https://doi.org/10.3390/molecules27196424
  89. Jinmei Xia, Wenhai Xiao, Xihuang Lin, Yiduo Zhou, Peng Qiu, Hongkun Si, Xiaorong Wu, Siwen Niu, Zhuhua Luo, Xianwen Yang. Ion Mobility-Derived Collision Cross-Sections Add Extra Capability in Distinguishing Isomers and Compounds with Similar Retention Times: The Case of Aphidicolanes. Marine Drugs 2022, 20 (9) , 541. https://doi.org/10.3390/md20090541
  90. Carter K. Asef, Markace Rainey, Brianna M. Garcia, Goncalo J. Gouveia, Amanda O. Shaver, Franklin E. Leach, Allison M. Morse, Arthur S. Edison, Lauren M. McIntyre, Facundo M. Fernández. Ion Mobility for Unknown Metabolite Identification: Hope or Hype?. 2022https://doi.org/10.1101/2022.08.26.505158
  91. María Moran‐Garrido, Sandra. M. Camunas‐Alberca, Alberto Gil‐de‐la Fuente, Antonio Mariscal, Ana Gradillas, Coral Barbas, Jorge Sáiz. Recent developments in data acquisition, treatment and analysis with ion mobility‐mass spectrometry for lipidomics. PROTEOMICS 2022, 22 (15-16) https://doi.org/10.1002/pmic.202100328
  92. Esra te Brinke, Ane Arrizabalaga-Larrañaga, Marco H. Blokland. Insights of ion mobility spectrometry and its application on food safety and authenticity: A review. Analytica Chimica Acta 2022, 1222 , 340039. https://doi.org/10.1016/j.aca.2022.340039
  93. Amy Li, Kelly M. Hines, Dylan H. Ross, James W. MacDonald, Libin Xu. Temporal changes in the brain lipidome during neurodevelopment of Smith–Lemli–Opitz syndrome mice. The Analyst 2022, 147 (8) , 1611-1621. https://doi.org/10.1039/D2AN00137C
  94. Yue-xin Qian, Dong-xue Zhao, Hong-da Wang, He Sun, Ying Xiong, Xiao-yan Xu, Wan-di Hu, Mei-yu Liu, Bo-xue Chen, Ying Hu, Xue Li, Mei-ting Jiang, Wen-zhi Yang, Xiu-mei Gao. An ion mobility-enabled and high-efficiency hybrid scan approach in combination with ultra-high performance liquid chromatography enabling the comprehensive characterization of the multicomponents from Carthamus tinctorius. Journal of Chromatography A 2022, 1667 , 462904. https://doi.org/10.1016/j.chroma.2022.462904
  95. Vanessa Hinnenkamp, Peter Balsaa, Torsten C. Schmidt. Target, suspect and non-target screening analysis from wastewater treatment plant effluents to drinking water using collision cross section values as additional identification criterion. Analytical and Bioanalytical Chemistry 2022, 414 (1) , 425-438. https://doi.org/10.1007/s00216-021-03263-1
  96. Frank Menger, Alberto Celma, Emma L. Schymanski, Foon Yin Lai, Lubertus Bijlsma, Karin Wiberg, Félix Hernández, Juan Vicente Sancho, Ahrens Lutz. Enhancing Spectral Quality in Complex Environmental Matrices: Supporting Suspect and Non-Target Screening in Zebra Mussels with Ion Mobility. SSRN Electronic Journal 2022, 165 https://doi.org/10.2139/ssrn.4149383
  97. Tlou T. Mosekiemang, Maria A. Stander, André de Villiers. Ultra-high pressure liquid chromatography coupled to travelling wave ion mobility-time of flight mass spectrometry for the screening of pharmaceutical metabolites in wastewater samples: Application to antiretrovirals. Journal of Chromatography A 2021, 1660 , 462650. https://doi.org/10.1016/j.chroma.2021.462650
  98. Katyeny Manuela da Silva, Elias Iturrospe, Joris Heyrman, Jeremy P. Koelmel, Matthias Cuykx, Tamara Vanhaecke, Adrian Covaci, Alexander L.N. van Nuijs. Optimization of a liquid chromatography-ion mobility-high resolution mass spectrometry platform for untargeted lipidomics and application to HepaRG cell extracts. Talanta 2021, 235 , 122808. https://doi.org/10.1016/j.talanta.2021.122808
  99. Emma L. Schymanski, Todor Kondić, Steffen Neumann, Paul A. Thiessen, Jian Zhang, Evan E. Bolton. Empowering large chemical knowledge bases for exposomics: PubChemLite meets MetFrag. Journal of Cheminformatics 2021, 13 (1) https://doi.org/10.1186/s13321-021-00489-0
  100. Armin Sebastian Guntner, Thomas Bögl, Franz Mlynek, Wolfgang Buchberger. Large-Scale Evaluation of Collision Cross Sections to Investigate Blood-Brain Barrier Permeation of Drugs. Pharmaceutics 2021, 13 (12) , 2141. https://doi.org/10.3390/pharmaceutics13122141
Download PDF
Go to Analytical Chemistry
Get e-Alerts
Get e-Alerts

Analytical Chemistry

Cite this: Anal. Chem. 2020, 92, 6, 4548–4557
Click to copy citationCitation copied!
https://doi.org/10.1021/acs.analchem.9b05772
Published February 25, 2020

Copyright © 2020 American Chemical Society. This publication is licensed under these Terms of Use.

Request reuse permissions

Article Views

3722

Altmetric

-

Citations

100
Learn about these metrics

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.

Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.

The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated.

  • Abstract

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 1

    Figure 1. (A) Counts and (B) comparison of agreement between measurements present in multiple sources. Agreement for all overlapping CCS values in blue. Agreement between DTIM CCS values in red. Agreement between TWIM CCS values in purple. Agreement between DTIM and TWIM CCS values in gold.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 2

    Figure 2. PCA projections of full CCS database onto principal axes 1, 2, and 3, colored by data set (A,B) or chemical classification (C,D). Correlation of the top three molecular descriptors contributing to separation along PC1 (E–G) and PC2 (H–J). hac = heavy atom count; m/z = mass to charge ratio; ao = acyclic oxygen count; hbd = H-bond donor atoms; ctv = cyclic trivalent nodes; r6 = 6-membered ring count.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 3

    Figure 3. PLS-RA projections of full CCS database onto axes 1, 2, colored by data set (A) or chemical classification (B). Correlation between PLS-RA projections along axis 1 and PCA projections along PC1 (C). Correlation between molecular descriptors and PLS-RA projections along axis 1 (blue) or CCS (red) for all compounds (D–K). hac = heavy atom count; m/z = mass to charge ratio; hbam = H-bond acceptor sites; c = carbon atom count; ao = acyclic oxygen count; asb = acyclic single bonds; asv = acyclic single valent nodes; adb = acyclic double bonds.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 4

    Figure 4. PCA projections of full CCS database onto principal axes 1, 2, and 3, colored by chemical class label (A, B), or by cluster (C, D). (E) Plot of CCS vs m/z for full CCS database, colored by cluster. (F) Central structures within each cluster. (G–I) Average predictive performance of models (lasso, forest, svr, respectively) by MDAE and RMSE from five independent trials, trained on the full CCS database (training set = blue, test set = red) or on individual cluster data sets (training set = purple, test set = gold).

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 5

    Figure 5. Workflow describing the process for training and validating the final prediction model. First, MQNs are generated from the compound SMILES structure in addition to the m/z and MS adduct, and this data is stored in a database. The complete data set is randomly partitioned into a training set and test set, preserving the approximate distribution of CCS values between the two sets. The training set is then fit using K-Means clustering to find the dominant groupings within the data set in terms of chemical similarity. The data from each assigned cluster is then used to train an individual predictive model that is specialized for that group of compounds. Finally, the overall CCS prediction performance of this set of models is validated using the test set data by first assigning each sample to one of the fitted clusters then predicting CCS using the corresponding predictive model.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 6

    Figure 6. (A–C) Complete performance metrics for final predictive model on training (blue) and test (red) data. (A) R2 (B) mean/median absolute error and root mean squared error (C) proportion of predictions falling within 1, 3, 5, and 10% of reference values. (D–F) Comparison of CCS prediction performance between final model (purple) and DeepCCS (gold) on all data sets used for training DeepCCS. (D) R2 (E) mean/median absolute error and root mean squared error (F) mean/median relative error (MRE and MDRE).

    High Resolution Image
    Download MS PowerPoint Slide

  • References

    This article references 46 other publications.

    1. 1
      Prakash, C.; Shaffer, C. L.; Nedderman, A. Mass Spectrom. Rev. 2007, 26, 340– 369,  DOI: 10.1002/mas.20128
      1
      Analytical strategies for identifying drug metabolites
      Prakash, Chandra; Shaffer, Christopher L.; Nedderman, Angus
      Mass Spectrometry Reviews (2007), 26 (3), 340-369CODEN: MSRVD3; ISSN:0277-7037. (John Wiley & Sons, Inc.)
      A review. With the dramatic increase in the no. of new chem. entities (NCEs) arising from combinatorial chem. and modern high-throughput bioassays, novel bioanal. techniques are required for the rapid detn. of the metabolic stability and metabolites of these NCEs. Knowledge of the metabolic site(s) of the NCEs in early drug discovery is essential for selecting compds. with favorable pharmacokinetic credentials and aiding medicinal chemists in modifying metabolic "soft spots". In development, elucidation of biotransformation pathways of a drug candidate by identifying its circulatory and excretory metabolites is vitally important to understand its physiol. effects. Mass spectrometry (MS) and NMR have played an invaluable role in the structural characterization and quantification of drug metabolites. Indeed, liq. chromatog. (LC) coupled with atm. pressure ionization (API) MS has now become the most powerful tool for the rapid detection, structure elucidation, and quantification of drug-derived material within various biol. fluids. Often, however, MS alone is insufficient to identify the exact position of oxidn., to differentiate isomers, or to provide the precise structure of unusual and/or unstable metabolites. In addn., an excess of endogenous material in biol. samples often suppress the ionization of drug-related material complicating metabolite identification by MS. In these cases, multiple anal. and wet chem. techniques, such as LC-NMR, enzymic hydrolysis, chem. derivatization, and hydrogen/deuterium-exchange (H/D-exchange) combined with MS are used to characterize the novel and isomeric metabolites of drug candidates. This review describes sample prepn. and introduction strategies to minimize ion suppression by biol. matrixes for metabolite identification studies, the application of various LC-tandem MS (LC-MS/MS) techniques for the rapid quantification and identification of drug metabolites, and future trends in this field.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXlvFCgu7Y%253D&md5=0bad5d31fb5d2252e9492253c9a0d172
    2. 2
      Blazenovic, I.; Kind, T.; Ji, J.; Fiehn, O. Metabolites 2018, 8, 31,  DOI: 10.3390/metabo8020031
      2
      Software tools and approaches for compound identification of LC-MS/MS data in metabolomics
      Blazenovic, Ivana; Kind, Tobias; Ji, Jian; Fiehn, Oliver
      Metabolites (2018), 8 (2), 31/1-31/23CODEN: METALU; ISSN:2218-1989. (MDPI AG)
      The annotation of small mols. remains a major challenge in untargeted mass spectrometry-based metabolomics. We here critically discuss structured elucidation approaches and software that are designed to help during the annotation of unknown compds. Only by elucidating unknown metabolites first is it possible to biol. interpret complex systems, to map compds. to pathways and to create reliable predictive metabolic models for translational and clin. research. These strategies include the construction and quality of tandem mass spectral databases such as the coalition of MassBank repositories and investigations of MS/MS matching confidence. We present in silico fragmentation tools such as MS-FINDER, CFM-ID, MetFrag, ChemDistiller and CSI:FingerID that can annotate compds. from existing structure databases and that have been used in the CASMI (crit. assessment of small mol. identification) contests. Furthermore, the use of retention time models from liq. chromatog. and the utility of collision cross-section modeling from ion mobility expts. are covered. Workflows and published examples of successfully annotated unknown compds. are included.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhvV2lt73I&md5=371fd01734c223fb65560768722c830c
    3. 3
      Clemmer, D. E.; Hudgins, R. R.; Jarrold, M. F. J. Am. Chem. Soc. 1995, 117, 10141– 10142,  DOI: 10.1021/ja00145a037
      3
      Naked Protein Conformations: Cytochrome c in the Gas Phase
      Clemmer, David E.; Hudgins, Robert R.; Jarrold, Martin F.
      Journal of the American Chemical Society (1995), 117 (40), 10141-2CODEN: JACSAT; ISSN:0002-7863. (American Chemical Society)
      The conformers present for naked cytochrome c ions in the gas phase were examd. using electrospray/ion mobility measurements. A no. of resolved features that correspond to distinct partially folded conformations were obsd. The abundance of the native conformer was negligible under the conditions employed. The obsd. conformers may be intermediates in the folding process.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXotVyqsL8%253D&md5=6f11c3d8b73925c5c6ff0ad475bd8a86
    4. 4
      von Helden, G.; Wyttenbach, T.; Bowers, M. T. Science 1995, 267, 1483– 1485,  DOI: 10.1126/science.267.5203.1483
      4
      Conformation of macromolecules in the gas phase: use of matrix-assisted laser desorption methods in ion chromatography
      Von Helden, Gert; Wyttenbach, Thomas; Bowers, Michael T.
      Science (Washington, D. C.) (1995), 267 (5203), 1483-5CODEN: SCIEAS; ISSN:0036-8075. (American Association for the Advancement of Science)
      Conformational data of macromols. in the gas phase were obtained by the coupling of matrix-assisted laser desorption ion source to an ion chromatograph. A series of polyethylene glycols (PEG), cationized by sodium ions were studied, in comparison with protonated bradykinin. Detailed modeling of Na+PEG9 with mol. mechanics methods indicates that the lowest energy structure has the Na+ ion "solvated" by the polymer chain with seven oxygen atoms as nearest neighbors. The agreement between the model and expt. is within 1% for Na+PEG9, Na+PEG13, and Na+PEG17, which strongly supported both the method and the deduced structures.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK2MXktlCltL4%253D&md5=6f8d9b91958b623e95eb2a121b7deeb9
    5. 5
      McLean, J. A.; Ruotolo, B. T.; Gillig, K. J.; Russell, D. H. Int. J. Mass Spectrom. 2005, 240, 301– 315,  DOI: 10.1016/j.ijms.2004.10.003
      5
      Ion mobility-mass spectrometry: a new paradigm for proteomics
      McLean, John A.; Ruotolo, Brandon T.; Gillig, Kent J.; Russell, David H.
      International Journal of Mass Spectrometry (2005), 240 (3), 301-315CODEN: IMSPF8; ISSN:1387-3806. (Elsevier B.V.)
      A review. Matrix-assisted laser desorption/ionization (MALDI) coupled with ion mobility-mass spectrometry (IM-MS) provides a rapid (μs-ms) means for the 2-dimensional (2D) sepn. of complex biol. samples (e.g., peptides, oligonucleotides, glycoconjugates, lipids, etc.), elucidation of solvent-free secondary structural elements (e.g., helixes, β-hairpins, random coils, etc.), rapid identification of post-translational modifications (e.g., phosphorylation, glycosylation, etc.) or ligation of small mols., and simultaneous and comprehensive sequencing information of biopolymers. In IM-MS, protein-identification information is complemented by structural characterization data, which is difficult to obtain using conventional proteomic techniques. New avenues for enhancing the figures of merit (e.g., sensitivity, limits of detection, dynamic range, and analyte selectivity) and optimizing IM-MS exptl. parameters are described in the context of deriving new information at the forefront of proteomics research.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2MXktF2rsA%253D%253D&md5=2d6eb1cf9e7418b1203a61ab094adc98
    6. 6
      Kanu, B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H. H., Jr. J. Mass Spectrom. 2008, 43, 1– 22,  DOI: 10.1002/jms.1383
      6
      Ion mobility-mass spectrometry
      Kanu, Abu B.; Dwivedi, Prabha; Tam, Maggie; Matz, Laura; Hill, Herbert H., Jr.
      Journal of Mass Spectrometry (2008), 43 (1), 1-22CODEN: JMSPFJ; ISSN:1076-5174. (John Wiley & Sons Ltd.)
      This review article compares and contrasts various types of ion mobility-mass spectrometers available today and describes their advantages for application to a wide range of analytes. Ion mobility spectrometry (IMS), when coupled with mass spectrometry, offers value-added data not possible from mass spectra alone. Sepn. of isomers, isobars, and conformers; redn. of chem. noise; and measurement of ion size are possible with the addn. of ion mobility cells to mass spectrometers. In addn., structurally similar ions and ions of the same charge state can be sepd. into families of ions which appear along a unique mass-mobility correlation line. This review describes the four methods of ion mobility sepn. currently used with mass spectrometry. They are (1) drift-time ion mobility spectrometry (DTIMS), (2) aspiration ion mobility spectrometry (AIMS), (3) differential-mobility spectrometry (DMS) which is also called field-asym. waveform ion mobility spectrometry (FAIMS) and (4) traveling-wave ion mobility spectrometry (TWIMS). DTIMS provides the highest IMS resolving power and is the only IMS method which can directly measure collision cross-sections. AIMS is a low resoln. mobility sepn. method but can monitor ions in a continuous manner. DMS and FAIMS offer continuous-ion monitoring capability as well as orthogonal ion mobility sepn. in which high-sepn. selectivity can be achieved. TWIMS is a novel method of IMS with a low resolving power but has good sensitivity and is well integrated into a com. mass spectrometer. One hundred and sixty refs. on ion mobility-mass spectrometry (IMMS) are provided.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXisFKntbk%253D&md5=44236351655e694207187c94eeca82e6
    7. 7
      Fenn, L. S.; Kliman, M.; Mahsut, A.; Zhao, S. R.; McLean, J. A. Anal. Bioanal. Chem. 2009, 394, 235– 244,  DOI: 10.1007/s00216-009-2666-3
      7
      Characterizing ion mobility-mass spectrometry conformation space for the analysis of complex biological samples
      Fenn, Larissa S.; Kliman, Michal; Mahsut, Ablatt; Zhao, Sophie R.; McLean, John A.
      Analytical and Bioanalytical Chemistry (2009), 394 (1), 235-244CODEN: ABCNBP; ISSN:1618-2642. (Springer)
      The conformation space occupied by different classes of biomols. measured by ion mobility-mass spectrometry (IM-MS) is described for utility in the characterization of complex biol. samples. Although the qual. sepn. of different classes of biomols. on the basis of structure or collision cross section is known, there is relatively little quant. cross-section information available for species apart from peptides. In this report, collision cross sections are measured for a large suite of biol. salient species, including oligonucleotides (n = 96), carbohydrates (n = 192), and lipids (n = 53), which are compared to reported values for peptides (n = 610). In general, signals for each class are highly correlated, and at a given mass, these correlations result in predicted collision cross sections that increase in the order oligonucleotides < carbohydrates < peptides < lipids. The specific correlations are described by logarithmic regressions, which best approx. the theor. trend of increasing collision cross section as a function of increasing mass. A statistical treatment of the signals obsd. within each mol. class suggests that the breadth of conformation space occupied by each class increases in the order lipids < oligonucleotides < peptides < carbohydrates. The utility of conformation space anal. in the direct anal. of complex biol. samples is described, both in the context of qual. mol. class identification and in fine structure examn. within a class. The latter is demonstrated in IM-MS sepns. of isobaric oligonucleotides, which are interpreted by mol. dynamics simulations.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXisVWqsro%253D&md5=febdb07ff7e0cfb10c0e9b5f22d5fa58
    8. 8
      Pringle, S. D.; Giles, K.; Wildgoose, J. L.; Williams, J. P.; Slade, S. E.; Thalassinos, K.; Bateman, R. H.; Bowers, M. T.; Scrivens, J. H. Int. J. Mass Spectrom. 2007, 261, 1– 12,  DOI: 10.1016/j.ijms.2006.07.021
      8
      An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/travelling wave IMS/oa-ToF instrument
      Pringle, Steven D.; Giles, Kevin; Wildgoose, Jason L.; Williams, Jonathan P.; Slade, Susan E.; Thalassinos, Konstantinos; Bateman, Robert H.; Bowers, Michael T.; Scrivens, James H.
      International Journal of Mass Spectrometry (2007), 261 (1), 1-12CODEN: IMSPF8; ISSN:1387-3806. (Elsevier B.V.)
      Ion mobility coupled with mass spectrometry has evolved into a powerful anal. technique for investigating the gas-phase structures of bio-mols. Here the authors present the mobility sepn. of some peptide and protein ions using a new hybrid quadrupole/travelling wave ion mobility separator/orthogonal acceleration time-of-flight instrument. Comparison of the mobility data obtained from the relatively new travelling wave sepn. device with data obtained using various other mobility separators demonstrate that while the mobility characteristics are similar, the new hybrid instrument geometry provides mobility sepn. without compromising the base sensitivity of the mass spectrometer. This capability facilitates mobility studies of samples at anal. significant levels.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhsFCru7c%253D&md5=1bcf7eba693142d9534f16ec8e978ec7
    9. 9
      May, J. C.; McLean, J. A. Anal. Chem. 2015, 87, 1422– 1436,  DOI: 10.1021/ac504720m
      9
      Ion Mobility-Mass Spectrometry: Time-Dispersive Instrumentation
      May, Jody C.; McLean, John A.
      Analytical Chemistry (Washington, DC, United States) (2015), 87 (3), 1422-1436CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      A review. The field of ion mobility-mass spectrometry (IM-MS) has grown with significant momentum in recent years in both fundamental advances and pioneering applications. A search of the terms "ion mobility" and "mass spectrometry" returns more than 2,000 papers, with over half of these being published in the past four years. This increased interest has been motivated in large part by improved technologies which have enabled contemporary IM-MS to be amendable to a variety of samples in biol. and medicine with high sensitivity, resolving power, and sample throughput.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXitFehs73P&md5=43635f295a34d972fde1f3fbe0cf4b83
    10. 10
      Mesleh, M. F.; Hunter, J. M.; Shvartsburg, A. A.; Schatz, G. C.; Jarrold, M. F. J. Phys. Chem. 1996, 100, 16082– 16086,  DOI: 10.1021/jp961623v
      10
      Structural Information from Ion Mobility Measurements: Effects of the Long-Range Potential
      Mesleh, M. F.; Hunter, J. M.; Shvartsburg, A. A.; Schatz, G. C.; Jarrold, M. F.
      Journal of Physical Chemistry (1996), 100 (40), 16082-16086CODEN: JPCHAX; ISSN:0022-3654. (American Chemical Society)
      In a no. of recent studies, information about the structure of large polyat. ions has been deduced from gas-phase ion mobility measurements by comparing mobilities measured in helium to those estd. for assumed geometries using a hard-sphere projection approxn. To examine the validity of this approach, we have compared mobilities calcd. using the hard-sphere projection approxn. for a range of fullerenes (C20-C240) to those detd. from trajectory calcns. with a more realistic He-fullerene potential. The He-fullerene potential we have employed, a sum of two-body 6-12 interactions plus a sum of ion-induced dipole interactions, was calibrated using the measured mobility of C60+ in helium over an 80-380 K temp. range. For the systems studied, the long-range interactions between the ion and buffer gas have a small, less than 10%, effect on the calcd. mobility at room temp. However, the effects are not insignificant, and in many cases it will be necessary to consider the long-range interactions if the correct structural assignments are to be made from measured ion mobilities.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XlsFCqu7w%253D&md5=d161824e4f762fb750c196d6270476d4
    11. 11
      Hinnenkamp, V.; Klein, J.; Meckelmann, S. W.; Balsaa, P.; Schmidt, T. C.; Schmitz, O. J. Anal. Chem. 2018, 90, 12042– 12050,  DOI: 10.1021/acs.analchem.8b02711
      11
      Comparison of CCS Values Determined by Traveling Wave Ion Mobility Mass Spectrometry and Drift Tube Ion Mobility Mass Spectrometry
      Hinnenkamp, Vanessa; Klein, Julia; Meckelmann, Sven W.; Balsaa, Peter; Schmidt, Torsten C.; Schmitz, Oliver J.
      Analytical Chemistry (Washington, DC, United States) (2018), 90 (20), 12042-12050CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Collision cross section (CCS, Ω) values detd. by ion mobility mass spectrometry (IM-MS) provide the study of ion shape in the gas phase and use of these as further identification criteria in anal. approaches. Databases of CCS values for a variety of mols. detd. by different instrument types are available. The comparability of CCS values detd. by a drift tube ion mobility mass spectrometer (DTIM-MS) and a traveling wave ion mobility mass spectrometer (TWIM-MS) was studied to test if a common database could be used across IM techniques. A total of 124 substances were measured with both systems and CCS values of [M + H]+ and [M + Na]+ adducts were compared. Deviations <1% were found for most substances, but some compds. show deviations up to 6.2%, which indicate that CCS databases cannot be used without care independently from the instrument type. Addnl., for several mols. [2M + Na]+ ions were formed during electrospray ionization, whereas a part of them disintegrates to [M + Na]+ ions after passing through the drift tube and before reaching the TOF region, resulting in 2 signals in their drift spectrum for the [M + Na]+ adduct. Finally, the impact of different LC-IM-MS settings (solvent compn., solvent flow rate, desolvation temp., and desolvation gas flow rate) were studied to test whether they have an influence on the CCS values or not. These conditions have no significant impact. Only for karbutilate changes in the drift spectrum could be obsd. with different solvent types and flow rates using the DTIM-MS system, which could be caused by the protonation at different sites in the mol.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhslaitbzN&md5=0f957ee087e93729d70c72d4df1f3347
    12. 12
      Stow, S. M.; Causon, T. J.; Zheng, X.; Kurulugama, R. T.; Mairinger, T.; May, J. C.; Rennie, E. E.; Baker, E. S.; Smith, R. D.; McLean, J. A.; Hann, S.; Fjeldsted, J. C. Anal. Chem. 2017, 89, 9048– 9055,  DOI: 10.1021/acs.analchem.7b01729
      12
      An Interlaboratory Evaluation of Drift Tube Ion Mobility-Mass Spectrometry Collision Cross Section Measurements
      Stow, Sarah M.; Causon, Tim J.; Zheng, Xueyun; Kurulugama, Ruwan T.; Mairinger, Teresa; May, Jody C.; Rennie, Emma E.; Baker, Erin S.; Smith, Richard D.; McLean, John A.; Hann, Stephan; Fjeldsted, John C.
      Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9048-9055CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Collision cross section (CCS) measurements resulting from ion mobility-mass spectrometry (IM-MS) expts. provide a promising orthogonal dimension of structural information in MS-based anal. sepns. As with any mol. identifier, interlab. standardization must precede broad range integration into anal. workflows. The authors present a ref. drift tube ion mobility mass spectrometer (DTIM-MS) where improvements on the measurement accuracy of exptl. parameters influencing IM sepns. provide standardized drift tube, nitrogen CCS values (DTCCSN2) for over 120 unique ion species with the lowest measurement uncertainty to date. The reproducibility of these DTCCSN2 values are evaluated across three addnl. labs. on a com. available DTIM-MS instrument. The traditional stepped field CCS method performs with a relative std. deviation of 0.29% for all ion species across the three addnl. labs. The calibrated single field CCS method, which is compatible with a wide range of chromatog. inlet systems, performs with an av., abs. bias of 0.54% to the standardized stepped field DTCCSN2 values on the ref. system. The low relative std. deviation and biases obsd. in this interlab. study illustrate the potential of DTIM-MS for providing a mol. identifier for a broad range of discovery based analyses.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhtbnL&md5=9081cdf22f75275dcb5f209741e49f9f
    13. 13
      Shvartsburg, A.; Jarrold, M. F. Chem. Phys. Lett. 1996, 261, 86– 91,  DOI: 10.1016/0009-2614(96)00941-4
      13
      An exact hard-spheres scattering model for the mobilities of polyatomic ions
      Shvartsburg, Alexandre A.; Jarrold, Martin F.
      Chemical Physics Letters (1996), 261 (1,2), 86-91CODEN: CHPLBC; ISSN:0009-2614. (Elsevier)
      We describe an exact hard-spheres scattering model for calcg. the gas-phase mobilities of polyat. ions. Ion mobility measurements have recently been used to deduce structural information for clusters and biomols. in the gas phase. In virtually all of the previous ion mobility studies, mobilities were evaluated for comparison with the exptl. data using a projection approxn. Comparison of the collision integrals calcd. using the exact hard-spheres scattering model with those estd. using the projection approxn. shows that large deviations, over 20%, occur for some geometries with grossly concave surfaces.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK28XlvFKnsbk%253D&md5=ee35ea2bf473e9e83188f21ada9d37ae
    14. 14
      Kim, H. I.; Kim, H.; Pang, E. S.; Ryu, E. K.; Beegle, L. W.; Loo, J. A.; Goddard, W. A.; Kanik, I. Anal. Chem. 2009, 81, 8289– 8297,  DOI: 10.1021/ac900672a
      14
      Structural Characterization of Unsaturated Phosphatidylcholines Using Traveling Wave Ion Mobility Spectrometry
      Kim, Hugh I.; Kim, Hyungjun; Pang, Eric S.; Ryu, Ernest K.; Beegle, Luther W.; Loo, Joseph A.; Goddard, William A.; Kanik, Isik
      Analytical Chemistry (Washington, DC, United States) (2009), 81 (20), 8289-8297CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      A no. of phosphatidylcholine (PC) cations spanning a mass range of 400-1000 Da are investigated using electrospray ionization mass spectrometry coupled with traveling wave ion mobility spectrometry (TWIMS). A high correlation between mass and mobility is demonstrated with satd. phosphatidylcholine cations in N2. A significant deviation from this mass-mobility correlation line is obsd. for the unsatd. PC cation. The authors found that the double bond in the acyl chain causes a 5% redn. in drift time. The drift time is reduced at a rate of ∼1% for each addnl. double bond. Theor. collision cross sections of PC cations exhibit good agreement with exptl. evaluated values. Collision cross sections are detd. using the recently derived relation between mobility and drift time in TWIMS stacked ring ion guide (SRIG) and compared to estd. collision cross sections using an empiric calibration method. Computational anal. was performed using the modified trajectory (TJ) method with nonspherical N2 mols. as the drift gas. The difference between estd. collision cross sections and theor. collision cross sections of PC cations is related to the sensitivity of the PC cation collision cross sections to the details of the ion-neutral interactions. The origin of the obsd. correlation and deviation between mass and mobility of PC cations is discussed in terms of the structural rigidity of these mols. using mol. dynamic simulations.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXhtFGlsLrL&md5=79ffad12c1e053d5450955fbccca9fdf
    15. 15
      Kim, H.; Kim, H. I.; Johnson, P. V.; Beegle, L. W.; Beauchamp, J. L.; Goddard, W. A.; Kanik, I. Anal. Chem. 2008, 80, 1928,  DOI: 10.1021/ac701888e
      15
      Experimental and Theoretical Investigation into the Correlation between Mass and Ion Mobility for Choline and Other Ammonium Cations in N2
      Kim, Hyungjun; Kim, Hugh I.; Johnson, Paul V.; Beegle, Luther W.; Beauchamp, J. L.; Goddard, William A.; Kanik, Isik
      Analytical Chemistry (Washington, DC, United States) (2008), 80 (6), 1928-1936CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      A no. of tertiary amine and quaternary ammonium cations spanning a mass range of 60-146 amu (trimethylamine, tetramethylammonium, trimethylethylammonium, N,N-dimethylaminoethanol, choline, N,N-dimethylglycine, betaine, acetylcholine, (3-carboxypropyl)trimethylammonium) were investigated using electrospray ionization ion mobility spectrometry. Measured ion mobilities demonstrate a high correlation between mass and mobility in N2. In addn., identical mobilities within exptl. uncertainties are obsd. for structurally dissimilar ions with similar ion masses. For example, dimethylethylammonium (88 amu) cations and protonated N,N-dimethylaminoethanol cations (90 amu) show identical mobilities (1.93 cm2 V-1 s-1) though N,N-dimethylaminoethanol contains a hydroxyl functional group while dimethylethylammonium only contains alkyl groups. Computational anal. was performed using the modified trajectory (TJ) method with nonspherical N2 mols. as the drift gas. The sensitivity of the ammonium cation collision cross sections to the details of the ion-neutral interactions was investigated and compared to other classes of org. mols. (carboxylic acids and abiotic amino acids). The specific charge distribution of the mol. ions in the investigated mass range has an insignificant affect on the collision cross section.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhvFChsbY%253D&md5=f0354c65a2e2274c042129ccdba8f7ca
    16. 16
      Campuzano, I.; Bush, M. F.; Robinson, C. V.; Beaumont, C.; Richardson, K.; Kim, H.; Kim, H. I. Anal. Chem. 2012, 84, 1026– 1033,  DOI: 10.1021/ac202625t
      16
      Structural Characterization of Drug-like Compounds by Ion Mobility Mass Spectrometry: Comparison of Theoretical and Experimentally Derived Nitrogen Collision Cross Sections
      Campuzano, Iain; Bush, Matthew F.; Robinson, Carol V.; Beaumont, Claire; Richardson, Keith; Kim, Hyungjun; Kim, Hugh I.
      Analytical Chemistry (Washington, DC, United States) (2012), 84 (2), 1026-1033CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      We present the use of drug-like mols. as a traveling wave (T-wave) ion mobility (IM) calibration sample set, covering the m/z range of 122.1-609.3, the nitrogen collision cross-section (ΩN2) range of 124.5-254.3 Å2 and the helium collision cross-section (ΩHe) range of 63.0-178.8 Å2. Abs. ΩN2 and ΩHe values for the drug-like calibrants and two diastereomers were measured using a drift-tube instrument with radio frequency (RF) ion confinement. T-wave drift-times for the protonated diastereomers betamethasone and dexamethasone are reproducibly different. Calibration of these drift-times yields T-wave ΩN2 values of 189.4 and 190.4 Å2, resp. These results demonstrate the ability of T-wave IM spectrometry to differentiate diastereomers differing in ΩN2 value by only 1 Å2, even though the resoln. of these IM expts. were ∼40 (Ω/ΔΩ). Demonstrated through d. functional theory optimized geometries and ionic electrostatic surface potential anal., the small but measurable mobility difference between the two diastereomers is mainly due to short-range van der Waals interactions with the neutral buffer gas and not long-range charge-induced dipole interactions. The exptl. RF-confining drift-tube and T-wave ΩN2 values were also evaluated using a nitrogen based trajectory method, optimized for T-wave operating temp. and pressures, incorporating addnl. scaling factors to the Lennard-Jones potentials. Exptl. ΩHe values were also compared to the original and optimized helium based trajectory methods.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3MXhsFOhtrvJ&md5=f5d0d911bc241a8c07aafbbe80139173
    17. 17
      Larriba, C.; Hogan, C. J., Jr. J. Phys. Chem. A 2013, 117, 3887– 3901,  DOI: 10.1021/jp312432z
      17
      Ion Mobilities in Diatomic Gases: Measurement versus Prediction with Non-Specular Scattering Models
      Larriba, Carlos; Hogan, Christopher J.
      Journal of Physical Chemistry A (2013), 117 (19), 3887-3901CODEN: JPCAFH; ISSN:1089-5639. (American Chemical Society)
      Ion/elec. mobility measurements of nanoparticles and polyat. ions are typically linked to particle/ion phys. properties through either application of the Stokes-Millikan relationship or comparison to mobilities predicted from polyat. models, which assume that gas mols. scatter specularly and elastically from rigid structural models. However, there is a discrepancy between these approaches; when specular, elastic scattering models (i.e., elastic-hard-sphere scattering, EHSS) are applied to polyat. models of nanometer-scale ions with finite-sized impinging gas mols., predictions are in substantial disagreement with the Stokes-Millikan equation. To rectify this discrepancy, a new approach is developed and tested for mobility calcns. using polyat. models in which non-specular (diffuse) and inelastic gas-mol. scattering is considered. Two distinct semiempirical models of gas-mol. scattering from particle surfaces were considered. In the first, which has been traditionally invoked in the study of aerosol nanoparticles, 91% of collisions are diffuse and thermally accommodating, and 9% are specular and elastic. In the second, all collisions are considered to be diffuse and accommodating, but the av. speed of the gas mols. reemitted from a particle surface is 8% lower than the mean thermal speed at the particle temp. Both scattering models attempt to mimic exchange between translational, vibrational, and rotational modes of energy during collision, as would be expected during collision between a nonmonoat. gas mol. and a nonfrozen particle surface. The mobility calcn. procedure was applied considering both hard-sphere potentials between gas mols. and the atoms within a particle and the long-range ion-induced dipole (polarization) potential. Predictions were compared to previous measurements in air near room temp. of multiply charged poly(ethylene glycol) (PEG) ions, which range in morphol. from compact to highly linear, and singly charged tetraalkylammonium cations. It was found that both non-specular, inelastic scattering rules lead to excellent agreement between predictions and exptl. mobility measurements (within 5% of each other) and that polarization potentials must be considered to make correct predictions for high-mobility particles/ions. Conversely, traditional specular, elastic scattering models were found to substantially overestimate the mobilities of both types of ions.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3sXjvFWhsb0%253D&md5=a40f1a9527fc4108f161648277010c9d
    18. 18
      Ewing, S. A.; Donor, M. T.; Wilson, J. W.; Prell, J. S. J. Am. Soc. Mass Spectrom. 2017, 28, 587– 596,  DOI: 10.1007/s13361-017-1594-2
      18
      Collidoscope: An Improved Tool for Computing Collisional Cross-Sections with the Trajectory Method
      Ewing, Simon A.; Donor, Micah T.; Wilson, Jesse W.; Prell, James S.
      Journal of the American Society for Mass Spectrometry (2017), 28 (4), 587-596CODEN: JAMSEF; ISSN:1044-0305. (Springer)
      Ion mobility-mass spectrometry (IM-MS) can be a powerful tool for detg. structural information about ions in the gas phase, from small covalent analytes to large, native-like or denatured proteins and complexes. For large biomol. ions, which may have a wide variety of possible gas-phase conformations and multiple charge sites, quant., phys. explicit modeling of collisional cross sections (CCSs) for comparison to IMS data can be challenging and time-consuming. We present a "trajectory method" (TM) based CCS calculator, named "Collidoscope," which utilizes parallel processing and optimized trajectory sampling, and implements both He and N2 as collision gas options. Also included is a charge-placement algorithm for detg. probable charge site configurations for protonated protein ions given an input geometry in PDB file format. Results from Collidoscope are compared with those from the current state-of-the-art CCS simulation suite, IMoS. Collidoscope CCSs are within 4% of IMoS values for ions with masses from ∼18 Da to ∼800 kDa. Collidoscope CCSs using X-ray crystal geometries are typically within a few percent of IM-MS exptl. values for ions with mass up to ∼3.5 kDa (melittin), and discrepancies for larger ions up to ∼800 kDa (GroEL) are attributed in large part to changes in ion structure during and after the electrospray process. Due to its phys. explicit modeling of scattering, computational efficiency, and accuracy, Collidoscope can be a valuable tool for IM-MS research, esp. for large biomol. ions.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXisF2mt7k%253D&md5=1f840c45579e5371f1601aadd1dc0e60
    19. 19
      Lee, J. W.; Lee, H. H. L.; Davidson, K. L.; Bush, M. F.; Kim, H. I. Analyst 2018, 143, 1786– 1796,  DOI: 10.1039/C8AN00270C
      19
      Structural characterization of small molecular ions by ion mobility mass spectrometry in nitrogen drift gas: improving the accuracy of trajectory method calculations
      Lee, Jong Wha; Lee, Hyun Hee L.; Davidson, Kimberly L.; Bush, Matthew F.; Kim, Hugh I.
      Analyst (Cambridge, United Kingdom) (2018), 143 (8), 1786-1796CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
      The investigation of ion structures based on a combination of ion mobility mass spectrometry (IM-MS) expts. and theor. collision cross section (CCS) calcns. has become important to many fields of research. However, the accuracy of current CCS calcns. for ions in nitrogen drift gas limits the information content of many expts. In particular, few studies have evaluated and attempted to improve the theor. tools for CCS calcn. in nitrogen drift gas. In this study, based on high-quality exptl. measurements and theor. modeling, a comprehensive evaluation of various aspects of CCS calcns. in nitrogen drift gas is performed. It is shown that the modification of the ion-nitrogen van der Waals (vdW) interaction potential enables accurate CCS predictions of 29 small ions with ca. 3% max. relative error. The present method exhibits no apparent systematic bias with respect to ion CCS (size) and dipole moment, suggesting that the method adequately describes the long-range interactions between the ions and the buffer gas. However, the method shows limitations in reproducing exptl. CCS at low temps. ( < 150 K) and for macromol. ions, and calcns. for these cases should be complemented by CCS calcn. methods in helium drift gas. This study presents an accurate and well-characterized CCS calcn. method for ions in nitrogen drift gas that is expected to become an important tool for ion structural characterization and mol. identification. The exptl. values reported here also provide a foundation for future studies aiming at developing more efficient computational tools.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXkslamtbY%253D&md5=906f8f6bc85485ff55f1a0bd26257c0f
    20. 20
      Zanotto, L.; Heerdt, G.; Souza, P. C. T.; Araujo, G.; Skaf, M. S. J. Comput. Chem. 2018, 39, 1675– 1681,  DOI: 10.1002/jcc.25199
      20
      High performance collision cross section calculation-HPCCS
      Zanotto, Leandro; Heerdt, Gabriel; Souza, Paulo C. T.; Araujo, Guido; Skaf, Munir S.
      Journal of Computational Chemistry (2018), 39 (21), 1675-1681CODEN: JCCHDD; ISSN:0192-8651. (John Wiley & Sons, Inc.)
      Since the com. introduction of Ion Mobility coupled with Mass Spectrometry (IM-MS) devices in 2003, a large no. of research labs. have embraced the technique. IM-MS is a fairly rapid expt. used as a mol. sepn. tool and to obtain structural information. The interpretation of IM-MS data is still challenging and relies heavily on theor. calcns. of the mol.'s collision cross section (CCS) against a buffer gas. Here, a new software (HPCCS) is presented, which performs CCS calcns. using high performance computing techniques. Based on the trajectory method, HPCCS can accurately calc. CCS for a great variety of mols., ranging from small org. mols. to large protein complexes, using helium or nitrogen as buffer gas with considerable gains in computer time compared to publicly available codes under the same level of theory. HPCCS is available as free software under the Academic Use License at . © 2018 Wiley Periodicals, Inc.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjs1Cltb8%253D&md5=8e74ea412b89c3f305f223fac23acf8c
    21. 21
      Colby, S. M.; Thomas, D. G.; Nunez, J. R.; Baxter, D. J.; Glaesemann, K. R.; Brown, J. M.; Pirrung, M. A.; Govind, N.; Teeguarden, J. G.; Metz, T. O.; Renslow, R. S. Anal. Chem. 2019, 91, 4346– 4356,  DOI: 10.1021/acs.analchem.8b04567
      21
      ISiCLE: A Quantum Chemistry Pipeline for Establishing in Silico Collision Cross Section Libraries
      Colby, Sean M.; Thomas, Dennis G.; Nunez, Jamie R.; Baxter, Douglas J.; Glaesemann, Kurt R.; Brown, Joseph M.; Pirrung, Meg A.; Govind, Niranjan; Teeguarden, Justin G.; Metz, Thomas O.; Renslow, Ryan S.
      Analytical Chemistry (Washington, DC, United States) (2019), 91 (7), 4346-4356CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      High-throughput, comprehensive, and confident identifications of metabolites and other chems. in biol. and environmental samples will revolutionize the understanding of the role these chem. diverse mols. play in biol. systems. Despite recent technol. advances, metabolomics studies still result in the detection of a disproportionate no. of features that cannot be confidently assigned to a chem. structure. This inadequacy is driven by the single most significant limitation in metabolomics, the reliance on ref. libraries constructed by anal. of authentic ref. materials with limited com. availability. To this end, the authors have developed the in silico chem. library engine (ISiCLE), a high-performance computing-friendly cheminformatics workflow for generating libraries of chem. properties. In the instantiation described here, the authors predict probable three-dimensional mol. conformers (i.e., conformational isomers) using chem. identifiers as input, from which collision cross sections (CCS) are derived. The approach employs first-principles simulation, distinguished by the use of mol. dynamics, quantum chem., and ion mobility calcns., to generate structures and chem. property libraries, all without training data. Importantly, optimization of ISiCLE included a refactoring of the popular MOBCAL code for trajectory-based mobility calcns., improving its computational efficiency by over 2 orders of magnitude. Calcd. CCS values were validated against 1983 exptl. measured CCS values and compared to previously reported CCS calcn. approaches. Av. calcd. CCS error for the validation set is 3.2% using std. parameters, outperforming other d. functional theory (DFT)-based methods and machine learning methods (e.g., MetCCS). An online database is introduced for sharing both calcd. and exptl. CCS values (metabolomics.pnnl.gov), initially including a CCS library with over 1 million entries. Finally, three successful applications of mol. characterization using calcd. CCS are described, including providing evidence for the presence of an environmental degrdn. product, the sepn. of mol. isomers, and an initial characterization of complex blinded mixts. of exposure chems. This work represents a method to address the limitations of small mol. identification and offers an alternative to generating chem. identification libraries exptl. by analyzing authentic ref. materials. All code is available at github.com/pnnl.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXivVOgsb0%253D&md5=786b8a6e8d764d48a9a4debfc31e4d59
    22. 22
      Zhou, Z.; Shen, X.; Tu, J.; Zhu, Z. J. Anal. Chem. 2016, 88, 11084– 11091,  DOI: 10.1021/acs.analchem.6b03091
      22
      Large-Scale Prediction of Collision Cross-Section Values for Metabolites in Ion Mobility-Mass Spectrometry
      Zhou, Zhiwei; Shen, Xiaotao; Tu, Jia; Zhu, Zheng-Jiang
      Analytical Chemistry (Washington, DC, United States) (2016), 88 (22), 11084-11091CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      The rapid development of metabolomics has significantly advanced health and disease related research. However, metabolite identification remains a major anal. challenge for untargeted metabolomics. While the use of collision cross-section (CCS) values obtained in ion mobility - mass spectrometry (IM-MS) effectively increases identification confidence of metabolites, it is restricted by the limited no. of available CCS values for metabolites. Here, the authors demonstrated the use of a machine-learning algorithm called support vector regression (SVR) to develop a prediction method that utilized 14 common mol. descriptors to predict CCS values for metabolites. In this work, the authors first exptl. measured CCS values (ΩN2) of ∼400 metabolites in nitrogen buffer gas, and used these values as training data to optimize the prediction method. The high prediction precision of this method was externally validated using an independent set of metabolites with a median relative error (MRE) of ∼3%, better than conventional theor. calcn. Using the SVR based prediction method, a large-scale predicted CCS database was generated for 35,203 metabolites in the Human Metabolome Database (HMDB). For each metabolite, five different ion adducts in pos. and neg. modes were predicted, accounting for 176,015 CCS values in total. Finally, improved metabolite identification accuracy was demonstrated using real biol. samples. Conclusively, the results proved that the SVR based prediction method can accurately predict nitrogen CCS values (ΩN2) of metabolites from mol. descriptors, and effectively improve identification accuracy and efficiency in untargeted metabolomics. The predicted CCS database, namely, MetCCS, is freely available on the Internet.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhslWjsbbM&md5=2e93305e7cce5f48dae1ed5fe2e861a9
    23. 23
      Zhou, Z.; Tu, J.; Xiong, X.; Shen, X.; Zhu, Z. J. Anal. Chem. 2017, 89, 9559– 9566,  DOI: 10.1021/acs.analchem.7b02625
      23
      LipidCCS: Prediction of Collision Cross-Section Values for Lipids with High Precision To Support Ion Mobility-Mass Spectrometry-Based Lipidomics
      Zhou, Zhiwei; Tu, Jia; Xiong, Xin; Shen, Xiaotao; Zhu, Zheng-Jiang
      Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9559-9566CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      The use of collision cross-section (CCS) values derived from ion mobility-mass spectrometry (IM-MS) has been proven to facilitate lipid identifications. Its utility is restricted by the limited availability of CCS values. Recently, the machine-learning algorithm-based prediction (e.g., MetCCS) is reported to generate CCS values in a large-scale. However, the prediction precision is not sufficient to differentiate lipids due to their high structural similarities and subtle differences on CCS values. To address this challenge, we developed a new approach, namely, LipidCCS, to precisely predict lipid CCS values. In LipidCCS, a set of mol. descriptors were optimized using bioinformatic approaches to comprehensively describe the subtle structure differences for lipids. The use of optimized mol. descriptors together with a large set of std. CCS values for lipids (458 in total) to build the prediction model significantly improved the precision. The prediction precision of LipidCCS was externally validated with median relative errors (MRE) of ∼1% using independent data sets across different instruments (Agilent DTIM-MS and Waters TWIM-MS) and labs. We also demonstrated that the improved precision in the predicted LipidCCS database (15,646 lipids and 63,434 CCS values in total) could effectively reduce false-pos. identifications of lipids. Common users can freely access our LipidCCS web server for the following: (1) the prediction of lipid CCS values directly from SMILES structure; (2) database search; and (3) lipid match and identification. We believe LipidCCS will be a valuable tool to support IM-MS-based lipidomics.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhurbN&md5=147694b1cc6fdc83de29883511f8185c
    24. 24
      Bijlsma, L.; Bade, R.; Celma, A.; Mullin, L.; Cleland, G.; Stead, S.; Hernandez, F.; Sancho, J. V. Anal. Chem. 2017, 89, 6583– 6589,  DOI: 10.1021/acs.analchem.7b00741
      24
      Prediction of Collision Cross-Section Values for Small Molecules: Application to Pesticide Residue Analysis
      Bijlsma, Lubertus; Bade, Richard; Celma, Alberto; Mullin, Lauren; Cleland, Gareth; Stead, Sara; Hernandez, Felix; Sancho, Juan V.
      Analytical Chemistry (Washington, DC, United States) (2017), 89 (12), 6583-6589CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      The use of collision cross-section (CCS) values obtained by ion mobility high-resoln. mass spectrometry has added a third dimension (alongside retention time and exact mass) to aid in the identification of compds. However, its utility is limited by the no. of exptl. CCS values currently available. This work demonstrates the potential of artificial neural networks (ANNs) for the prediction of CCS values of pesticides. The predictor, based on eight software-chosen mol. descriptors, was optimized using CCS values of 205 small mols. and validated using a set of 131 pesticides. The relative error was within 6% for 95% of all CCS values for protonated mols., resulting in a median relative error less than 2%. In order to demonstrate the potential of CCS prediction, the strategy was applied to spinach samples. It notably improved the confidence in the tentative identification of suspect and nontarget pesticides.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXosVOqsr0%253D&md5=a3328a65ce4adc8c838b7aee5aa6c2d4
    25. 25
      Soper-Hopper, M. T.; Petrov, A. S.; Howard, J. N.; Yu, S. S.; Forsythe, J. G.; Grover, M. A.; Fernandez, F. M. Chem. Commun. (Cambridge, U. K.) 2017, 53, 7624– 7627,  DOI: 10.1039/C7CC04257D
      25
      Collision cross section predictions using 2-dimensional molecular descriptors
      Soper-Hopper, M. T.; Petrov, A. S.; Howard, J. N.; Yu, S.-S.; Forsythe, J. G.; Grover, M. A.; Fernandez, F. M.
      Chemical Communications (Cambridge, United Kingdom) (2017), 53 (54), 7624-7627CODEN: CHCOFS; ISSN:1359-7345. (Royal Society of Chemistry)
      Traditional methods for deriving computationally-generated collision cross sections for comparisons with ion mobility-mass spectrometry data require 3-dimensional energy-minimized structures and are often time consuming, preventing high throughput implementation. Here, we introduce a method to predict ion mobility collision cross sections of lipids and peptide analogs important in prebiotic chem. and other fields. Using less than 100 2-D mol. descriptors this approach resulted in prediction errors of less than 2%.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhtVCltbjP&md5=d5d2efa7af0ed7e4c83680770fb8ba1b
    26. 26
      Mollerup, C. B.; Mardal, M.; Dalsgaard, P. W.; Linnet, K.; Barron, L. P. J. Chromatogr., A 2018, 1542, 82– 88,  DOI: 10.1016/j.chroma.2018.02.025
      26
      Prediction of collision cross section and retention time for broad scope screening in gradient reversed-phase liquid chromatography-ion mobility-high resolution accurate mass spectrometry
      Mollerup, Christian Brinch; Mardal, Marie; Dalsgaard, Petur Weihe; Linnet, Kristian; Barron, Leon Patrick
      Journal of Chromatography A (2018), 1542 (), 82-88CODEN: JCRAEY; ISSN:0021-9673. (Elsevier B.V.)
      Exact mass, retention time (RT), and collision cross section (CCS) are used as identification parameters in liq. chromatog. coupled to ion mobility high resoln. accurate mass spectrometry (LC-IM-HRMS). Targeted screening analyses are now more flexible and can be expanded for suspect and nontargeted screening. These allow for tentative identification of new compds., and in-silico predicted ref. values are used for improving confidence and filtering false-pos. identifications. In this work, predictions of both RT and CCS values are performed with machine learning using artificial neural networks (ANNs). Prediction was based on mol. descriptors, 827 RTs, and 357 CCS values from pharmaceuticals, drugs of abuse, and their metabolites. ANN models for the prediction of RT or CCS sep. were examd., and the potential to predict both from a single model was investigated for the first time. The optimized combined RT-CCS model was a four-layered multilayer perceptron ANN, and the 95th prediction error percentiles were within 2 min RT error and 5% relative CCS error for the external validation set (n = 36) and the full RT-CCS dataset (n = 357). 88.6% (n = 733) of predicted RTs were within 2 min error for the full dataset. Overall, when using 2 min RT error and 5% relative CCS error, 91.9% (n = 328) of compds. were retained, while 99.4% (n = 355) were retained when using at least one of these thresholds. This combined prediction approach can therefore be useful for rapid suspect/nontargeted screening involving HRMS, and will support current workflows.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjtFamsbw%253D&md5=4d52e6dad7704a31434f4ff47e50923a
    27. 27
      Plante, P. L.; Francovic-Fontaine, E.; May, J. C.; McLean, J. A.; Baker, E. S.; Laviolette, F.; Marchand, M.; Corbeil, J. Anal. Chem. 2019, 91, 5191– 5199,  DOI: 10.1021/acs.analchem.8b05821
      27
      Predicting Ion Mobility Collision Cross-Sections Using a Deep Neural Network: DeepCCS
      Plante, Pier-Luc; Francovic-Fontaine, Elina; May, Jody C.; McLean, John A.; Baker, Erin S.; Laviolette, Francois; Marchand, Mario; Corbeil, Jacques
      Analytical Chemistry (Washington, DC, United States) (2019), 91 (8), 5191-5199CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Untargeted metabolomic measurements using mass spectrometry are a powerful tool for uncovering new small mols. with environmental and biol. importance. The small mol. identification step, however, still remains an enormous challenge due to fragmentation difficulties or unspecific fragment ion information. Current methods to address this challenge are often dependent on databases or require the use of NMR, which have their own difficulties. The use of the gas-phase collision cross section (CCS) values obtained from ion mobility spectrometry (IMS) measurements were recently demonstrated to reduce the no. of false pos. metabolite identifications. While promising, the amt. of empirical CCS information currently available is limited, thus predictive CCS methods need to be developed. In this article, the authors expand upon current exptl. IMS capabilities by predicting the CCS values using a deep learning algorithm. The authors successfully developed and trained a prediction model for CCS values requiring only information about a compd.'s SMILES notation and ion type. The use of data from five different labs. using different instruments allowed the algorithm to be trained and tested on more than 2400 mols. The resulting CCS predictions were found to achieve a coeff. of detn. of 0.97 and median relative error of 2.7% for a wide range of mols. Furthermore, the method requires only a small amt. of processing power to predict CCS values. Considering the performance, time, and resources necessary, as well as its applicability to a variety of mols., this model was able to outperform all currently available CCS prediction algorithms.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmtFOisLY%253D&md5=d735e5f80e99301113280f7c878314a5
    28. 28
      Hines, K.; Herron, J.; Xu, L. J. Lipid Res. 2017, 58, 809– 819,  DOI: 10.1194/jlr.D074724
      28
      Assessment of altered lipid homeostasis by HILIC-ion mobility-mass spectrometry-based lipidomics
      Hines, Kelly M.; Herron, Josi; Xu, Libin
      Journal of Lipid Research (2017), 58 (4), 809-819CODEN: JLPRAW; ISSN:0022-2275. (American Society for Biochemistry and Molecular Biology)
      Ion mobility-mass spectrometry (IM-MS) has proven to be a highly informative technique for the characterization of lipids from cells and tissues. We report the combination of hydrophilic-interaction liq. chromatog. (HILIC) with traveling-wave IM-MS (TWIM-MS) for comprehensive lipidomics anal. Main lipid categories such as glycerolipids, sphingolipids, and glycerophospholipids are sepd. on the basis of their lipid backbones in the IM dimension, whereas subclasses of each category are mostly sepd. on the basis of their headgroups in the HILIC dimension, demonstrating the orthogonality of HILIC and IM sepns. Using our previously established lipid calibrants for collision cross-section (CCS) measurements in TWIM, we measured over 250 CCS values covering 12 lipid classes in pos. and neg. modes. The coverage of the HILIC-IM-MS method is demonstrated in the anal. of Neuro2a neuroblastoma cells exposed to benzalkonium chlorides (BACs) with C10 or C16 alkyl chains, which we have previously shown to affect gene expression related to cholesterol and lipid homeostasis. We found that BAC exposure resulted in significant changes to several lipid classes, including glycerides, sphingomyelins, phosphatidylcholines, and phosphatidylethanolamines. Our results indicate that BAC exposure modifies lipid homeostasis in a manner that is dependent upon the length of the BAC alkyl chain.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXlsFSjtLg%253D&md5=b204c490ee0b8fe634cb08a50accae69
    29. 29
      Hines, K. M.; Waalkes, A.; Penewit, K.; Holmes, E. A.; Salipante, S. J.; Werth, B. J.; Xu, L. mSphere 2017, 2, e00492-17  DOI: 10.1128/mSphere.00492-17
      There is no corresponding record for this reference.
    30. 30
      Hines, K. M.; Xu, L. Chem. Phys. Lipids 2019, 219, 15– 22,  DOI: 10.1016/j.chemphyslip.2019.01.007
      30
      Lipidomic consequences of phospholipid synthesis defects in Escherichia coli revealed by HILIC-ion mobility-mass spectrometry
      Hines, Kelly M.; Xu, Libin
      Chemistry and Physics of Lipids (2019), 219 (), 15-22CODEN: CPLIA4; ISSN:0009-3084. (Elsevier Ireland Ltd.)
      Our understanding of phospholipid biosynthesis in Gram-pos. and Gram-neg. bacteria is derived from the prototypical Gram-neg. organism Escherichia coli. The inner and outer membranes of E. coli are largely composed of phosphatidylethanolamine (PE), minor amts. of phosphatidylglycerol (PG) and cardiolipin (CL). We report here the utility of hydrophilic interaction liq. chromatog. (HILIC) paired with ion mobility-mass spectrometry (IM-MS) for the comprehensive anal. of the E. coli lipidome. Using strains with chromosomal deletions in the PG and CL synthesis genes pgsA and clsABC, resp., we show that defective phospholipid biosynthesis in E. coli results in fatty-acid specific changes in select lipid classes and the presence of the minor triacylated phospholipids, acylphosphatidyl glycerol (acylPG) and N-acylphosphatidylethanolamine (N-acylPE). Notably, acylPGs were accumulated in the clsABC-KO strain, but were absent in other mutant strains. The sepn. of 1-lyso and 2-lyso-phosphatidylethanolamines (lysoPEs) is demonstrated in both the HILIC and IM dimensions. Using our previously validated calibration method, collision cross section values of nearly 200 phospholipids found in E. coli were detd. on a traveling wave IM-MS platform, including newly reported values for cardiolipins, positional isomers of lysoPEs, acylPGs and N-acylPEs.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXhvFOhurY%253D&md5=d2b158f0b956c6c56a794f80fe5d7133
    31. 31
      Groessl, M.; Graf, S.; Knochenmuss, R. Analyst 2015, 140, 6904– 6911,  DOI: 10.1039/C5AN00838G
      31
      High resolution ion mobility-mass spectrometry for separation and identification of isomeric lipids
      Groessl, M.; Graf, S.; Knochenmuss, R.
      Analyst (Cambridge, United Kingdom) (2015), 140 (20), 6904-6911CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
      Lipidomics is a particularly difficult anal. challenge due to the no. and importance of isomeric species that are known or postulated in biol. samples. Current sepn. and identification techniques are too often insufficiently powerful, slow or ambiguous. High resoln., low field ion mobility coupled to mass spectrometry is shown here to have sufficient performance to represent a new alternative for lipidomics. For the first time, drift-tube ion mobility sepn. of lipid isomers that differ only in position of the acyl chain, position of the double bond or double bond geometry is demonstrated. Differences in collision cross sections of <1% are sufficient for baseline sepn. The same level of performance is maintained in complex biol. mixts. More than 130 high-precision reduced mobility and collision cross section values were also detd. for a range of lipids. Such data can be the basis of a new lipidomics workflow, as the appropriate libraries are developed.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhtlKntbvP&md5=dd7628f57b198a94f5520191a96f9435
    32. 32
      Hines, K. M.; Ross, D. H.; Davidson, K. L.; Bush, M. F.; Xu, L. Anal. Chem. 2017, 89, 9023– 9030,  DOI: 10.1021/acs.analchem.7b01709
      32
      Large-Scale Structural Characterization of Drug and Drug-Like Compounds by High-Throughput Ion Mobility-Mass Spectrometry
      Hines, Kelly M.; Ross, Dylan H.; Davidson, Kimberly L.; Bush, Matthew F.; Xu, Libin
      Analytical Chemistry (Washington, DC, United States) (2017), 89 (17), 9023-9030CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Ion mobility-mass spectrometry (IM-MS) can provide orthogonal information, i.e. m/z and collision cross section (CCS), for the identification of drugs and drug metabolites. However, only a small no. of CCS values are available for drugs, which limits the use of CCS as an identification parameter and the assessment of structure-function relationships of drugs using IM-MS. Here the authors report the development of a rapid workflow for the measurement of CCS values of a large no. of drug or drug-like mols. in nitrogen on the widely available traveling wave IM-MS (TWIM-MS) platform. Using a combination of small mol. and polypeptide CCS calibrants, the authors successfully detd. the nitrogen CCS values of 1425 drug or drug-like mols. in the MicroSource Discovery Systems' Spectrum Collection using flow injection anal. of 384-well plates. Software was developed to streamline data extn., processing, and calibration. The authors found that the overall drug collection covers a wide CCS range for the same masses, suggesting a large structural diversity of these drugs. However, individual drug classes appear to occupy a narrow and unique space in the CCS-mass 2D spectrum, suggesting a tight structure-function relationship for each class of drugs with a specific target. The authors obsd. bimodal distributions for several antibiotic species due to multiple protomers, including the known fluoroquinolone protomers and the newer finding of cephalosporin protomers. Lastly, the authors demonstrated the utility of the high-throughput method and drug CCS database by quickly and confidently confirming the active component in a pharmaceutical product.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXht1GhtbfE&md5=6dac3abb0c1050b8370a8e79066a7d34
    33. 33
      May, J. C.; Goodwin, C. R.; Lareau, N. M.; Leaptrot, K. L.; Morris, C. B.; Kurulugama, R. T.; Mordehai, A.; Klein, C.; Barry, W.; Darland, E.; Overney, G.; Imatani, K.; Stafford, G. C.; Fjeldsted, J. C.; McLean, J. A. Anal. Chem. 2014, 86, 2107– 2116,  DOI: 10.1021/ac4038448
      33
      Conformational Ordering of Biomolecules in the Gas Phase: Nitrogen Collision Cross Sections Measured on a Prototype High Resolution Drift Tube Ion Mobility-Mass Spectrometer
      May, Jody C.; Goodwin, Cody R.; Lareau, Nichole M.; Leaptrot, Katrina L.; Morris, Caleb B.; Kurulugama, Ruwan T.; Mordehai, Alex; Klein, Christian; Barry, William; Darland, Ed; Overney, Gregor; Imatani, Kenneth; Stafford, George C.; Fjeldsted, John C.; McLean, John A.
      Analytical Chemistry (Washington, DC, United States) (2014), 86 (4), 2107-2116CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Ion mobility-mass spectrometry measurements which describe the gas-phase scaling of mol. size and mass are of both fundamental and pragmatic utility. Fundamentally, such measurements expand the authors' understanding of intrinsic intramol. folding forces in the absence of solvent. Practically, reproducible transport properties, such as gas-phase collision cross-section (CCS), are anal. useful metrics for identification and characterization purposes. Here, the authors report 594 CCS values obtained in nitrogen drift gas on an electrostatic drift tube ion mobility-mass spectrometry (IM-MS) instrument. The instrument platform is a newly developed prototype incorporating a uniform-field drift tube bracketed by electrodynamic ion funnels and coupled to a high resoln. quadrupole time-of-flight mass spectrometer. The CCS values reported here are of high exptl. precision (±0.5% or better) and represent four chem. distinct classes of mols. (quaternary ammonium salts, lipids, peptides, and carbohydrates), which enables structural comparisons to be made between mols. of different chem. compns. for the rapid "omni-omic" characterization of complex biol. samples. Comparisons made between helium and nitrogen-derived CCS measurements demonstrate that nitrogen CCS values are systematically larger than helium values; however, general sepn. trends between chem. classes are retained regardless of the drift gas. These results underscore that, for the highest CCS accuracy, care must be exercised when using helium-derived CCS values to calibrate measurements obtained in nitrogen, as is the common practice in the field.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXpslehtQ%253D%253D&md5=9e420afba857551e3d389ad7e038187a
    34. 34
      Paglia, G.; Williams, J. P.; Menikarachchi, L.; Thompson, J. W.; Tyldesley-Worster, R.; Halldorsson, S.; Rolfsson, O.; Moseley, A.; Grant, D.; Langridge, J.; Palsson, B. O.; Astarita, G. Anal. Chem. 2014, 86, 3985– 3993,  DOI: 10.1021/ac500405x
      34
      Ion Mobility Derived Collision Cross Sections to Support Metabolomics Applications
      Paglia, Giuseppe; Williams, Jonathan P.; Menikarachchi, Lochana; Thompson, J. Will; Tyldesley-Worster, Richard; Halldorsson, Skarphedinn; Rolfsson, Ottar; Moseley, Arthur; Grant, David; Langridge, James; Palsson, Bernhard O.; Astarita, Giuseppe
      Analytical Chemistry (Washington, DC, United States) (2014), 86 (8), 3985-3993CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Metabolomics is a rapidly evolving anal. approach in life and health sciences. The structural elucidation of the metabolites of interest remains a major anal. challenge in the metabolomics workflow. Here, we investigate the use of ion mobility as a tool to aid metabolite identification. Ion mobility allows for the measurement of the rotationally averaged collision cross-section (CCS), which gives information about the ionic shape of a mol. in the gas phase. We measured the CCSs of 125 common metabolites using traveling-wave ion mobility-mass spectrometry (TW-IM-MS). CCS measurements were highly reproducible on instruments located in three independent labs. (RSD < 5% for 99%). We also detd. the reproducibility of CCS measurements in various biol. matrixes including urine, plasma, platelets, and red blood cells using ultra performance liq. chromatog. (UPLC) coupled with TW-IM-MS. The mean RSD was < 2% for 97% of the CCS values, compared to 80% of retention times. Finally, as proof of concept, we used UPLC-TW-IM-MS to compare the cellular metabolome of epithelial and mesenchymal cells, an in vitro model used to study cancer development. Exptl. detd. and computationally derived CCS values were used as orthogonal anal. parameters in combination with retention time and accurate mass information to confirm the identity of key metabolites potentially involved in cancer. Thus, our results indicate that adding CCS data to searchable databases and to routine metabolomics workflows will increase the identification confidence compared to traditional anal. approaches.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXksVWisbs%253D&md5=361ab03df449c68121984ec0f1cb4241
    35. 35
      Zheng, X.; Aly, N. A.; Zhou, Y.; Dupuis, K. T.; Bilbao, A.; Paurus, V. L.; Orton, D. J.; Wilson, R.; Payne, S. H.; Smith, R. D.; Baker, E. S. Chem. Sci. 2017, 8, 7724– 7736,  DOI: 10.1039/C7SC03464D
      35
      A structural examination and collision cross section database for over 500 metabolites and xenobiotics using drift tube ion mobility spectrometry
      Zheng, Xueyun; Aly, Noor A.; Zhou, Yuxuan; Dupuis, Kevin T.; Bilbao, Aivett; Paurus, Vanessa L.; Orton, Daniel J.; Wilson, Ryan; Payne, Samuel H.; Smith, Richard D.; Baker, Erin S.
      Chemical Science (2017), 8 (11), 7724-7736CODEN: CSHCCN; ISSN:2041-6520. (Royal Society of Chemistry)
      The confident identification of metabolites and xenobiotics in biol. and environmental studies is an anal. challenge due to their immense dynamic range, vast chem. space and structural diversity. Ion mobility spectrometry (IMS) is widely used for small mol. analyses since it can sep. isomeric species and be easily coupled with front end sepns. and mass spectrometry for multidimensional characterizations. However, to date IMS metabolomic and exposomic studies have been limited by an inadequate no. of accurate collision cross section (CCS) values for small mols., causing features to be detected but not confidently identified. In this work, we utilized drift tube IMS (DTIMS) to directly measure CCS values for over 500 small mols. including primary metabolites, secondary metabolites and xenobiotics. Since DTIMS measurements do not need calibrant ions or calibration like some other IMS techniques, they avoid calibration errors which can cause problems in distinguishing structurally similar mols. All measurements were performed in triplicate in both pos. and neg. polarities with nitrogen gas and seven different elec. fields, so that relative std. deviations (RSD) could be assessed for each mol. and structural differences studied. The primary metabolites analyzed to date have come from key metab. pathways such as glycolysis, the pentose phosphate pathway and the tricarboxylic acid cycle, while the secondary metabolites consisted of classes such as terpenes and flavonoids, and the xenobiotics represented a range of mols. from antibiotics to polycyclic arom. hydrocarbons. Different CCS trends were obsd. for several of the diverse small mol. classes and when urine features were matched to the database, the addn. of the IMS dimension greatly reduced the possible no. of candidate mols. This CCS database and structural information are freely available for download at http://panomics.pnnl.gov/metabolites/ with new mols. being added frequently.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2sXhsFOlu7bE&md5=08bd6a4ed911bea19834302d123923b4
    36. 36
      Zhou, Z.; Xiong, X.; Zhu, Z. J. Bioinformatics 2017, 33, 2235– 2237,  DOI: 10.1093/bioinformatics/btx140
      36
      MetCCS predictor: a web server for predicting collision cross-section values of metabolites in ion mobility-mass spectrometry based metabolomics
      Zhou, Zhiwei; Xiong, Xin; Zhu, Zheng-Jiang
      Bioinformatics (2017), 33 (14), 2235-2237CODEN: BOINFP; ISSN:1460-2059. (Oxford University Press)
      In metabolomics, rigorous structural identification of metabolites presents a challenge for bioinformatics. The use of collision cross-section (CCS) values of metabolites derived from ion mobility-mass spectrometry effectively increases the confidence of metabolite identification, but this technique suffers from the limit no. of available CCS values. Currently, there is no software available for rapidly generating the metabolites' CCS values. Here, we developed the first web server, namely, MetCCS Predictor, for predicting CCS values. It can predict the CCS values of metabolites using mol. descriptors within a few seconds. Common users with limited background on bioinformatics can benefit from this software and effectively improve the metabolite identification in metabolomics.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhvV2nur7N&md5=5ca86ebe753f0fadca8cc4f26ba0e1d7
    37. 37
      Nichols, C. M.; May, J. C.; Sherrod, S. D.; McLean, J. A. Analyst 2018, 143, 1556– 1559,  DOI: 10.1039/C8AN00056E
      37
      Automated flow injection method for the high precision determination of drift tube ion mobility collision cross sections
      Nichols, Charles M.; May, Jody C.; Sherrod, Stacy D.; McLean, John A.
      Analyst (Cambridge, United Kingdom) (2018), 143 (7), 1556-1559CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
      The field of ion mobility-based omics studies requires high-quality collision cross section (CCS) libraries to effectively utilize CCS as a mol. descriptor. Abs. CCS values with the highest precision are obtained on drift tube instruments by measuring the drift time of ions at multiple drift voltages, commonly referred to as a 'stepped field' expt. However, generating large scale abs. CCS libraries from drift tube instruments is time consuming due to the current lack of high-throughput methods. This communication reports a fully automated stepped-field method to acquire abs. CCS on com. available equipment. Using a drift tube ion mobility-mass spectrometer (DTIM-MS) coupled to a minimally modified liq. chromatog. (LC) system, CCS values can be measured online with a carefully timed flow injection anal. (FIA) expt. Results demonstrate that the FIA stepped-field method yields CCS values which are of high anal. precision (<0.4% relative std. deviation, RSD) and accuracy (≤0.4% difference) comparable to CCS values obtained using traditional direct-infusion stepped-field expts. This high-throughput CCS method consumes very little sample vol. (20 μL) and will expedite the generation of large-scale CCS libraries to support mol. identification within global untargeted studies.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXks1Ois78%253D&md5=15f77a4e81c957c39c5019980edd1c13
    38. 38
      Righetti, L.; Bergmann, A.; Galaverna, G.; Rolfsson, O.; Paglia, G.; Dall’Asta, C. Anal. Chim. Acta 2018, 1014, 50– 57,  DOI: 10.1016/j.aca.2018.01.047
      38
      Ion mobility-derived collision cross section database: Application to mycotoxin analysis
      Righetti, Laura; Bergmann, Andreas; Galaverna, Gianni; Rolfsson, Ottar; Paglia, Giuseppe; Dall'Asta, Chiara
      Analytica Chimica Acta (2018), 1014 (), 50-57CODEN: ACACAM; ISSN:0003-2670. (Elsevier B.V.)
      The recent hyphenation of ion mobility spectrometry (IMS) with high resoln. mass spectrometry (HRMS) has risen as a powerful technique for both targeted and non-targeted screening, reducing background noise and allowing sepn. of isomeric and isobaric compds. Nevertheless, such an approach remains largely unexplored in food safety applications, such as mycotoxin anal. To implement ion mobility in routinely MS-based mycotoxin workflows, searchable databases with collusion cross section (CCS) values and accurate mass-values are required. This paper provides for the first time a traveling-wave IMS (TWIMS)-derived CCS database for mycotoxins, including more than 100 CCS values. The measurements showed high reproducibility (RSD<2%) across different instrumental conditions as well as several complex cereal matrixes, showing a mean inter-matrix precision of RSD <0.9%. As a proof of concept, the database was applied to the anal. of several spiked as well as naturally incurred cereal-based samples. In addn., the effect of adducts on the drift time was studied in a series of mycotoxins in order to understand potential deviations from expected drift time behaviors. Overall, our study confirmed that CCS values represent a physicochem. property that can be used alongside the traditional mol. identifiers of precursor ion accurate mass, fragment ions, isotopic pattern, and retention time.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXisFelu74%253D&md5=7258333f8021aa0c8868bb53aa0a7d3d
    39. 39
      Leaptrot, K. L.; May, J. C.; Dodds, J. N.; McLean, J. A. Nat. Commun. 2019, 10, 985,  DOI: 10.1038/s41467-019-08897-5
      39
      Ion mobility conformational lipid atlas for high confidence lipidomics
      Leaptrot Katrina L; May Jody C; Dodds James N; McLean John A
      Nature communications (2019), 10 (1), 985 ISSN:.
      Lipids are highly structurally diverse molecules involved in a wide variety of biological processes. Here, we use high precision ion mobility-mass spectrometry to compile a structural database of 456 mass-resolved collision cross sections (CCS) of sphingolipid and glycerophospholipid species. Our CCS database comprises sphingomyelin, cerebroside, ceramide, phosphatidylethanolamine, phosphatidylcholine, phosphatidylserine, and phosphatidic acid classes. Primary differences observed are between lipid categories, with sphingolipids exhibiting 2-6% larger CCSs than glycerophospholipids of similar mass, likely a result of the sphingosine backbone's restriction of the sn1 tail length, limiting gas-phase packing efficiency. Acyl tail length and degree of unsaturation are found to be the primary structural descriptors determining CCS magnitude, with degree of unsaturation being four times as influential per mass unit. The empirical CCS values and previously unmapped quantitative structural trends detailed in this work are expected to facilitate prediction of CCS in broadscale lipidomics research.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3cfos1GitA%253D%253D&md5=f4b60c23db2a9889cc167c49d40519e8
    40. 40
      Blaženović, I.; Shen, T.; Mehta, S. S.; Kind, T.; Ji, J.; Piparo, M.; Cacciola, F.; Mondello, L.; Fiehn, O. Anal. Chem. 2018, 90 (18), 10758– 10764,  DOI: 10.1021/acs.analchem.8b01527
      40
      Increasing Compound Identification Rates in Untargeted Lipidomics Research with Liquid Chromatography Drift Time-Ion Mobility Mass Spectrometry
      Blazenovic, Ivana; Shen, Tong; Mehta, Sajjan S.; Kind, Tobias; Ji, Jian; Piparo, Marco; Cacciola, Francesco; Mondello, Luigi; Fiehn, Oliver
      Analytical Chemistry (Washington, DC, United States) (2018), 90 (18), 10758-10764CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Unknown metabolites represent a bottleneck in untargeted metabolomics research. Ion mobility-mass spectrometry (IM-MS) facilitates lipid identification because it yields collision cross section (CCS) information that is independent from mass or lipophilicity. To date, only a few CCS values are publicly available for complex lipids such as phosphatidylcholines, sphingomyelins, or triacylglycerides. This scarcity of data limits the use of CCS values as an identification parameter that is orthogonal to mass, MS/MS, or retention time. A combination of lipid descriptors was used to train five different machine learning algorithms for automatic lipid annotations, combining accurate mass (m/z), retention time (RT), CCS values, carbon no., and unsatn. level. Using a training data set of 429 true pos. lipid annotations from four lipid classes, 92.7% correct annotations overall were achieved using internal cross-validation. The trained prediction model was applied to an unknown milk lipidomics data set and allowed for class 3 level annotations of most features detected in this application set according to Metabolomics Stds. Initiative (MSI) reporting guidelines.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhsVOltrvO&md5=9bb9b150edfe9444863e2a7274fa0c98
    41. 41
      Hines, K. M.; May, J. C.; McLean, J. A.; Xu, L. Anal. Chem. 2016, 88, 7329– 7336,  DOI: 10.1021/acs.analchem.6b01728
      41
      Evaluation of Collision Cross Section Calibrants for Structural Analysis of Lipids by Traveling Wave Ion Mobility-Mass Spectrometry
      Hines, Kelly M.; May, Jody C.; McLean, John A.; Xu, Libin
      Analytical Chemistry (Washington, DC, United States) (2016), 88 (14), 7329-7336CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Collision cross section (CCS) measurement of lipids using traveling wave ion mobility-mass spectrometry (TWIM-MS) is of high interest to the lipidomics field. However, currently available calibrants for CCS measurement using TWIM are predominantly peptides that display quite different phys. properties and gas-phase conformations from lipids, which could lead to large CCS calibration errors for lipids. Here we report the direct CCS measurement of a series of phosphatidylcholines (PCs) and phosphatidylethanolamines (PEs) in nitrogen using a drift tube ion mobility (DTIM) instrument and an evaluation of the accuracy and reproducibility of PCs and PEs as CCS calibrants for phospholipids against different classes of calibrants, including polyalanine (PolyAla), tetraalkylammonium salts (TAA), and hexakis(fluoroalkoxy)phosphazines (HFAP), in both pos. and neg. modes in TWIM-MS anal. We demonstrate that structurally mismatched calibrants lead to larger errors in calibrated CCS values while the structurally matched calibrants, PCs and PEs, gave highly accurate and reproducible CCS values at different traveling wave parameters. Using the lipid calibrants, the majority of the CCS values of several classes of phospholipids measured by TWIM are within 2% error of the CCS values measured by DTIM. The development of phospholipid CCS calibrants will enable high-accuracy structural studies of lipids and add an addnl. level of validation in the assignment of identifications in untargeted lipidomics expts.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhtVSltb7J&md5=6c37b06c11ffbdc24ac1bfc9961f9e40
    42. 42
      Reymond, J.-L.; Awale, M. ACS Chem. Neurosci. 2012, 3, 649– 657,  DOI: 10.1021/cn3000422
      42
      Exploring Chemical Space for Drug Discovery Using the Chemical Universe Database
      Reymond, Jean-Louis; Awale, Mahendra
      ACS Chemical Neuroscience (2012), 3 (9), 649-657CODEN: ACNCDM; ISSN:1948-7193. (American Chemical Society)
      Herein we review our recent efforts in searching for bioactive ligands by enumeration and virtual screening of the unknown chem. space of small mols. Enumeration from first principles shows that almost all small mols. (>99.9%) have never been synthesized and are still available to be prepd. and tested. We discuss open access sources of mols., the classification and representation of chem. space using mol. quantum nos. (MQN), its exhaustive enumeration in form of the chem. universe generated databases (GDB), and examples of using these databases for prospective drug discovery. MQN-searchable GDB, PubChem, and DrugBank are freely accessible at www.gdb.unibe.ch.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XmtValu70%253D&md5=bbbe3c0931328f3796ce999189374864
    43. 43
      Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V.; Vanderplas, J.; Passos, A.; Cournapeau, D.; Brucher, M.; Perrot, M.; Duchesnay, E. J. Mach Learn Res. 2011, 12, 2825– 2830
      There is no corresponding record for this reference.
    44. 44
      Picache, J. A.; Rose, B. S.; Balinski, A.; Leaptrot, K. L.; Sherrod, S. D.; May, J. C.; McLean, J. A. Chem. Sci. 2019, 10, 983– 993,  DOI: 10.1039/C8SC04396E
      44
      Collision cross section compendium to annotate and predict multi-omic compound identities
      Picache, Jaqueline A.; Rose, Bailey S.; Balinski, Andrzej; Leaptrot, Katrina L.; Sherrod, Stacy D.; May, Jody C.; McLean, John A.
      Chemical Science (2019), 10 (4), 983-993CODEN: CSHCCN; ISSN:2041-6520. (Royal Society of Chemistry)
      Ion mobility mass spectrometry (IM-MS) expands the analyte coverage of existing multi-omic workflows by providing an addnl. sepn. dimension as well as a parameter for characterization and identification of mols. - the collision cross section (CCS). This work presents a large, Unified CCS compendium of >3800 exptl. acquired CCS values obtained from traceable mol. stds. and measured with drift tube ion mobility-mass spectrometers. An interactive visualization of this compendium along with data analytic tools have been made openly accessible. Represented in the compendium are 14 structurally-based chem. super classes, consisting of a total of 80 classes and 157 subclasses. Using this large data set, regression fitting and predictive statistics have been performed to describe mass-CCS correlations specific to each chem. ontol. These structural trends provide a rapid and effective filtering method in the traditional untargeted workflow for identification of unknown biochem. species. The utility of the approach is illustrated by an application to metabolites in human serum, quantified trends of which were used to assess the probability of an unknown compd. belonging to a given class. CCS-based filtering narrowed the chem. search space by 60% while increasing the confidence in the remaining isomeric identifications from a single class, thus demonstrating the value of integrating predictive analyses into untargeted expts. to assist in identification workflows. The predictive abilities of this compendium will improve in specificity and expand to more chem. classes as addnl. data from the IM-MS community is contributed. Instructions for data submission to the compendium and criteria for inclusion are provided.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXitlSqtrbN&md5=177c56f0b3557ca849e5f224ffd5a224
    45. 45
      Hernandez-Mesa, M.; Le Bizec, B.; Monteau, F.; Garcia-Campana, A. M.; Dervilly-Pinel, G. Anal. Chem. 2018, 90, 4616– 4625,  DOI: 10.1021/acs.analchem.7b05117
      45
      Collision cross section (CCS) database: An additional measure to characterize steroids
      Hernandez-Mesa, Maykel; Le Bizec, Bruno; Monteau, Fabrice; Garcia-Campana, Ana M.; Dervilly-Pinel, Gaud
      Analytical Chemistry (Washington, DC, United States) (2018), 90 (7), 4616-4625CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Ion mobility spectrometry enhances the performance characteristics of liq. chromatog.-mass spectrometry workflows intended to steroid profiling by providing a new sepn. dimension and a novel characterization parameter, the so-called collision cross section (CCS). This work proposes the first CCS database for 300 steroids (i.e., endogenous, including phase I and phase II metabolites, and exogenous synthetic compds.), which involves 1080 ions and covers the CCS of 127 androgens, 84 estrogens, 50 corticosteroids, and 39 progestagens. This large database provides information related to all the ionized species identified for each steroid in pos. electrospray ionization mode as well as for estrogens in neg. ionization mode. CCS values have been measured using nitrogen as drift gas in the ion mobility cell. Generally, direct correlation exists between mass-to-charge ratio (m/z) and CCS because both are related parameters. However, several steroids mainly steroid glucuronides and steroid esters have been characterized as more compact or elongated mols. than expected. In such cases, CCS results in addnl. relevant information to retention time and mass spectral data for the identification of steroids. Moreover, several isomeric steroid pairs (e.g., 5β-androstane-3,17-dione and 5α-androstane-3,17-dione) have been sepd. based on their CCS differences. These results indicate that adding the CCS to databases in anal. workflows increases selectivity, thus improving the confidence in steroids anal. Consequences in terms of identification and quantification are discussed. Quality criteria and a construction of an interlab. reproducibility approach are also reported for the obtained CCS values. The CCS database described here is made publicly available.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXktlGisL8%253D&md5=b73301562f0f191833005d2a5f3b9fc2
    46. 46
      Sumner, L. W.; Amberg, A.; Barrett, D.; Beale, M. H.; Beger, R.; Daykin, C. A.; Fan, T. W. M.; Fiehn, O.; Goodacre, R.; Griffin, J. L.; Hankemeier, T.; Hardy, N.; Harnly, J.; Higashi, R.; Kopka, J.; Lane, A. N.; Lindon, J. C.; Marriott, P.; Nicholls, A. W.; Reily, M. D.; Thaden, J. J.; Viant, M. R. Metabolomics 2007, 3 (3), 211– 221,  DOI: 10.1007/s11306-007-0082-2
      46
      Proposed minimum reporting standards for chemical analysis. Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI)
      Sumner, Lloyd W.; Amberg, Alexander; Barrett, Dave; Beale, Michael H.; Beger, Richard; Daykin, Clare A.; Fan, Teresa W.-M.; Fiehn, Oliver; Goodacre, Royston; Griffin, Julian L.; Hankemeier, Thomas; Hardy, Nigel; Harnly, James; Higashi, Richard; Kopka, Joachim; Lane, Andrew N.; Lindon, John C.; Marriott, Philip; Nicholls, Andrew W.; Reily, Michael D.; Thaden, John J.; Viant, Mark R.
      Metabolomics (2007), 3 (3), 211-221CODEN: METAHQ; ISSN:1573-3882. (Springer)
      There is a general consensus that supports the need for standardized reporting of metadata or information describing large-scale metabolomics and other functional genomics data sets. Reporting of std. metadata provides a biol. and empirical context for the data, facilitates exptl. replication, and enables the reinterrogation and comparison of data by others. Accordingly, the Metabolomics Stds. Initiative is building a general consensus concerning the min. reporting stds. for metabolomics expts. of which the Chem. Anal. Working Group (CAWG) is a member of this community effort. This article proposes the min. reporting stds. related to the chem. anal. aspects of metabolomics expts. including: sample prepn., exptl. anal., quality control, metabolite identification, and data pre-processing. These min. stds. currently focus mostly upon mass spectrometry and NMR spectroscopy due to the popularity of these techniques in metabolomics. However, addnl. input concerning other techniques is welcomed and can be provided via the CAWG online discussion forum at http://msi-workgroups.sourceforge.net/ or http://[email protected]. Further, community input related to this document can also be provided via this electronic forum.
      >> More from SciFinder ®
      https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhtFOntbvL&md5=e636dad276f9f1b2e756db169368eb53

  • Supporting Information

    Supporting Information

    The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.9b05772.

    • All code used to assemble the CCS database and train predictive models is available on GitHub (https://github.com/dylanhross/c3sdb); SI includes Experimental Section and schema; MS adduct encodings, molecular quantum numbers (MQNs), top three features contributing to separation along PC3, CCS prediction accuracy of LipidCCS on different chemical classes, and feature selection trials; and additional Results and Discussion (PDF)


    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.