Analysis of Liquid Chromatography–Mass Spectrometry Data with an Elastic Net Multivariate Curve Resolution Strategy for Sparse Spectral Recovery

Use of Fisher's Ratio assisted multivariate curve resolution- alternating least squares for discovery-based analysis using ultrahigh pressure liquid chromatography-high resolution mass spectrometry

Non-targeted analysis of complex chemical mixtures can be difficult considering the convoluted nature of the matrix and the potential unknown chemical differences between samples or classes of samples. Ultrahigh pressure liquid chromatography coupled to quadrupole time-of-flight mass spectrometry (UHPLC-QTOF) is an ideal technique to probe chemical differences for a wide variety of samples. While UHPLC-QTOF can discover minute chemical differences down to low part per billion (ppb) concentrations with a high degree of confidence, the application of high-resolution mass spectrometry can yield massive amounts of information (∼ 10 gb per sample) that cannot be analyzed manually. Therefore, the application of chemometric techniques is mandatory for the interrogation of complex samples. Fisher's ratio (FR) assisted multivariate curve resolution-alternating least squares (MCR-ALS) was used to the discover and identify the chemical differences between two classes of materials: 1) a pond water matrix and 2) the matrix spiked with a pharmaceutical standard mix containing 17 compounds. Thirteen of the seventeen spiked compounds were discovered using FR analysis, and then five were successfully deconvoluted using MCR-ALS wherein the number of curves chosen were automatically determined using singular value decomposition (SVD). The use of an automated FR assisted MCR-ALS will aid in discovering trace levels of chemical components without the need for the researcher to provide potentially biased input which will aid in non-targeted workflow.

Optimization of Parameters for ROI Data Compression for Nontargeted Analyses Using LC-HRMS

Nontargeted analyses of low-concentration analytes in the information-rich data collected by liquid chromatography with high-resolution mass spectrometry detection can be challenging to accomplish in an efficient and comprehensive manner. The aim of this study is to demonstrate a workflow involving targeted parameter optimization for entire chromatograms using region of interest (ROI) data compression uncoupled from a subsequent tile-based Fisher ratio (F-ratio) analysis, a supervised discovery-based method, for the discovery of low-concentration analytes. Soil samples spiked with 18 pesticides at nominal concentrations ranging from 0.1 to 50 ppb for a total of six sample classes served as challenging samples to demonstrate the overall workflow. Optimization of two parameters proved to be the most critical for ROI data compression: the signal threshold parameter and the admissible mass deviation parameter. The parameter optimization method workflow we introduce is based upon spiking known analytes into a representative sample and determining the number of detectable spikes and the Δppm for various combinations of the signal threshold and admissible mass deviation, where Δppm is the absolute value of the difference between the theoretical m/z and the ROI m/z. Once optimal parameters are determined providing the lowest average Δppm and the greatest number of detectable analytes, the optimized parameters can be utilized for the intended analysis. Herein, tile-based F-ratio analysis was performed on the ROI compressed data of all spiked soil samples first by applying ROI parameters recommended in the literature, referred to herein as the initial ROI parameters, and finally by the combination of the two optimized parameters. Using the initial ROI parameters, three pesticides were discovered, whereas all 18 spiked pesticides were discovered by optimizing both ROI parameters.

Peak Identification and Quantification by Proteomic Mass Spectrogram Decomposition

Recent advances in the liquid chromatography/mass spectrometry (LC/MS) technology have improved the sensitivity, resolution, and speed of proteome analysis, resulting in increasing demand for more sophisticated algorithms to interpret complex mass spectrograms. Here, we propose a novel statistical method, proteomic mass spectrogram decomposition (ProtMSD), for joint identification and quantification of peptides and proteins. Given the proteomic mass spectrogram and the reference mass spectra of all possible peptide ions associated with proteins as a dictionary, ProtMSD estimates the chromatograms of those peptide ions under a group sparsity constraint without using the conventional careful preprocessing (e.g., thresholding and peak picking). We show that the method was significantly improved using protein-peptide hierarchical relationships, isotopic distribution profiles, reference retention times of peptide ions, and prelearned mass spectra of noise. We examined the concept of database search, library search, and match-between-runs. Our ProtMSD showed excellent agreements of 3277 peptide ions (94.79%) and 493 proteins (98.21%) with Mascot/Skyline for an Escherichia coli proteome sample and of 4460 peptide ions (103%) and 588 proteins (101%) with match-between-runs by MaxQuant for a yeast proteome sample. This is the first attempt to use a matrix decomposition technique as a tool for LC/MS-based proteome identification and quantification.

Self-Optimizing Support Vector Elastic Net

Chemometrics is widely used to solve various quantitative and qualitative problems in analytical chemistry. A self-optimizing chemometrics method facilitates scientists to exploit the advantages of chemometrics. In this report, a parameter-free support vector elastic net that self-optimizes two key regularization constants, i.e., λ for L2 regularization and t for L1 regularization, is developed and referred to as self-optimizing support vector elastic net (SOSVEN). Response surface modeling (RSM) and bootstrapped Latin partitions (BLPs) are incorporated for the optimization. Responses at a set of design points over the ranges of the two factors are evaluated with an internal BLP validation using a calibration set. A 2-dimensional interpolation with a cubic spline fits a response surface to determine the best condition that gives the best-estimated response. The SOSVEN with RSM had comparable performances with the one tuned by grid search, while the RSM is more efficient. The developed SOSVEN was compared with two parameter-free chemometrics methods, super partial least-squares regression (sPLSR) and super support vector regression (sSVR) for calibration, and sPLS-discriminant analysis (sPLS-DA) and support vector classification (SVC) for classification. For calibration, the SOSVEN with RSM worked equivalently well or better than the other two self-optimizing methods for the evaluations using meat and hemp oil data sets. For classification, a reference wine data set and mass spectra of different marijuana extracts were used. The three classifiers had similar performances to identify the cultivars of wines with nearly 98% of accuracy. The SOSVEN significantly outperformed sPLS-DA and SVC to classify the mass spectra of marijuana extracts with an overall accuracy of 97%. These results demonstrated excellent abilities of SOSVEN for classification and calibration.

Recent applications of chemometrics in one- and two-dimensional chromatography

The proliferation of increasingly more sophisticated analytical separation systems, often incorporating increasingly more powerful detection techniques, such as high-resolution mass spectrometry, causes an urgent need for highly efficient data-analysis and optimization strategies. This is especially true for comprehensive two-dimensional chromatography applied to the separation of very complex samples. In this contribution, the requirement for chemometric tools is explained and the latest developments in approaches for (pre-)processing and analyzing data arising from one- and two-dimensional chromatography systems are reviewed. The final part of this review focuses on the application of chemometrics for method development and optimization.

Urine Metabolic Fingerprints Encode Subtypes of Kidney Diseases

Metabolic fingerprints of biofluids encode diverse diseases and particularly urine detection offers complete non-invasiveness for diagnostics of the future. Present urine detection affords unsatisfactory performance and requires advanced materials to extract molecular information, due to the limited biomarkers and high sample complexity. Herein, we report plasmonic polymer@Ag for laser desorption/ionization mass spectrometry (LDI-MS) and sparse-learning-based metabolic diagnosis of kidney diseases. Using only 1 μL of urine without enrichment or purification, polymer@Ag afforded urine metabolic fingerprints (UMFs) by LDI-MS in seconds. Analysis by sparse learning discriminated lupus nephritis from various other non-lupus nephropathies and controls. We combined UMFs with urine protein levels (UPLs) and constructed a new diagnostic model to characterize subtypes of kidney diseases. Our work guides urine-based diagnosis and leads to new personalized analytical tools for other diseases.

ROIMCR: a powerful analysis strategy for LC-MS metabolomic datasets

BACKGROUND

The analysis of LC-MS metabolomic datasets appears to be a challenging task in a wide range of disciplines since it demands the highly extensive processing of a vast amount of data. Different LC-MS data analysis packages have been developed in the last few years to facilitate this analysis. However, most of these strategies involve chromatographic alignment and peak shaping and often associate each "feature" (i.e., chromatographic peak) with a unique m/z measurement. Thus, the development of an alternative data analysis strategy that is applicable to most types of MS datasets and properly addresses these issues is still a challenge in the metabolomics field.

RESULTS

Here, we present an alternative approach called ROIMCR to: i) filter and compress massive LC-MS datasets while transforming their original structure into a data matrix of features without losing relevant information through the search of regions of interest (ROIs) in the m/z domain and ii) resolve compressed data to identify their contributing pure components without previous alignment or peak shaping by applying a Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) analysis. In this study, the basics of the ROIMCR method are presented in detail and a detailed description of its implementation is also provided. Data were analyzed using the MATLAB (The MathWorks, Inc., www.mathworks.com ) programming and computing environment. The application of the ROIMCR methodology is described in detail, with an example of LC-MS data generated in a lipidomic study and with other examples of recent applications.

CONCLUSIONS

The methodology presented here combines the benefits of data filtering and compression based on the searching of ROI features, without the loss of spectral accuracy. The method has the benefits of the application of the powerful MCR-ALS data resolution method without the necessity of performing chromatographic peak alignment or modelling. The presented method is a powerful alternative to other existing data analysis approaches that do not use the MCR-ALS method to resolve LC-MS data. The ROIMCR method also represents an improved strategy compared to the direct applications of the MCR-ALS method that use less-powerful data compression strategies such as binning and windowing. Overall, the strategy presented here confirms the usefulness of the ROIMCR chemometrics method for analyzing LC-MS untargeted metabolomics data.

An initial estimation method using cosine similarity for multivariate curve resolution: application to NMR spectra of chemical mixtures

Mixture spectra is decomposed into pure spectra without prior knowledge, and the MCR calculation refines the spectra and provides the concentrations.