Efficiency of automated peak identification with daily used commercial software for X‐ray photoelectron spectra ‐ report from VAMAS/TWA 2/A 9 project

In the first step of X‐ray photoelectron spectroscopy (XPS) analysis, photoelectron peaks in a survey spectrum for the specimen of interest are generally identified as arising from particular elements and electron energy levels. This peak identification is performed using sophisticated software on commercial XPS instruments. For the Versailles Project on Advanced Materials and Standards (VAMAS)/Technical Working Area (TWA) 2/A 9 project, automated peak‐identification efficiency using commercial software utilities has been investigated with simulated test XPS spectra. This efficiency has been investigated with reference peaks that were detected by eye in our previous work [SIA, 2008, 40, 1337]. We found (i) one software that identified the binding energy (BE) at the peak maximum for a fitted curve in the peak region, a second that reported the BE stored in its database, and a third that reported the BE as the local maximum, (ii) software does not identify some weak peaks even though they were obviously detected by analysts, (iii) only one major peak of a doublet appeared to be in software database when the energy separation is less than about 4 eV such as 3.6 eV for Au 4f, and (iv) the identification efficiency is larger than 90% when the ratio between peak height and noise amplitude is greater than 10. We suggest several things for improving software design to assist non‐expert analysts. Copyright © 2013 John Wiley & Sons, Ltd.     

Evaluation of uncertainties in X‐ray photoelectron spectroscopy intensities associated with different methods and procedures for background subtraction. I. Spectra for monochromatic Al X‐ray

We report uncertainties in X‐ray photoelectron spectroscopy (XPS) intensities arising from commonly used methods and procedures for subtraction of the spectral background. These uncertainties were determined from a comparison of XPS intensities reported by volunteer analysts from 28 institutions and the corresponding intensities expected for a set of simulated XPS spectra. We analyzed peak intensities from 32 sets of data for a group of 12 spectra that had been simulated for a monochromated Al Kα source. Each reported intensity was compared with an expected intensity for the particular integration limits chosen by each analyst and known from the simulation design. We present ratios of the reported intensities to the expected intensities for the background‐subtraction methods chosen by the analysts. These ratios were close to unity in most cases, as expected, but deviations were found in the results from some analysts, particularly if the main peak was asymmetrical or if shakeup was present. We showed that better results for the Shirley, Tougaard, and linear backgrounds were obtained when analysts determined peak intensities over certain energy ranges or integration limits. We then were able to recommend integration limits that should be a useful guide in the determination of peak intensities for other XPS spectra. The use of relatively narrow integration limits with the Shirley and linear backgrounds, however, will lead to measures of peak intensities that are less than the total intensities. Although these measures may be satisfactory for some quantitative analyses, errors in quantitative XPS analyses can occur if there are changes in XPS lineshapes or shakeup fractions with change of chemical state. The use of curve‐fitting equations to fit an entire spectrum will generally exclude the shakeup contribution to the intensity of the main peak, and no account will be taken of any variation in the shakeup fraction with change of chemical state. Published in 2009 by John Wiley & Sons, Ltd. Certain commercial products are identified to specify the formats in which the test spectra were distributed and the software with which the test spectra were analyzed by participants. This identification does not imply that the products are endorsed or recommended by the National Institute of Standards and Technology, or that they are necessarily the most suitable for the purposes described.     

Evaluation of uncertainties in X‐ray photoelectron spectroscopy intensities associated with different methods and procedures for background subtraction. II. Spectra for unmonochromated Al and Mg X‐rays

We report uncertainties in X‐ray photoelectron spectroscopy (XPS) intensities arising from commonly used methods and procedures for subtraction of the spectral background. These uncertainties were determined from a comparison of XPS intensities reported by volunteer analysts and the corresponding intensities expected for a set of simulated XPS spectra. We analyzed peak intensities from 16 sets of data (submitted from 15 institutions) for a group of 12 spectra that had been simulated for an unmonochromated Al‐Kα source and similar intensities from 20 sets of data (submitted from 17 institutions) that had been simulated for an unmonochromated Mg‐Kα source. Each reported intensity was compared with an expected intensity for the particular integration limits chosen by each analyst and known from the simulation design. We present ratios of the reported intensities to the expected intensities for the background‐subtraction methods chosen by the analysts. These ratios were close to unity in most cases, as expected, but deviations were found in the results from some analysts, particularly if shakeup was present. We showed that better results for the Shirley and Tougaard backgrounds were obtained when analysts determined peak intensities over certain energy ranges or integration limits. We then were able to suggest integration limits that should be a useful guide in the determination of peak intensities for other XPS spectra. The use of relatively narrow integration limits with the Shirley and linear backgrounds, however, will lead to measures of peak intensity that are less than the total intensities. Although these measures may be satisfactory for some quantitative analyses, errors in quantitative XPS analyses can occur if there are changes in XPS lineshapes or shakeup fractions with change of chemical state. The use of curve‐fitting equations to fit an entire spectrum will generally exclude the shakeup contribution to the intensity of the main peak, and any variation in the shakeup fraction with change of chemical state will not be taken into account. Published in 2009 by John Wiley & Sons, Ltd.     

Automated discrimination of polymerase chain reaction products with closely related sequences by software-based detection of characteristic peaks in product ion spectra

A computer-based method is described for automated detection of peaks in product ion spectra that allows discrimination of structurally related polymerase chain reaction (PCR) products. PCR products of K-ras mutants having single nucleotide substitutions and isomeric sequence changes in positions 1 and 2 of codon 12 (e.g. TGT and GTT) were used as a model system. SpecDiff, a tool for differentiating pairs of mass spectra by identifying peaks that either differ in relative intensity between spectra or only appear in one of a pair of spectra, was created to help automate detection. This program was demonstrated to have great utility in detection of mutations and could also be useful as a general tool for differentiating other molecules of closely related structure.     

A new method to analyze matrix-assisted laser desorption/ionization time-of-flight peptide profiling mass spectra

In protein and peptide mass spectrometry in which profiling of peaks is involved, their masses and intensities are important characteristics. Because of the relative low reproducibility of peak intensities associated with complex samples in matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS), it is difficult to accurately assess the number of peaks and their intensities. In this study we evaluate these two characteristics for tryptic digests of cerebro-spinal fluid. We observed that the reproducibility of peak intensities was relatively poor (CV = 42%) and that additional normalization or spiking did not lead to a large improvement (CV = 30%). Moreover, at least seven mass spectra per sample were required to obtain a reliable peak list. An improvement of the sensitivity (i.e., eventually more peaks are detected) is observed if more replicates per sample are measured. We conclude that the reproducibility and sensitivity of peptide profiling can be significantly improved by a combination of measuring at least seven spectra per sample with a dichotomous scoring of the intensities. This approach will aid the analysis of large numbers of mass spectra of patient samples in a reproducible way for the detection and validation of candidate biomarkers.     

Development of an algorithm for peak detection in comprehensive two-dimensional chromatography

A method for peak detection in two-dimensional chromatography is presented. The algorithm applies first the methods developed for peak detection in one-dimensional chromatography to detect peaks in one dimension. In a second step, a decision tree is applied to decide which one-dimensional peaks are originated from the same compound and have to be 'merged' into one two-dimensional peak. To this end, different features of the peaks (second-dimension peak regions and second-dimension retention times) are compared and different criteria (common peak regions, retention time differences, unimodality in the first dimension) are applied. Different options can be used, depending on the nature of the data. The user controls this decision tree by establishing several options and "switches". The algorithm was tested with GCxGC chromatograms obtained for a commercial air-freshener sample, detecting and merging the modulated peaks belonging to the same compound. Recommendations for the set of options and switches are given. A utility that calculates and sums peak areas from merged peaks is added to facilitate automated quantification. Although the algorithm was developed for GCxGC, its application to comprehensive two-dimensional liquid chromatography (LCxLC) data should at most require minor modifications.     

On the risk of false positive identification using multiple ion monitoring in qualitative mass spectrometry: Large-scale intercomparisons with a comprehensive mass spectral library

Analysts involved in qualitative mass spectrometry have long debated the minimum data requirements for demonstrating that signals from an unknown sample are identical to those from a known compound. Often this process is carried out by comparing a few selected ions acquired by multiple ion monitoring (MIM), with due allowance for expected variability in response. In a few past experiments with electron-ionization mass spectrometry (EI-MS), the number of ions selected and the allowable variability in relative abundance were tested by comparing one spectrum against a library of mass spectra, where library spectra served to represent potential false positive signals in an analysis. We extended these experiments by carrying out large-scale intercomparisons between thousands of spectra and a library of one hundred thousand EI mass spectra. The results were analyzed to gain insights into the identification confidence associated with various numbers of selected ions. A new parameter was investigated for the first time, to take into account that a library spectrum with a different base peak than the search spectrum may still cause a false positive identification. The influence of peak correlation among the specific ions in all the library mass spectra was also studied. Our computations showed that (1) false positive identifications can result from similar compounds, or low-abundance peaks in unrelated compounds if the method calls for detection at very low levels; (2) a MIM method's identification confidence improves in a roughly continuous manner as more ions are monitored, about one order of magnitude for each additional ion selected; (3) full scan spectra still represent the best alternative, if instrument sensitivity is adequate. The use of large scale intercomparisons with a comprehensive library is the only way to provide direct evidence in support of these conclusions, which otherwise depend on the judgment and experience of individual analysts. There are implications for residue chemists who would rely on standardized confirmation criteria to assess the validity of a given confirmatory method. For example, standardized confirmation criteria should not be used in the absence of interference testing and rational selection of diagnostic ions.     

Automatic poisson peak harvesting for high throughput protein identification

High throughput identification of proteins by peptide mass fingerprinting requires an efficient means of picking peaks from mass spectra. Here, we report the development of a peak harvester to automatically pick monoisotopic peaks from spectra generated on matrix-assisted laser desorption/ionisation time of flight (MALDI-TOF) mass spectrometers. The peak harvester uses advanced mathematical morphology and watershed algorithms to first process spectra to stick representations. Subsequently, Poisson modelling is applied to determine which peak in an isotopically resolved group represents the monoisotopic mass of a peptide. We illustrate the features of the peak harvester with mass spectra of standard peptides, digests of gel-separated bovine serum albumin, and with Escherictia coli proteins prepared by two-dimensional polyacrylamide gel electrophoresis. In all cases, the peak harvester proved effective in its ability to pick similar monoisotopic peaks as an experienced human operator, and also proved effective in the identification of monoisotopic masses in cases where isotopic distributions of peptides were overlapping. The peak harvester can be operated in an interactive mode, or can be completely automated and linked through to peptide mass fingerprinting protein identification tools to achieve high throughput automated protein identification.     

Simplifying LR‐HSQC spectra using a triple‐quantum filter: The LR‐HTQC experiment

Long‐range heteronuclear single quantum correlation (LR‐HSQC) experiments may be applied for detecting long‐range correlations but suffer from two disadvantages, common to all heteronuclear long‐range correlation experiments: (i) The information density in LR‐HSQC spectra may be too high to be used directly without “filtering out” shorter range correlations, and (ii) often, substantial differences in intensity among cross peaks exist, potentially hampering the visualization of weak, often crucial cross peaks. In this contribution, we propose a modified LR‐HSQC experiment, the LR‐HTQC experiment (Long‐Range Heteronuclear Triple Quantum Correlation) that partially solves the problems aforementioned. We show theoretically and experimentally that the LR‐HTQC experiment removes the intense cross peaks of CH spin pairs, substantially reduces the medium intensity of cross peaks originating from CHH' spin systems, whereas the typically weak intensity of cross peaks of CHH'H″ and C(H)_n, _{n > 3} spin systems is less affected. Consequently, the LR‐HTQC experiment affords simplified long‐range heteronuclear shift correlation spectra and scales down large intensity differences among different types of cross peaks, although a certain general reduction of signal intensities has to be accepted.     

Outlier detection for the Generalized Rank Annihilation Method in HPLC-DAD analysis

The Generalized Rank Annihilation Method (GRAM) is a second-order calibration method that is used in chromatography to quantify analytes that coelute with interferences. For a correct quantification, the peak of the analyte in the standard and in the test sample must be aligned and have the same shape (i.e., have a trilinear structure). Variations in retention time and shape between the two peaks may cause the test sample to behave as an outlier and produce an incorrect prediction. This situation cannot be detected by checking the coincidence of the recovered spectrum with the known spectrum of the analyte because the spectral domain is not affected. It cannot be detected either by checking if the recovered profile is correct (i.e., unimodal and positive). Several plots are presented to detect such outliers. The first plot compares the particular elution profiles in the standard and in the test sample that are recovered by least-squares regression from the spectra estimated by GRAM. The calculated elution profiles from both peaks should coincide. A second plot uses the elution profiles and spectra calculated by GRAM to define the vector space spanned by the interferences. The measured peaks in the standard and in the test sample are projected onto the space that is orthogonal to the space spanned by the interferences. These projections are proportional (up to the noise) if data are trilinear. The proportionality is checked graphically from the first singular vector of the projected peaks, or from the plot of the orthogonal signal versus the net sensitivity. The use of these graphs is shown for simulated data and for the determination of 4-nitrophenol in river water samples with liquid chromatography/UV-Vis detection.     

Mass spectrometry of DNA mixtures by laser ablation from frozen aqueous solution.

We report time-of-flight mass spectra of test mixtures of six single-stranded DNA segments. The segments range in size from 8 to 60 nucleotides (molecular weight range 2413 to 18,602 Da). The best mass spectra were obtained by pulsed laser ablation of thin frozen films of an aqueous solution of the mixture from an oxidized copper substrate. These mass spectra are dominated by the molecular-ion peak for each DNA segment, and show little evidence of fragmentation, peak broadening or cluster formation. In contrast, mass spectra obtained using UV laser ablation from an anthranilic acid matrix yield broad peaks with evidence of fragmentation, and DNA segments longer than 26 nucleotides are difficult to detect.     

Percolation-to-droplets transition during spinodal decomposition in polymer blends, morphology analysis

Phase separation kinetics of the off-critical mixture of polystyrene and poly(methylphenylsiloxane) is studied by the time-resolved light scattering and optical microscopy. The results from the light scattering experiments are correlated with the images obtained by the optical microscopic observation in order to find characteristic features of the scattering intensity during the percolation-to-droplets morphology transition. At the beginning of the spinodal decomposition process only a bicontinuous network is present in the system and the light scattering intensity has only one peak. The network coarsens and at the same time small droplets appear in the system resulting in a growth of the scattering intensity at very small wave vectors. When the large network starts to break up into disjoint elongated domains a second peak in the scattering intensity appears. Finally, both peaks merge into a single peak at zero wave vector, indicating a complete transformation of elongated domains into spherical droplets of variable sizes. The comparison of the direct microscopic observations with the light scattering spectra shows that the process of breaking up of the bicontinuous network starts when the growth of the first peak, corresponding to the bicontinuous pattern, becomes very slow (essentially pinned down).     

Separation and detection of individual submicron particles by capillary electrophoresis with laser-light-scattering detection

Separation and detection of individual submicron polystyrene spheres using capillary electrophoresis with laser-light-scattering detection has been demonstrated. Electrophoretically separated particles were passed through a focused laser beam and light scattered from individual particles was collected at 90 degrees. Each diameter of polystyrene spheres injected (from 110 to 992 nm) resulted in the observation of a well-defined migration window containing multiple peaks, each arising from the light scattered by an individual particle. The migration time window for individual particles of a particular size corresponded well to the migration time of a peak from a population of particles of the same size detected using a UV absorbance detector. The electrophoretic mobility and scattered light intensity were determined for each particle detected. The average scattered light intensity for each particle size was consistent with Mie scattering theory. Particles as small as 110 nm in diameter were detected individually using this method, but particles with a diameter of 57 nm could not be individually detected. The number of single particle scattering events was counted and compared to the theoretical number of particles injected electrokinetically, and the detection efficiency determined ranged from 38 to 57% for polystyrene spheres of different sizes. The laser-light-scattering detection method was directly compared to laser-induced fluorescence detection using fluorescent polystyrene microspheres. The number of particles detected individually by each method was in agreement.     

Probing Invisible,Excited Protein States by Non‐Uniformly Sampled Pseudo‐4D CEST Spectroscopy

Chemical exchange saturation transfer (CEST) NMR spectroscopy is a powerful tool for studies of slow timescale protein dynamics. Typical experiments are based on recording a large number of 2D data sets and quantifying peak intensities in each of the resulting planes. A weakness of the method is that peaks must be resolved in 2D spectra, limiting applications to relatively small proteins. Resolution is significantly improved in 3D spectra but recording uniformly sampled data is time‐prohibitive. Here we describe non‐uniformly sampled HNCO‐based pseudo‐4D CEST that provides excellent resolution in reasonable measurement times. Data analysis is done through fitting in the time domain, without the need of reconstructing the frequency dimensions, exploiting previously measured accurate peak positions in reference spectra. The methodology is demonstrated on several protein systems, including a nascent form of superoxide dismutase that is implicated in neurodegenerative disease.     

Probing Invisible,Excited Protein States by Non‐Uniformly Sampled Pseudo‐4D CEST Spectroscopy

Chemical exchange saturation transfer (CEST) NMR spectroscopy is a powerful tool for studies of slow timescale protein dynamics. Typical experiments are based on recording a large number of 2D data sets and quantifying peak intensities in each of the resulting planes. A weakness of the method is that peaks must be resolved in 2D spectra, limiting applications to relatively small proteins. Resolution is significantly improved in 3D spectra but recording uniformly sampled data is time‐prohibitive. Here we describe non‐uniformly sampled HNCO‐based pseudo‐4D CEST that provides excellent resolution in reasonable measurement times. Data analysis is done through fitting in the time domain, without the need of reconstructing the frequency dimensions, exploiting previously measured accurate peak positions in reference spectra. The methodology is demonstrated on several protein systems, including a nascent form of superoxide dismutase that is implicated in neurodegenerative disease.     

Profiling and analysis of multiple compounds in rhubarb decoction after processing by wine steaming using UHPLC–Q‐TOF‐MS coupled with multiple statistical strategies

Rhubarb is one of the most popular traditional Chinese medicines and has been used for thousands of years in many Asian countries. Prepared rhubarb is obtained by steaming raw rhubarb with glutinous rice wine until it turned black in appearance both inside and outside. After processing, the therapeutic effects of prepared rhubarb change a lot. To find out the exact compound changes of the chemical profile in a decoction of rhubarb after processing and to clarify the material basis of the changed therapeutic effects, an ultra‐high performance liquid chromatography with quadrupole time‐of‐flight mass spectrometry method coupled with automated data analysis software and statistical strategy was developed. As a result, 63 peaks in raw rhubarb and 54 peaks in prepared rhubarb were detected, and a total of 45 chemical compounds were identified. The analysis data were subjected to a principle component analysis and a t‐test. Based on the results, 16 peaks were found to be the main contributors to the significant difference (p < 0.05) between raw and prepared rhubarb. Compared with raw rhubarb, the content of 15 components in prepared rhubarb was lower, while only rhein (1,8‐dihydroxy‐3‐carboxy anthraquinone) showed a higher intensity.     

Detection of the reduced forms of radical adducts on the ESR trace using HPLC‐electrochemical detector‐ultraviolet absorption detector‐electron spin resonance‐MS

To detect and identify the electron spin resonance (ESR) silent forms of the α‐(4‐pyridyl‐1‐oxide)‐N‐tert‐butylnitrone (4‐POBN) radical adducts, an electrochemical detector (ECD) was employed as a reactor in the HPLC‐ECD‐UV absorption detector‐ESR‐MS (HPLC‐ECD‐UV‐ESR‐MS). The ECD was employed to regenerate the radical forms from the reduced forms. The reduced forms of the 4‐POBN/pentyl radical adducts were analyzed using the HPLC‐ECD‐UV‐ESR‐MS. On addition of the ECD applied potential of +0.3 V, a peak appeared on the ESR trace of the HPLC‐ECD‐UV‐ESR‐MS analyses, indicating that the radical forms are regenerated from the reduced forms. The HPLC‐ECD‐UV‐ESR‐MS analyses were also performed for the reaction mixtures of phenylhydrazine with CuCl₂. Two peaks (peaks I and II) were detected on the UV trace (300 nm) of the HPLC‐ECD‐UV‐ESR‐MS. The mass spectra showed that the peak I and peak II compounds are radical and reduced forms of the 4‐POBN/phenyl radical adducts under the ECD applied potential of 0.0 V. Peak I was only detected on the ESR trace under the ECD applied potential of 0.0 V. In addition to peak I, peak II appeared on the ESR trace under the ECD applied potential of +0.3 V, indicating that the reduced forms are oxidized to the corresponding radical forms.     

Evaluation of peak overlap in migration-time distributions determined by organelle capillary electrophoresis: Type-II error analogy based on statistical-overlap theory

Organelles commonly are separated by capillary electrophoresis (CE) with laser-induced-fluorescence detection. Usually, it is assumed that peaks observed in the CE originate from single organelles, with negligible occurrence of peak overlap. Under this assumption, migration-time and mobility distributions are obtained by partitioning the CE into different regions and counting the number of observed peaks in each region. In this paper, criteria based on statistical-overlap theory (SOT) are developed to test the assumption of negligible peak overlap and to predict conditions for its validity. For regions of the CE having constant peak density, the numbers of peaks (i.e., intensity profiles of single organelles) and observed peaks (i.e., maxima) are modeled by probability distributions. For minor peak overlap, the distributions partially merge, and their mergence is described by an analogy to the Type-II error of hypothesis testing. Criteria are developed for the amount of peak overlap, at which the number of observed peaks has an 85% or 90% probability of lying within the 95% confidence interval of the number of peaks of single organelles. For this or smaller amounts of peak overlap, the number of observed peaks is a good approximation to the number of peaks. A simple procedure is developed for evaluating peak overlap, requiring determination of only the peak standard deviation, the duration of the region occupied by peaks, and the number of observed peaks in the region. The procedure can be applied independently to each region of the partitioned CE. The procedure is applied to a mitochondrial CE.     

New algorithms for processing and peak detection in liquid chromatography/mass spectrometry data

Two new algorithms for automated processing of liquid chromatography/mass spectrometry (LC/MS) data are presented. These algorithms were developed from an analysis of the noise and artifact distribution in such data. The noise distribution was analyzed by preparing histograms of the signal intensity in LC/MS data. These histograms are well fit by a sum of two normal distributions in the log scale. One new algorithm, median filtering, provides increased performance compared to averaging adjacent scans in removing noise that is not normally distributed in the linear scale. Another new algorithm, vectorized peak detection, provides increased robustness with respect to variation in the noise and artifact distribution compared to methods based on determining an intensity threshold for the entire dataset. Vectorized peak detection also permits the incorporation of existing algorithms for peak detection in ion chromatograms and/or mass spectra. The application of these methods to LC/MS spectra of complex biological samples is described.     

An automatic peak finding method for LC‐MS data using Gaussian second derivative filtering

A highly automated procedure for localising and characterising peaks in the chromatographic time domain of LC‐MS data has been developed. The work was initiated by an identified need to facilitate the detection and tracking of chromatographic peaks during method development for the analysis of impurities in pharmaceutical products. The algorithm is mainly based on a digital filter for which the settings are automatically adapted to the data set under study. The procedure was evaluated for synthetic data sets with various S/N levels, peak widths and baseline properties. It was found that even for the worst case tested with S/N=10 and a high variability in the baseline, 94% of the simulated analytical peaks could be detected without producing any false‐positive identifications. Furthermore, the number of correctly estimated peak heights and peak widths falling within a 10% error of the true values were 94 and 91%, respectively. For experimental data sets, peak height, and width estimations were more difficult, but the processed reconstructions showed an excellent agreement with the analytical signals of the raw data, and also a clearly improved visualisation in total ion‐ and base‐peak chromatograms.