Search

A method for performing untargeted metabolomic analysis of human serum has been developed based on protein precipitation followed by Ultra Performance Liquid Chromatography and Time-of-Flight mass spectrometry (UPLC-TOF-MS). This method was specifically designed to fulfill the requirements of a long-term metabolomic study, spanning more than 3 years, and it was subsequently thoroughly evaluated for robustness and repeatability. We describe here the observed drift in instrumental performance over time and its improvement with adjustment of the length of analytical block. The optimal setup for our purpose was further validated against a set of serum samples from 30 healthy individuals. We also assessed the reproducibility of chromatographic columns with the same chemistry of stationary phase from the same manufacturer but from different production batches. The results have allowed the authors to prepare SOPs for "fit for purpose" long-term UPLC-MS metabolomic studies, such as are being employed in the HUSERMET project. This method allows the acquisition of data and subsequent comparison of data collected across many months or years.

Metabolomics seeks to measure potentially all the metabolites in a biological sample, and consequently, we need to develop and optimize methods to increase significantly the number of metabolites we can detect. We extended the closed-loop (iterative, automated) optimization system that we had previously developed for one-dimensional GC-TOF-MS (O'Hagan, S.; Dunn, W. B.; Brown, M.; Knowles, J. D.; Kell, D. B. Anal. Chem. 2005, 77, 290−303) to comprehensive two-dimensional (GC×GC) chromatography. The heuristic approach used was a multiobjective version of the efficient global optimization algorithm. In just 300 automated runs, we improved the number of metabolites observable relative to those in 1D GC by some 3-fold. The optimized conditions allowed for the detection of over 4000 raw peaks, of which some 1800 were considered to be real metabolite peaks and not impurities or peaks with a signal/noise ratio of less than 5. A variety of computational methods served to explain the basis for the improvement. This closed-loop optimization strategy is a generic and powerful approach for the optimization of any analytical instrumentation.

Direct injection electrospray ionization mass spectrometry (ESI-MS) without prior analyte separation was investigated for the analysis of whole cell suspensions of bacteria. Thirty-six strains of aerobic endospore-forming bacteria, consisting of six Bacillus species and one Brevibacillus species, were studied. ESI was performed in the positive ion mode on the bacterial suspensions. Several peaks in the range of 250−1500 m/z were observed to contribute to variations in the spectral information among the species. Application of cluster analysis to the spectral data showed that this ESI-MS technique was capable of discriminating strains of the species B. subtilis. This investigation demonstrates the feasibility of measuring liquid samples with minimal sample preparation that can be useful for discrimination at the subspecies level. A change in the cone potential in the electrospray ion source was found to influence the spectral information of representative strains of all of the seven species tested. This ha...

Thirty-six strains of aerobic endospore-forming bacteria confirmed by polyphasic taxonomic methods to belong to Bacillus amyloliquefaciens, Bacillus cereus, Bacillus licheniformis, Bacillus megaterium, Bacillus subtilis (including Bacillus niger and Bacillus globigii), Bacillus sphaericus, and Brevi laterosporus were grown axenically on nutrient agar, and vegetative and sporulated biomasses were analyzed by Curie-point pyrolysis mass spectrometry (PyMS) and diffuse reflectance-absorbance Fourier-transform infrared spectroscopy (FT-IR). Chemometric methods based on rule induction and genetic programming were used to determine the physiological state (vegetative cells or spores) correctly, and these methods produced mathematical rules which could be simply interpreted in biochemical terms. For PyMS it was found that m/z 105 was characteristic and is a pyridine ketonium ion (C6H3ON+) obtained from the pyrolysis of dipicolinic acid (pyridine-2,6-dicarboxylic acid; DPA), a substance found in spores but not in vegetative cells; this was confirmed using pyrolysis-gas chromatography/mass spectrometry. In addition, a pyridine ring vibration at 1447-1439 cm-1 from DPA was found to be highly characteristic of spores in FT-IR analysis. Thus, although the original data sets recorded hundreds of spectral variables from whole cells simultaneously, a simple biomarker can be used for the rapid and unequivocal detection of spores of these organisms.

A technique for the analysis of multivariate data by genetic programming (GP) is described, with particular reference to the quantitative analysis of orange juice adulteration data collected by pyrolysis mass spectrometry (PyMS). The dimensionality of the input space was reduced by ranking variables according to product moment correlation or mutual information with the outputs. The GP technique as described gives predictive errors equivalent to, if not better than, more widespread methods such as partial least squares and artificial neural networks but additionally can provide a means for easing the interpretation of the correlation between input and output variables. The described application demonstrates that by using the GP method for analyzing PyMS data the adulteration of orange juice with 10% sucrose solution can be quantified reliably over a 0-20% range with an RMS error in the estimate of ∼1%.

For pyrolysis mass spectrometry (PyMS) to be used for the routine identification of microorganisms, for quantifying determinands in biological and biotechnological systems, and in the production of useful mass spectral libraries, it is paramount that newly acquired spectra be compared to those previously collected. Neural network and other multivariate calibration models have been used to relate mass spectra to the biological features of interest. As commonly observed, however, mass spectral fingerprints showed a lack of long-term reproducibility, due to instrumental drift in the mass spectrometer; when identical materials were analyzed by PyMS at dates from 4 to 20 months apart, neural network models produced at earlier times could not be used to give accurate estimates of determinand concentrations or bacterial identities. Neural networks, however, can be used to correct for pyrolysis mass spectrometer instrumental drift itself, so that neural network or other multivariate calibration models created using previously collected data can be used to give accurate estimates of determinand concentration or the nature of bacteria (or, indeed, other materials) from newly acquired pyrolysis mass spectra. This approach is not limited solely to pyrolysis mass spectrometry but is generally applicable to any analytical tool which is prone to instrumental drift, such as IR, ESR, NMR and other spectroscopies, and gas and liquid chromatography, as well as other types of mass spectrometry.

Binary mixtures of the protein lysozyme with glycogen, of DNA or RNA in glycogen, and the tertiary mixture of cells of the bacteria Bacillus subtilh, Escherichia coli, and Staphylococcusaurecls were subjected to pyrolysis mass spectrometry. To analyze the pyrolysis mass spectra so as to obtain quantitative information representative of the complex components of the mixtures, partial least-squares regression (PLS), principal components regression (PCR), and fully interconnected feedforward artificial neural networks (ANNs) were studied. In the latter case, the weights were modified using the standard back-propagation algorithm, and the nodes used a sigmoidal squashing function. It was found that each of the methods could be used to provide calibration models which gave excellent predictions for the concentrations of determinands in samples on which they had not been trained. Neural networks were found to provide the most accurate predictions. We also report that scaling the individual nodes on the input layer of ANNs significantly decreased the time taken for the ANNs to learn. Removing masses of low intensity, which perhaps mainly contributed noise to the pyrolysis mass spectra, had little effect on the accuracy of the ANN predictions though could dramatically speed up the learning process (by more than 100-fold) and slightly improved the accuracy of PLS calibrations.

Variable selection enhances the understanding and interpretability of multivariate classification models. A new chemometric method based on the selection of the most important variables in discriminant partial least-squares (VS-DPLS) analysis is described. The suggested method is a simple extension of DPLS where a small number of elements in the weight vector w is retained for each factor. The optimal number of DPLS factors is determined by cross-validation. The new algorithm is applied to four different high-dimensional spectral data sets with excellent results. Spectral profiles from Fourier transform infrared spectroscopy and pyrolysis mass spectrometry are used. To investigate the uniqueness of the selected variables an iterative VS-DPLS procedure is performed. At each iteration, the previously found selected variables are removed to see if a new VS-DPLS classification model can be constructed using a different set of variables. In this manner, it is possible to determine regions rather than individual variables that are important for a successful classification.

A method for the preparation and GC-TOF-MS analysis of human serum samples has been developed and evaluated for application in long-term metabolomic studies. Serum samples were deproteinized using 3:1 methanol/serum, dried in a vacuum concentrator, and chemically derivatized in a two-stage process. Samples were analyzed by GC-TOF-MS with a 25 min analysis time. In addition, quality control (QC) samples were used to quantify process variability. Optimization of chemical derivatization was performed. Products were found to be stable for 30 h after derivatization. An assessment of within-day repeatability and within-week reproducibility demonstrates that excellent performance is observed with our developed method. Analyses were consistent over a 5 month period. Additional method testing, using spiked serum samples, showed the ability to define metabolite differences between samples from a population and samples spiked with metabolites standards. This methodology allows the continuous acquisition and application of data acquired over many months in long-term metabolomic studies, including the HUSERMET project (http://www.husermet.org/).

Metabolomics seeks to measure potentially all the metabolites in a biological sample, and consequently, we need to develop and optimize methods to increase significantly the number of metabolites we can detect. We extended the closed-loop (iterative, automated) optimization system that we had previously developed for one-dimensional GC-TOF-MS (O'Hagan, S.; Dunn, W. B.; Brown, M.; Knowles, J. D.; Kell, D. B. Anal. Chem. 2005, 77, 290-303) to comprehensive two-dimensional (GCxGC) chromatography. The heuristic approach used was a multiobjective version of the efficient global optimization algorithm. In just 300 automated runs, we improved the number of metabolites observable relative to those in 1D GC by some 3-fold. The optimized conditions allowed for the detection of over 4000 raw peaks, of which some 1800 were considered to be real metabolite peaks and not impurities or peaks with a signal/noise ratio of less than 5. A variety of computational methods served to explain the basis for the improvement. This closed-loop optimization strategy is a generic and powerful approach for the optimization of any analytical instrumentation.

The number of instrumental parameters controlling modern analytical apparatus can be substantial, and varying them systematically to optimize a particular chromatographic separation, for example, is out of the question because of the astronomical number of combinations that are possible (i.e., the "search space" is very large). However, heuristic methods, such as those based on evolutionary computing, can be used to explore such search spaces efficiently. We here describe the implementation of an entirely automated (closed-loop) strategy for doing this and apply it to the optimization of gas chromatographic separations of the metabolomes of human serum and of yeast fermentation broths. Without human intervention, the Robot Chromatographer system (i) initializes the settings on the instrument, (ii) controls the analytical run, (iii) extracts the variables defining the analytical performance (specifically the number of peaks, signal/noise ratio, and run time), (iv) chooses (via the PESA-II multiobjective genetic algorithm), and (v) programs the next series of instrumental settings, the whole continuing in an iterative cycle until suitable sets of optimal conditions have been established. Genetic programming was used to remove noise peaks and to establish the basis for the improvements observed. The system showed that the number of peaks observable depended enormously on the conditions used and served to increase them by as much as 3-fold (e.g., to over 950 in human serum) while in many cases maintaining or reducing the run time and preserving excellent signal/noise ratios. The evolutionary closed-loop machine learning strategy we describe is generic to any type of analytical optimization.

We studied the effects of electrospray mass spectrometric instrumental settings on the relative and absolute detection of individual proteins in a five-component mixture. Conditions that were effective for a given protein could be very poor for the others, and vice versa, such that to a good approximation it was possible to find conditions for selective detection of individual proteins in a complex mixture without prior analytical separation. Some of these could be rationalized on the basis of the known biophysical properties of the individual proteins. The ability to vary the conditions of a mass spectrometric detection method on-line provides an important degree of freedom for the selective detection, and hence discrimination, of individual proteins and peptides in complex mixtures and has implications in proteomics, in particular with respect to top-down strategies for proteomic characterizations.

Optimizing experimental conditions for the effective analysis of intact proteins by mass spectrometry is challenging, as many analytical factors influence the spectral quality, often in very different ways for different proteins and especially with complex protein mixtures. We show that genetic search methods are highly effective in this kind of optimization and that it was possible in 6 generations with a total of500 experiments out of some 10(14) to find good combinations of experimental variables (electrospray ionization mass spectral settings) that would not have been detected by optimizing each variable alone (i.e., the search space is epistatic). Moreover, by inspecting the evolution of the variables to be optimized using genetic programming, we discovered an important relationship between two of the mass spectrometer settings that accounts for much of this success. Specifically, the conditions that were evolved included very low values of skimmer 1 voltage (the sample cone) and a skimmer 2 voltage (extraction cone) above a threshold that would nevertheless minimize the potential difference between the sample and extraction skimmers. The discovery of this relationship demonstrates the hypothesis-generating ability of genetic search in optimization processes where the size of the search space means that little or no a priori knowledge of the optimal conditions is available.