Tapping the Full Potential of Infrared Spectroscopy for the Analysis of Technical Lignins.

We present an approach to overcome the challenges associated with the increasing demand of high-throughput characterization of technical lignins, a key resource in emerging bioeconomies. Our approach offers a resort from the lack of direct, simple, and low-cost analytical techniques for lignin characterization by employing multivariate calibration models based on infrared (IR) spectroscopy to predict structural properties of lignins (i. e., functionality, molar mass). By leveraging a comprehensive database of over 500 well-characterized technical lignin samples - a factor of 10 larger than previously used sets - our chemometric models achieved high levels of quality and statistical confidence for the determination of different functional group contents (RMSEPs of 4-16 %). However, the statistical moments of the molar mass distribution are still best determined by size-exclusion chromatography. Analyses of over 500 technical lignins offered also a great opportunity to provide information on the general variability in kraft lignins and lignosulfonates (from different origins). Overall, the effected savings in analysis time (>7 h), resources, and required sample mass combined with non-destructiveness of the measurement satisfy key demands for efficient high-throughput lignin analyses. Finally, we discuss the advantages, disadvantages, and limitations of our approach, along with critical insights into the associated chemical-analytical and spectroscopic challenges.

Chemometrics in analytical chemistry-part II: modeling, validation, and applications.

The contribution of chemometrics to important stages throughout the entire analytical process such as experimental design, sampling, and explorative data analysis, including data pretreatment and fusion, was described in the first part of the tutorial "Chemometrics in analytical chemistry." This is the second part of a tutorial article on chemometrics which is devoted to the supervised modeling of multivariate chemical data, i.e., to the building of calibration and discrimination models, their quantitative validation, and their successful applications in different scientific fields. This tutorial provides an overview of the popularity of chemometrics in analytical chemistry.

Chemometrics in analytical chemistry-part I: history, experimental design and data analysis tools.

Chemometrics has achieved major recognition and progress in the analytical chemistry field. In the first part of this tutorial, major achievements and contributions of chemometrics to some of the more important stages of the analytical process, like experimental design, sampling, and data analysis (including data pretreatment and fusion), are summarised. The tutorial is intended to give a general updated overview of the chemometrics field to further contribute to its dissemination and promotion in analytical chemistry.

Prediction of liquid chromatographic retention behavior based on quantum chemical parameters using supervised self organizing maps.

Self organizing maps (SOMs) in a supervised mode were applied for prediction of liquid chromatographic retention behavior of chemical compounds based on their quantum chemical information. The proposed algorithm was simple and required only a small alteration of the standard SOM algorithm. The application was illustrated by the prediction of the retention indices of bifunctionally substituted N-benzylideneanilines (NBA) and the prediction of the retention factors of some pesticides. Although the predictive ability of the supervised SOM could not be significantly greater than that of some previously established neural network methods, such as a radial basis function (RBF) neural network and a back-propagation artificial neural network (ANN), the main advantage of the proposed method was the ability to reveal non-linear structure of the model. The complex relationships between samples could be visualized using U-matrix and the influence of each variable on the predictive model could be investigated using component planes-which can provide chemical insight.

PIXE analysis of PM2.5 and PM(2.5-10) for air quality assessment of Islamabad, Pakistan: application of chemometrics for source identification.

A Gent sampler was used to collect 379 pairs of filters from Nilore, a suburban area of Islamabad city. The study was designed to assess the concentration variations of trace elements in fine and coarse particulate matter due to anthropogenic activities and naturally occurring events. Source identification was performed by applying MATLAB software for principal component analysis (PCA), and cluster analysis (CA). The average fine and coarse particulate masses during the study period were 15.1 ± 11.9 and 37.3 ± 28.0 µg/m(3) respectively which complies with the 24-h air quality limits set by the government of Pakistan. The application of PCA to PM(2.5) data suggests the PM contribution from sources such as soil, automobile exhaust and coal combustion, road dust and wearing of tyres, wood combustion, biomass burning and fertilizers and fungicides whereas for the PM(2.5-10) data shows signatures of suspended soil, automobile exhaust, road dust and wearing of tyres, wood and biomass burning, refuse incineration, Ni smelter, fertilizers and fungicides are obtained. Cluster analysis of PM(2.5) and PM(2.5-10) datasets reveals that there are mainly three contributory pollution sources and these are suspended soil particles, automobile related sources and wood and coal combustion.

Self organising maps for visualising and modelling.

The paper describes the motivation of SOMs (Self Organising Maps) and how they are generally more accessible due to the wider available modern, more powerful, cost-effective computers. Their advantages compared to Principal Components Analysis and Partial Least Squares are discussed. These allow application to non-linear data, are not so dependent on least squares solutions, normality of errors and less influenced by outliers. In addition there are a wide variety of intuitive methods for visualisation that allow full use of the map space. Modern problems in analytical chemistry include applications to cultural heritage studies, environmental, metabolomic and biological problems result in complex datasets. Methods for visualising maps are described including best matching units, hit histograms, unified distance matrices and component planes. Supervised SOMs for classification including multifactor data and variable selection are discussed as is their use in Quality Control. The paper is illustrated using four case studies, namely the Near Infrared of food, the thermal analysis of polymers, metabolomic analysis of saliva using NMR, and on-line HPLC for pharmaceutical process monitoring.

Automated peak detection and matching algorithm for gas chromatography-differential mobility spectrometry.

A gas chromatography-differential mobility spectrometer (GC-DMS) involves a portable and selective mass analyzer that may be applied to chemical detection in the field. Existing approaches examine whole profiles and do not attempt to resolve peaks. A new approach for peak detection in the 2D GC-DMS chromatograms is reported. This method is demonstrated on three case studies: a simulated case study; a case study of headspace gas analysis of Mycobacterium tuberculosis (MTb) cultures consisting of three matching GC-DMS and GC-MS chromatograms; a case study consisting of 41 GC-DMS chromatograms of headspace gas analysis of MTb culture and media.

Self-Organizing Maps and Support Vector Regression as aids to coupled chromatography: illustrated by predicting spoilage in apples using volatile organic compounds.

The paper describes the application of SOMs (Self-Organizing Maps) and SVR (Support Vector Regression) to pattern recognition in GC-MS (gas chromatography-mass spectrometry). The data are applied to two groups of apples, one which is a control and one which has been inoculated with Penicillium expansum and which becomes spoiled over the 10-day period of the experiment. GC-MS of SPME (solid phase microextraction) samples of volatiles from these apples were recorded, on replicate samples, over time, to give 58 samples used for pattern recognition and a peak table obtained. A new approach for finding the optimum SVR parameters called differential evolution is described. SOMs are presented in the form of two-dimensional maps. This paper shows the potential of using machine learning methods for pattern recognition in analytical chemistry, particularly as applied to food chemistry and biology where trends are likely to be non-linear.

Analysis of volatile organic compounds in human saliva by a static sorptive extraction method and gas chromatography-mass spectrometry.

Human saliva not only helps control oral health (with anti-microbial proteins), but it may also play a role in chemical communication. As is the case with other mammalian species, human saliva contains peptides, proteins, and numerous volatile organic compounds (VOCs). A high-throughput analytical method is described for profiling a large number of saliva samples to screen the profiles of VOCs. Saliva samples were collected in a non-stimulated fashion. The method utilized static stir bar extraction followed by gas chromatography-mass spectrometry (GC-MS). The method provided excellent reproducibility for a wide range of salivary compounds, including alcohols, aldehydes, ketones, carboxylic acids, esters, amines, amides, lactones, and hydrocarbons. Furthermore, substantial overlap of salivary VOCs and the previously reported skin VOCs in the same subject group was found in this study by using pattern recognition analyses. Sensitivity, precision, and reproducibility of the method suggest that this technique has potential in physiological, metabolomic, pharmacokinetic, forensic, and toxicological studies of small organic compounds where a large number of human saliva samples are involved.

Self-organizing map quality control index.

A new approach for process monitoring is described, the self-organizing map quality control (SOMQC) index. The basis of the method is that SOM maps are formed from normal operating condition (NOC) samples, using a leave-one-out approach. The distances (or dissimilarities) of the left out sample can be determined to all the units in the map, and the nth percentile measured distance of the left out sample is used to provide a null distribution of NOC distances which is generated using the Hodges-Lehmann method. The nth percentile distance of a test sample to a map generated from all NOC samples can be measured and compared to the null distribution at a given confidence level to determine whether the sample can be judged as out of control. The approach described in this paper is applied to online high-performance liquid chromatography (HPLC) measurements of a continuous pharmaceutical process and is compared to other established methods including Q and D statistics and support vector domain description. The SOMQC has advantages in that there is no requirement for multinormality in the NOC samples, or for linear models, or to perform principal components analysis (PCA) prior to the analysis with concomitant issues about choosing the number of PCs. It also provides information about which variables are important using component planes. The influence of extreme values in the background data set can also be tuned by choosing the distance percentile.

Mouse urinary biomarkers provide signatures of maturation, diet, stress level, and diurnal rhythm.

Body fluids such as urine potentially contain a wealth of information pertaining to age, sex, social and reproductive status, physiologic state, and genotype of the donor. To explore whether urine could encode information regarding environment, physiology, and development, we compared the volatile compositions of mouse urine using solid-phase microextraction and gas chromatography-mass spectrometry (SPME-GC/MS). Specifically, we identified volatile organic compounds (VOCs) in individual urine samples taken from inbred C57BL/6J-H-2(b) mice under several experimental conditions-maturation state, diet, stress, and diurnal rhythms, designed to mimic natural variations. Approximately 1000 peaks (i.e., variables) were identified per comparison and of these many were identified as potential differential biomarkers. Consistent with previous findings, we found groups of compounds that vary significantly and consistently rather than a single unique compound to provide a robust signature. We identified over 49 new predictive compounds, in addition to identifying several published compounds, for maturation state, diet, stress, and time-of-day. We found a considerable degree of overlap in the chemicals identified as (potential) biomarkers for each comparison. Chemometric methods indicate that the strong group-related patterns in VOCs provide sufficient information to identify several parameters of natural variations in this strain of mice including their maturation state, stress level, and diet.

Support vector machines for classification and regression.

The increasing interest in Support Vector Machines (SVMs) over the past 15 years is described. Methods are illustrated using simulated case studies, and 4 experimental case studies, namely mass spectrometry for studying pollution, near infrared analysis of food, thermal analysis of polymers and UV/visible spectroscopy of polyaromatic hydrocarbons. The basis of SVMs as two-class classifiers is shown with extensive visualisation, including learning machines, kernels and penalty functions. The influence of the penalty error and radial basis function radius on the model is illustrated. Multiclass implementations including one vs. all, one vs. one, fuzzy rules and Directed Acyclic Graph (DAG) trees are described. One-class Support Vector Domain Description (SVDD) is described and contrasted to conventional two- or multi-class classifiers. The use of Support Vector Regression (SVR) is illustrated including its application to multivariate calibration, and why it is useful when there are outliers and non-linearities.

Supervised self organizing maps for classification and determination of potentially discriminatory variables: illustrated by application to nuclear magnetic resonance metabolomic profiling.

The article describes the extension of the self organizing maps discrimination index (SOMDI) for cases where there are more than two classes and more than one factor that may influence the group of samples by using supervised SOMs to determine which variables and how many are responsible for the different types of separation. The methods are illustrated by an application in the area of metabolic profiling, consisting of a nuclear magnetic resonance (NMR) data set of 96 samples of human saliva, which is characterized by three factors, namely, whether the sample has been treated or not, 16 donors, and 3 sampling days, differing for each donor. The sampling days can be considered a null factor as they should have no significant influence on the metabolic profile. Methods for supervised SOMs involve including a classifier for organizing the map, and we report a method for optimizing this by using an additional weight that determines the relative importance of the classifier relative to the overall experimental data set in order to avoid overfitting. Supervised SOMs can be obtained for each of the three factors, and we develop a multiclass SOM discrimination index (SOMDI) to determine which variables (or regions of the NMR spectra) are considered significant for each of the three potential factors. By dividing the data iteratively into training and test sets 100 times, we define variables as significant for a given factor if they have a positive SOMDI in the training set for the factor and class of interest over all iterations.

Pattern recognition of inductively coupled plasma atomic emission spectroscopy of human scalp hair for discriminating between healthy and hepatitis C patients.

Inductively Coupled Plasma Atomic Emission Spectroscopy measurements of six trace elements were performed on the scalp hair of 155 donors, 73 of which have been diagnosed with Hepatitis C and 82 Controls. Principal Components Analysis (PCA) was employed to visualise the separation between groups and show the relationship between the elements and the diseased state. Pattern recognition methods for classification involving Quadratic Discriminant Analysis and Partial Least Squares Discriminant Analysis (PLS-DA) were applied to the data. The number of significant components for both PCA and PLS were determined using the bootstrap. The stability of training set models were determined by repeatedly splitting the data into training and test sets and employing visualisation for two components models: the percent classification ability (CC), predictive ability (PA) and model stability (MS) were computed for test and training sets.

Variable selection using iterative reformulation of training set models for discrimination of samples: application to gas chromatography/mass spectrometry of mouse urinary metabolites.

The paper discusses variable selection as used in large metabolomic studies, exemplified by mouse urinary gas chromatography of 441 mice in three experiments to detect the influence of age, diet, and stress on their chemosignal. Partial least squares discriminant analysis (PLS-DA) was applied to obtain class models, using a procedure of 20,000 iterations including the bootstrap for model optimization and random splits into test and training sets for validation. Variables are selected using PLS regression coefficients on the training set using an optimized number of components obtained from the bootstrap. The variables are ranked in order of significance, and the overall optimal variables are selected as those that appear as highly significant over 100 different test and training set splits. Cost/benefit analysis of performing the model on a reduced number of variables is also illustrated. This paper provides a strategy for properly validated methods for determining which variables are most significant for discriminating between two groups in large metabolomic data sets avoiding the common pitfall of overfitting if variables are selected on a combined training and test set and also taking into account that different variables may be selected each time the samples are split into training and test sets using iterative procedures.

Consensus multivariate methods in gas chromatography mass spectrometry and denaturing gradient gel electrophoresis: MHC-congenic and other strains of mice can be classified according to the profiles of volatiles and microflora in their scent-marks.

House mice (Mus domesticus) communicate using scent-marks, and the chemical and microbial composition of these 'extended phenotypes' are both influenced by genetics. This study examined how the genes of the major histocompatibility complex (MHC) and background genes influence the volatile compounds (analysed with Gas Chromatography Mass Spectrometry or GC/MS) and microbial communities (analysed using Denaturating Gradient Gel Electrophoresis or DGGE) in scent-marks produced by congenic strains of mice. The use of Consensus Principal Components Analysis is described and shows relationships between the two types of fingerprints (GC/MS and DGGE profiles). Classification methods including Support Vector Machines and Discriminant Partial Least Squares suggest that mice can be classified according to both background strain and MHC-haplotype. As expected, the differences among the mice were much greater between strains that vary at both MHC and background loci than the congenics, which differ only at the MHC. These results indicate that the volatiles in scent-marks provide information about genetic similarity of the mice, and support the idea that the production of these genetically determined volatiles is influenced by commensal microflora. This paper describes the application of consensus methods to relate two blocks of analytical data.

Multilevel simultaneous component analysis for fault detection in multicampaign process monitoring: application to on-line high performance liquid chromatography of a continuous process.

A continuous process is monitored by on-line HPLC in four separate campaigns, ranging in duration from 10 to 104 h. Methods are reported that allow the study of variation over all four campaigns using Multilevel Simultaneous Components Analysis, which separates out the within- and between-campaign variation. In order to obtain control charts, Q- and D-statistics are combined with a within-campaign submodel (Simultaneous Components Analysis) to obtain a single model that is based only on within-campaign variation over all four campaigns.

Dynamic analysis of on-line high-performance liquid chromatography for multivariate statistical process control.

A continuous process was studied over 83.32 h using on-line high-performance liquid chromatography, involving the acquisition of 252 chromatograms. A method for analysis of these data using multivariate statistical process control on peak tables, in real-time, is described. The normal operating condition (NOC) region of the process was identified using evolving principal components analysis to be between 5.77 and 8.13 h. 19 out of the 37 peaks detected throughout the process were found in the NOC region, the remainder representing undesirable contaminants found elsewhere in the process. A major challenge is to develop the peak table as the process evolves, which is dynamically updated as new peaks are detected after the NOC region: this approach involving an "unlocked" peak table is contrasted to an approach using a "locked" peak table where only peaks detected during the NOC region are included in the model. In addition, results are compared to those obtained using baseline corrected and aligned chromatograms, using a NOC region of 5.85-8.33h. D- and Q-charts were obtained. It is shown that the "unlocked" peak table detects out of control samples best and provides good diagnostic insight into problems with the process.

Self Organising Maps for distinguishing polymer groups using thermal response curves obtained by dynamic mechanical analysis.

Self Organising Maps are described including the U-Matrix, component planes, hit histograms, quality indicators as mean quantisation error and topological error. Software was written in Matlab and several new approaches for visualising multiclass maps are employed. The method is applied to a dataset consisting of the Dynamic Mechanical Analysis of 293 polymers, involving heating the polymers over a temperature range of -51 degrees C to 270 degrees C. These can be characterised in three different ways (a) amorphous or semi-crystalline (b) as 9 groups (c) as 30 grades.