支持向量机结台主成分分析辅助激光诱导击穿光谱技术识别鲜肉品种

为提高激光诱导击穿光谱技术(Laser-induced breakdown spectroscopy,LIBS)对鲜肉品种的识别率,采用支持向量机结合主成分分析算法辅助LIBS技术对鲜肉品种进行识别.对鲜肉切片用载玻片压平,采用LIBS技术对鲜肉组织(猪肉、牛肉和鸡肉)表面进行光谱数据的采集,每种鲜肉采集150幅光谱并进行随机排列,取前75幅光谱作为训练集建立模型,后75幅作为测试集测试建模结果.研究选取K、Ca、Na、Mg、Al、H、O等元素的49条归一化谱线数据进行主成分分析,并用所得数据建立支持向量机分类模型.结果表明,通过主成分分析降维,输入变量从49个优化减少到18个,模型建模速度从88.91 s降至55.52 s,提高了支持向量机的建模效率;并使预测集的平均识别率提高到89.11%.本研究为激光诱导击穿光谱技术在鲜肉品种快速分类领域提供了方法和数据参考.     

交叉响应传感器鉴别β受体激动剂及其混合体系

开发了一种鉴别β受体激动剂的新型阵列传感器。该传感器由8种传感物质构成,使用96孔板酶标仪采集响应数据,结合主成分分析(PCA)、分层聚类分析(HCA)、判别分析(LDA)等模式识别方法进行数据处理,对5类β受体激动剂及其混合物进行检测。PCA结果表明,该传感器主要是基于空间结构以及氢键作用实现对β受体激动剂的识别;HCA结果显示,93个分析样本归类正确;LDA结果显示,该传感器对于β受体激动剂识别的准确率达98.9%。本方法在β受体激动剂的检测中有潜在应用价值。     

Multivariate design

In multivariate data analysis such as principal components analysis (PCA) and projections to latent structures (PLS), it is essential that the training set systems (objects) are selected to provide data with substantial information for model parametrization, and to represent properly any future situations where the multilvariate model is used for predictions. In the framework of multivariate projections (PCA, SIMCA and PLS), elementary concepts of statistical design (fractional factorials and composite designs) can be used with the latent variables (PC or PLS scores) as design variables. The plan of action thus becomes: (1) problem formulation (specify aim and model, make a conceptual division of the investigated system into subsystems); (2) collection of multivariate data for each type of subsystems; (3) estimation of the practical dimensionality of the data for each type of subsystems by PC or PLS analysis; (4) use of the PC or PLS scores (t) as design variables in the combination of subsystems to systems in the training set; (5) measurement of responses (Y); (6) analysis of data by PCA or PLS; (7) interpretation of results with possible feedback to steps 1, 2 or 3. The procedures are illustrated by two problems: a structure/activity relationship for a family of peptides, and optimization of an organic synthesis with respect to system variables (solvent, substrate, co-reactant_) and process variables (temperature, reactant concentrations).     

Multivariate monitoring of soybean oil ethanolysis by FTIR

In this work, an analytical procedure was developed to monitor the ethanolysis of degummed soybean oil (DSO) using Fourier-transformed mid-infrared spectroscopy (FTIR) and methods of multivariate analysis such as principal component analysis (PCA) and partial least squares regression (PLS). The triglycerides (reagents) and ethyl esters (products) involved in ethanolysis were shown to have similar FTIR spectra. However, when the FTIR spectra derived from seven standard mixtures of triolein and ethyl oleate were treated by PCA at the region that represents the CO stretching vibration of ester groups (1700-1800 cm⁻¹), only two principal components (PC) were shown to capture 99.95% of the total spectral variance (92.37% for the former and 7.58% for the latter PC). This observation supported the development of a multivariate calibration model that was based on the PLS regression of the FTIR data. The prevision capability of this model was measured against 40 reaction aliquots whose ester content was previously determined by size exclusion chromatography. Only small discrepancies were observed when the two experimental data sets were treated by linear regression (R²=0.9837) and these deviations were attributed to the occurrence of non-modeled transient species in the reaction mixture (reaction intermediates), particularly at short reaction times. Therefore, the FTIR/PLS model was shown to be a fast and accurate method to predict reaction yields and to follow the in situ kinetics of soybean oil ethanolysis.     

Chemometrics‐enhanced laser‐induced thermal emission detection of PETN and other explosives on various substrates

Infrared emissions (IREs) of samples of pentaerythritol tetranitrate (PETN) deposited as contamination residues on various substrates were measured to generate models for the detection and discrimination of the important nitrate ester from the emissions of the substrates. Mid‐infrared emissions were generated by heating the samples remotely using laser‐induced thermal emission (LITE). Chemometrics multivariate analysis techniques such as principal component analysis (PCA), soft independent modeling by class analogy (SIMCA), partial least squares‐discriminant analysis (PLS‐DA), support vector machines (SVMs), and neural network (NN) were employed to generate the models for the classification and discrimination of PETN IREs from substrate thermal emissions. PCA exhibited less variability for the LITE spectra of PETN/substrates. SIMCA was able to predict only 44.7% of all samples, while SVM proved to be the most effective statistical analysis routine, with a discrimination performance of 95%. PLS‐DA and NN achieved prediction accuracies of 94% and 88%, respectively. High sensitivity and specificity values were achieved for five of the seven substrates investigated. Copyright © 2015 John Wiley & Sons, Ltd.     

OPLS in batch monitoring – Opens up new opportunities

In batch statistical process control (BSPC), data from a number of “good” batches are used to model the evolution (trajectory) of the process and they also define model control limits, against which new batches may be compared. The benchmark methods used in BSPC include partial least squares (PLS) and principal component analysis (PCA).     

Spectrophotometric determination of metal ions in electroplating solutions in the presence of EDTA with the aid of multivariate calibration and artificial neural networks

Metal ions such as Co(II), Ni(II), Cu(II), Fe(III) and Cr(III), which are commonly present in electroplating baths at high concentrations, were analysed simultaneously by a spectrophotometric method modified by the inclusion of the ethylenediaminetetraacetate (EDTA) solution as a chromogenic reagent. The prediction of the metal ion concentrations was facilitated by the use of an orthogonal array design to build a calibration data set consisting of absorption spectra collected in the 370-760 nm range from solution mixtures containing the five metal ions earlier. With the aid of this data set, calibration models were built based on 10 different chemometrics methods such as classical least squares (CLS), principal component regression (PCR), partial least squares (PLS), artificial neural networks (ANN) and others. These were tested with the use of a validation data set constructed from synthetic solutions of the five metal ions. The analytical performance of these chemometrics methods were characterized by relative prediction errors and recoveries (%). On the basis of these results, the computational methods were ranked according to their performances using the multi-criteria decision making procedures preference ranking organization method for enrichment evaluation (PROMETHEE) and geometrical analysis for interactive aid (GAIA). PLS and PCR models applied to the spectral data matrix that used the first derivative pre-treatment were the preferred methods. They together with ANN-radial basis function (RBF) and PLS were applied for analysis of results from some typical industrial samples analysed by the EDTA-spectrophotometric method described. DPLS, DPCR and the ANN-RBF chemometrics methods performed particularly well especially when compared with some target values provided by industry.     

Direct detection of trimethylamine in meat food products using ion mobility spectrometry

Biogenic amines are degradation products generated by bacteria in meat products. These amines can indicate bacterial contamination or have a carcinogenic effect to humans consuming spoiled meats; therefore, their rapid detection is essential. Trimethylamine (TMA) is a good target for the detection of biogenic amines because its volatility. TMA was directly detected in meat food products using ion mobility spectrometry (IMS). TMA concentrations were measured in chicken meat juice for a quantitative evaluation of the meat decaying process. The lowest detected TMA concentration in chicken juice was 0.6 ± 0.2 ng and the lowest detected signal for TMA in a standard aqueous solution was 0.6 ng. IMS data were processed using partial least squares (PLS) and Fuzzy rule-building expert system (FuRES). Using these two chemometric methods, trimethylamine concentrations of different days of meat spoilage can be separated, indicating the decaying of meat products. Comparing the two methods, FuRES provided a better classification of different days of meat spoilage.     

An optimization‐based undeflated PLS (OUPLS) method to handle missing data in the training set

Advances in sensory systems have led to many industrial applications with large amounts of highly correlated data, particularly in chemical and pharmaceutical processes. With these correlated data sets, it becomes important to consider advanced modeling approaches built to deal with correlated inputs in order to understand the underlying sources of variability and how this variability will affect the final quality of the product. Additional to the correlated nature of the data sets, it is also common to find missing elements and noise in these data matrices. Latent variable regression methods such as partial least squares or projection to latent structures (PLS) have gained much attention in industry for their ability to handle ill‐conditioned matrices with missing elements. This feature of the PLS method is accomplished through the nonlinear iterative PLS (NIPALS) algorithm, with a simple modification to consider the missing data. Moreover, in expectation maximization PLS (EM‐PLS), imputed values are provided for missing data elements as initial estimates, conventional PLS is then applied to update these elements, and the process iterates to convergence. This study is the extension of previous work for principal component analysis (PCA), where we introduced nonlinear programming (NLP) as a means to estimate the parameters of the PCA model. Here, we focus on the parameters of a PLS model. As an alternative to modified NIPALS and EM‐PLS, this paper presents an efficient NLP‐based technique to find model parameters for PLS, where the desired properties of the parameters can be explicitly posed as constraints in the optimization problem of the proposed algorithm. We also present a number of simulation studies, where we compare effectiveness of the proposed algorithm with competing algorithms. Copyright © 2014 John Wiley & Sons, Ltd.     

Quick monitoring of coloring agents in highly consumed candies using multivariate calibration

Multivariate calibration are gaining popularity in assaying food matrices. Partial least squares is a powerful multivariate calibration method that used to build a quantitative relationship between measured variables and a property of interest (i.e., concentration) of the system under study. Partial least squares PLS calibration along with UV/vis spectral data was efficient to account for indirect food matrix and direct interference effects resulted from overlapping food dyes. PLS was able to quantify tartrazine TAT, allura red AR, sunset yellow SY and brilliant black BB that added to wide selection sugar-based candies. The results indicated that 70% of samples containing single dye while 8% containing TAT-SY mix and certain samples containing TAT + SY + AR + BB. Lollypops were found to contain high levels of AR (77–120 mg/kg) and TAT (56–166 mg/kg). The maximum adulteration was 50% observed in lollypops. PLS calibration was workable to predict colorants with prediction errors of 7%. Using PLS, dyes were detected down to 0.1 mg/L with acceptable accuracy and precision. PLS showed comparable performance with liquid chromatography for dyes quantification and can substitute laborious chromatography for quick detection of coloring agents in candies.     

1H-NMR-Based Metabolomics: An Integrated Approach for the Detection of the Adulteration in Chicken,Chevon, Beef and Donkey Meat

Meat is a rich source of energy that provides high-value animal protein, fats, vitamins, minerals and trace amounts of carbohydrates. Globally, different types of meats are consumed to fulfill nutritional requirements. However, the increasing burden on the livestock industry has triggered the mixing of high-price meat species with low-quality/-price meat. This work aimed to differentiate different meat samples on the basis of metabolites. The metabolic difference between various meat samples was investigated through Nuclear Magnetic Resonance spectroscopy coupled with multivariate data analysis approaches like principal component analysis (PCA) and orthogonal partial least square-discriminant analysis (OPLS-DA). In total, 37 metabolites were identified in the gluteal muscle tissues of cow, goat, donkey and chicken using ¹H-NMR spectroscopy. PCA was found unable to completely differentiate between meat types, whereas OPLS-DA showed an apparent separation and successfully differentiated samples from all four types of meat. Lactate, creatine, choline, acetate, leucine, isoleucine, valine, formate, carnitine, glutamate, 3-hydroxybutyrate and α-mannose were found as the major discriminating metabolites between white (chicken) and red meat (chevon, beef and donkey). However, inosine, lactate, uracil, carnosine, format, pyruvate, carnitine, creatine and acetate were found responsible for differentiating chevon, beef and donkey meat. The relative quantification of differentiating metabolites was performed using one-way ANOVA and Tukey test. Our results showed that NMR-based metabolomics is a powerful tool for the identification of novel signatures (potential biomarkers) to characterize meats from different sources and could potentially be used for quality control purposes in order to differentiate different meat types.     

Metabolomics of meat exudate: Its potential to evaluate beef meat conservation and aging

In this study we analyzed the exudate of beef to evaluate its potential as non invasive sampling for nuclear magnetic resonance (NMR) based metabolomic analysis of meat samples. Exudate, as the natural juice from raw meat, is an easy to obtain matrix that it is usually collected in small amounts in commercial meat packages. Although meat exudate could provide complete and homogeneous metabolic information about the whole meat piece, this sample has been poorly studied. Exudates from 48 beef samples of different breeds, cattle and storage times have been studied by ¹H NMR spectroscopy. The liquid exudate spectra were compared with those obtained by High Resolution Magic Angle Spinning (HRMAS) of the original meat pieces. The close correlation found between both spectra (>95% of coincident peaks in both registers; Spearman correlation coefficient = 0.945) lead us to propose the exudate as an excellent alternative analytical matrix with a view to apply meat metabolomics. 60 metabolites could be identified through the analysis of mono and bidimensional exudate spectra, 23 of them for the first time in NMR meat studies. The application of chemometric tools to analyze exudate dataset has revealed significant metabolite variations associated with meat aging. Hence, NMR based metabolomics have made it possible both to classify meat samples according to their storage time through Principal Component Analysis (PCA), and to predict that storage time through Partial Least Squares (PLS) regression.     

Chemometric Analysis of Fatty Acid Composition of Raw Chicken,Beef, and Pork Meat with Plant Extract Addition during Refrigerated Storage

During the shelf-life, meat undergoes a number of processes that negatively affect the quality of the product, including fatty acid composition. The application of various plant extracts in meat could affect the changes of fatty acids during storage. Thus, the aim of this study was to investigate the effect of various spice and herb extracts on fatty acid composition in raw pork, beef, and chicken meat when stored at 4 °C for 13 days. Based on multivariate statistical analysis, two datasets were extracted from each type of meat. One dataset included samples with allspice, bay leaf, black seed, cardamom, caraway, clove, and nutmeg with the high share of total MUFA (monounsaturated fatty acids) in chicken and pork meat and high MUFA and PUFA (polyunsaturated fatty acids) contribution in beef meat after storage. The second dataset included basil, garlic, onion, oregano, rosemary, and thyme with high PUFA share in chicken and pork meat and high SFA (saturated fatty acids) contribution in beef meat. From the regression analysis, a significant effect of time on fatty acid composition in meat was reported. Generally, the rates of fatty acid changes were dependent on the plant extract incorporated into the meat. The most visible effect of plant extracts was obtained in chicken meat. In chicken meat with plant extracts, the rates of SFA and PUFA changes with time were slower compared to the control sample. In summary, the fatty acid composition of intramuscular fat varied during storage, and the addition of plant extracts significantly affected the rate of these changes, which was dependent on the meat matrix.     

Simultaneous determination of tin, germanium and molybdenum by diode array detection-flow injection analysis with partial least squares calibration model

Simultaneous determination of tin, germanium and molybdenum in food samples has been established by flow injection-charge coupled detector (CCD) diode array detection spectrophotometry with partial least squares (PLS) algorithm. The method was based on the chromogenic reaction of metal ions and salicylflurone in the presence of cetyltrimethyl ammonium bromide. The overlapping spectra of these complexes are collected by CCD diode array detector and the multi-wavelength absorbance data are processed using partial least squares algorithm. The reaction conditions and analytical parameters of flow injection analysis have been investigated. The method was applied to directly determine Ge, Mo and Sn in several food samples after digestion with satisfactory results. The recoveries of spiked samples were 80.0-102.0% for tin, 86.3-92.0% for germanium and 83.2-95.2% for molybdenum, and the relative standard deviations for samples were 4.4-7.8%. Molybdenum in certified reference material of cattle liver was determined by the proposed method (n=8). The differential values between determined and guarantee values were within the given uncertain value ranges (t=1.687, P>0.05 for t-test). The samples of mung bean, kelp and pork liver were analyzed by the proposed method and inductively couple plasma-atomic emission spectroscopy (ICP-AES) method. The determination results of the two methods are in good agreement. The sampling rate is 30 samples h⁻¹.     

Regression with multiple regressor arrays

Extension of standard regression to the case of multiple regressor arrays is given via the Kronecker product. The method is illustrated using ordinary least squares regression (OLS) as well as the latent variable (LV) methods principal component regression (PCR) and partial least squares regression (PLS). Denoting the method applied to PLS as mrPLS, the latter was shown to explain as much or more variance for the first LV relative to the comparable L‐partial least squares regression (L‐PLS) model. The same relationship holds when mrPLS is compared to PLS or n‐way partial least squares (N‐PLS) and the response array is 2‐way or 3‐way, respectively, where the regressor array corresponding to the first mode of the response array is 2‐way and the second mode regressor array is an identity matrix. In a comparison with N‐PLS using fragrance data, mrPLS proved superior in a validation sense when model selection was used. Though the focus is on 2‐way regressor arrays, the method can be applied to n‐way regressors via N‐PLS. Copyright © 2007 John Wiley & Sons, Ltd.     

FT-MIR and Raman spectroscopy coupled to multivariate analysis for the detection of clenbuterol in murine model

A fast and simple screening method for the determination of clenbuterol at the ppb level in a murine model was demonstrated by Mid Infrared (MIR) and Raman spectroscopy in conjunction with multivariate analysis. In order to build the calibration models to quantify clenbuterol in rat meat, mixtures of rat meat and clenbuterol were prepared in a range of 5-10,000 ppb. Partial Least Square (PLS) analysis was used to build the calibration model. The results shown that Mid Infrared and Raman spectroscopy were efficient, but Mid Infrared (R(2) = 0.966 and SEC = 0.27) were superior to Raman (R(2) = 0.914 and SEC = 1.167). The SIMCA model developed showed 100% classification rate of rat meat samples with or without clenbuterol. The results were confirmed with contaminated meat samples from animals treated with clenbuterol. Chemometric models represent an attractive option for meat quality screening without sample pretreatments which can identify veterinary medicinal products at the ppb level.     

Multi-way PLS modeling of structure-activity data by incorporating electrostatic and lipophilic potentials on molecular surface

We devised and elaborated a surface-based three-dimensional-quantitative structure-activity relationship (3D-QSAR) method, which had been proposed in the previous study. This approach can be applied to more general case where both the electrostatic and lipophilic potentials on molecular surface simultaneously change. The 3D coordinates of all sampling points on molecular surface are projected into a 2D map by Kohonen neural network (KNN). Each node in the map is coded by the associated molecular electrostatic potential (MEP) or molecular lipophilic potential (MLP) values. The electrostatic and lipophilic KNN maps are generated for each compound and the four-way array is constructed by collecting two KNN maps of all samples. The correlation between four-way array and biological activity is examined by four-way partial least-squares (PLS). For validation, the structure-activity data of estrogen receptor antagonists was investigated. The four-way PLS model gave the high statistics at calibration and validation stages. The coefficients of the four-way PLS model back-projected on molecular surface had a reasonable 3D distribution and it was nicely consistent with active site of the estrogen receptor which was recently made clear by X-ray crystallography.     

Multi-way PLS modeling of structure–activity data by incorporating electrostatic and lipophilic potentials on molecular surface

We devised and elaborated a surface-based three-dimensional-quantitative structure–activity relationship (3D-QSAR) method, which had been proposed in the previous study. This approach can be applied to more general case where both the electrostatic and lipophilic potentials on molecular surface simultaneously change. The 3D coordinates of all sampling points on molecular surface are projected into a 2D map by Kohonen neural network (KNN). Each node in the map is coded by the associated molecular electrostatic potential (MEP) or molecular lipophilic potential (MLP) values. The electrostatic and lipophilic KNN maps are generated for each compound and the four-way array is constructed by collecting two KNN maps of all samples. The correlation between four-way array and biological activity is examined by four-way partial least-squares (PLS). For validation, the structure–activity data of estrogen receptor antagonists was investigated. The four-way PLS model gave the high statistics at calibration and validation stages. The coefficients of the four-way PLS model back-projected on molecular surface had a reasonable 3D distribution and it was nicely consistent with active site of the estrogen receptor which was recently made clear by X-ray crystallography.