Preliminary study on the application of visible–near infrared spectroscopy and chemometrics to classify Riesling wines from different countries

Visible (VIS) and near infrared (NIR) spectroscopy combined with chemometrics was used in an attempt to classify commercial Riesling wines from different countries (Australia, New Zealand, France and Germany). Commercial Riesling wines (n = 50) were scanned in the VIS and NIR regions (400–2500 nm) in a monochromator instrument, in transmission mode. Principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA) and stepwise linear discriminant analysis (SLDA) based on PCA scores were used to classify Riesling wines according to their country of origin. Full cross validation (leave-one-out) was used as the validation method when classification models were developed. PLS-DA models correctly classified 97.5%, 80% and 70.5% of the Australian, New Zealand and European (France and Germany) Riesling wines, respectively. SLDA calibration models correctly classified 86%, 67%, 67% and 87.5% of the Australian, New Zealand, French and German Riesling wines, respectively. These results demonstrated that the VIS and NIR spectra contain information that when used with chemometrics allow discrimination between wines from different countries. To further validate the ability of VIS–NIR to classify white wine samples, a larger sample set will be required.     

Can spectroscopy geographically classify Sauvignon Blanc wines from Australia and New Zealand?

The combination of UV, visible (Vis), near-infrared (NIR) and mid-infrared (MIR) spectroscopy with multivariate data analysis was explored as a tool to classify commercial Sauvignon Blanc (Vitis vinifera L., var. Sauvignon Blanc) wines from Australia and New Zealand. Wines (n = 64) were analysed in transmission using UV, Vis, NIR and MIR regions of the electromagnetic spectrum. Principal component analysis (PCA), soft independent modelling of class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA) were used to classify Sauvignon Blanc wines according to their geographical origin using full cross validation (leave-one-out) as a validation method. Overall PLS-DA models correctly classified 86% of the wines from New Zealand and 73%, 86% and 93% of the Australian wines using NIR, MIR and the concatenation of NIR and MIR, respectively. Misclassified Australian wines were those sourced from the Adelaide Hills of South Australia. These results demonstrate the potential of combining spectroscopy with chemometrics data analysis techniques as a rapid method to classify Sauvignon Blanc wines according to their geographical origin.     

Varietal Differentiation of Grape Juice Based on the Analysis of Near- and Mid-infrared Spectral Data

The aim of this study was to evaluate the usefulness of visible (VIS), near-infrared reflectance (NIR) and mid-infrared (MIR) spectroscopy combined with pattern recognition methods as tools to differentiate grape juice samples from commercial Australian Chardonnay (n = 121) and Riesling (n = 91) varieties. Principal component analysis (PCA), partial least squares discriminant analysis and linear discriminant analysis (LDA) were applied to classified grape juice samples according to variety based on both NIR and MIR spectra using full cross-validation (leave-one-out) as a validation method. Overall, LDA models correctly classify 86% and 80% of the grape juice samples according to variety using MIR and VIS-NIR, respectively. The results from this study demonstrated that spectral differences exist between the juice samples from different varietal origins and confirmed that the infrared (IR) spectrum contains information able to discriminate among samples. Furthermore, analysis and interpretation of the eigenvectors from the PCA models developed verified that the IR spectrum of the grape juice has enough information to allow the prediction of the variety. These results also suggested that IR spectroscopy coupled with pattern recognition methods holds the necessary information for a successful classification of juice samples of different varieties.     

Classification and Authentication of Barley (<Emphasis Type="Italic">Hordeum vulgare</Emphasis>) Malt Varieties: Combining Attenuated Total Reflectance Mid-infrared Spectroscopy with Chemometrics

Attenuated total reflectance (ATR) mid-infrared (MIR) spectroscopy combined with chemometrics was explored as a tool to classify and authenticate Australian barley varieties. Grain samples (n = 162) were sourced from eight commercial barley varieties and analysed in the MIR range. Principal component analysis (PCA), discriminant partial least squares regression (PLS-DA), linear discriminant analysis (LDA) and soft independent modelling of class analogy (SIMCA) were used to classify the barley grain samples according to variety. PLS-DA correctly classified barley varieties between 91 and 100 %. The results have demonstrated the usefulness of ATR-MIR spectroscopy combined with chemometrics as a rapid method to classify barley grain samples according to their variety. Although MIR is not routinely available at the receival point in most of the cereal trade companies, it has the potential to be used in breeding programmes.     

Discrimination Between Shaoxing Wines and Other Chinese Rice Wines by Near-Infrared Spectroscopy and Chemometrics

Shaoxing rice wine (also called Shaoxing wine) is the most well-known Chinese rice wine in China. The common fraudulent practice in the commercialization of Chinese rice wine is to sell wines from different geographical origins under the denomination of Shaoxing rice wine. In this study, the use of near-infrared (NIR) spectroscopy combined with chemometrics as a rapid tool for the discrimination of Chinese rice wine from three geographical origins (“Fujian”, “non-Shaoxing”, “Shaoxing”) has been preliminarily investigated. NIR spectra were collected in transmission mode in the wavelength range of 800–2,500 nm. Discriminant models were developed by principal component analysis (PCA), discriminant analysis (DA), and discriminant partial least-squares analysis (DPLS). The chemical properties of Chinese rice wine were also investigated to find out the difference between samples from three varied origins. The results showed that good classification could be obtained after spectral pre-treatment. The percentage of samples correctly classified by both DA and DPLS methods in calibration and validation set was 97.2% and 100%, respectively. The results demonstrated that NIR could be used as a simple and rapid technique to distinguish Shaoxing wines from non-Shaoxing wines and Fujian wines. To further validate the ability of NIR spectroscopy, more samples should be incorporated to build a more robust model.     

Detection of adulterants such as sweeteners materials in honey using near-infrared spectroscopy and chemometrics

Near-infrared (NIR) spectroscopy combined with chemometrics methods has been used to detect adulteration of honey samples. The sample set contained 135 spectra of authentic (n = 68) and adulterated (n = 67) honey samples. Spectral data were compressed using wavelet transformation (WT) and principal component analysis (PCA), respectively. In this paper, five classification modeling methods including least square support vector machine (LS-SVM), support vector machine (SVM), back propagation artificial neural network (BP-ANN), linear discriminant analysis (LDA), and K-nearest neighbors (KNN) were adopted to correctly classify pure and adulterated honey samples. WT proved more effective than PCA, as a means for variables selection. Best classification models were achieved with LS-SVM. A total accuracy of 95.1% and the area under the receiver operating characteristic curves (AUC) of 0.952 for test set were obtained by LS-SVM. The results showed that WT-LS-SVM can be as a rapid screening technique for detection of this type of honey adulteration with good accuracy and better generalization.     

Classification of extra virgin olive oils according to their genetic variety using linear discriminant analysis of sterol profiles established by ultra-performance liquid chromatography with mass spectrometry detection

A method to classify extra virgin olive oils (EVOOs) according to their genetic variety using sterol profiles obtained by ultra-performance liquid chromatography (UPLC) with atmospheric pressure chemical ionization mass spectrometry (APCI-MS) detection has been developed. The optimal separation conditions were obtained using a gradient acetonitrile/water (0.01% acetic acid) at a flow rate of 0.8 mL min^− 1 and 10 °C. Linear discriminant analysis (LDA) models were constructed with the 11 UPLC-APCI-MS sterol peaks taken from the selective ion recording mode chromatograms. Ratios of the peak areas selected by pairs were used as predictors. With the sequential application of two LDA models and using a 95% probability, the EVOO samples belonging to seven genetic varieties mainly produced at La Comunitat Valenciana, Spain, were correctly classified with a prediction capability higher than 97%.     

Biogenic Amine Content in Red Wines from Different Protected Designations of Origin of Southern Italy: Chemometric Characterization and Classification

Biogenic amines are contaminants naturally present in wines. Their occurrence is influenced by several factors including oenological and agricultural practices, grape variety, and geographical origin. For these reasons, they have been chosen as marker to characterize and classify 56 Italian red wines belonging to four protected designations of origin (PDO) from Southern Italy. Principal component analysis and cluster analysis were applied on data obtained by HPLC/RF in order to highlight the natural grouping of samples. Afterward, linear discriminant analysis and partial least squares were used to classify the wines according to their PDO. Biogenic amines are demonstrated to be a reliable and useful marker for the characterization and classification of the four Southern Italian PDOs investigated. Both the linear discriminant analysis (LDA) and the partial least squares discriminant analysis (PLS-DA) achieved 100 % of wines correctly classified and predicted. Therefore, the determination of these compounds in red wines can play an important role in wine quality assessment, by providing information for the prevention of potential detrimental effects on health and for the characterization of PDO labeled wines.     

Tracing the geographical origin of honeys based on volatile compounds profiles assessment using pattern recognition techniques

The goal of this study was to examine the possibility of verifying the geographical origin of honeys based on the profiles of volatile compounds. A head-space solid phase microextraction (SPME) combined with comprehensive two-dimensional gas chromatography–time-of-flight mass spectrometry (GC × GC–TOFMS) was used to analyze the volatiles in honeys with various geographical and floral origins. Once the analytical data were collected, supervised pattern recognition techniques were applied to construct classification/discrimination rules to predict the origin of samples on the basis of their profiles of volatile compounds. Specifically, linear discriminant analysis (LDA), soft independent modeling of class analogies (SIMCA), discriminant partial least squares (DPLS) and support vector machines (SVM) with the recently proposed Pearson VII universal kernel (PUK) were used in our study to discriminate between Corsican and non-Corsican honeys. Although DPLS and LDA provided models with high sensitivities and specificities, the best performance was achieved by the SVM using PUK. The results of this study demonstrated that GC × GC–TOFMS combined with methods like LDA, DPLS and SVM can be successfully applied to detect mislabeling of Corsican honeys.     

Mid infrared and fluorescence spectroscopies coupled with factorial discriminant analysis technique to identify sheep milk from different feeding systems

Mid infrared spectroscopy (MIR) combined with multivariate data analysis was used to discriminate between ewes milk samples according to their feeding systems (controls, ewes fed scotch bean and ewes fed soybean). The MIR spectra were scanned throughout the first 11 weeks of the lactation stage. When factorial discriminant analysis (FDA) with leave one-out cross-validation was applied, separately, to the three spectral regions in the MIR (i.e. 3000–2800, 1700–1500 and 1500–900 cm⁻¹), the classification rate was not satisfactory. Therefore, the first principal component (PCs) scores (corresponding to 3, 10 and 10 for, respectively, the 3000–2800, 1700–1500 and 1500–900 cm⁻¹) of the principal component analysis (PCA) extracted from each of the data sets were pooled (concatenated) into a single matrix and analysed by FDA. Correct classification amounting to 71.7% was obtained. Finally, the same procedure was applied to the MIR and fluorescence data sets and 98% of milk samples were found to be correctly classified. Milk samples belonging to control and soybean groups were 100% correctly classified. Regarding milk samples originating from the scotch bean group, only 2 out of 33 samples were misclassified. It was concluded that concatenation of the data sets collected from the two spectroscopic techniques is an efficient tool for authenticating milk samples according to their feeding systems, regardless of the lactation stage.     

Selecting the best set of gas chromatography-derived fatty acids to discriminate between two finishing diets using linear discriminant analysis

Linear discriminant analysis (LDA) was investigated as a method for identifying the type of finishing diet fed to bulls (n = 169) based on gas chromatography fatty acid (FA) analysis. The bulls were fed ad libitum a high concentrate diet comprised of a cereal–soybean meal based concentrate plus straw offered separately (HC) or a total mixed ration made of cereal, soybean meal, maize silage and straw (TMR). Eleven variables (10 FA and one FA ratio) were selected as statistically significant predictors out of 41 variables tested. The Mahalanobis squared distance between the HC and TMR groups was 3.386 and F-test of the distance was highly significant (P > 0.001). In cross-validated classification matrices, 18 cases were misclassified in the HC group and 16 cases were misclassified in the TMR group. As a result, 79.9% of original grouped cases were classified correctly. We concluded that it was possible to classify beef samples according to their finishing diets using LDA.     

电子鼻对酿酒酵母菌株产香特性的评价

利用电子鼻PEN3 系统对不同酿酒酵母酿制葡萄酒的芳香成分进行检测分析。通过电子鼻系统动态采集葡萄酒试样的芳香成分,利用主成分分析(PCA)、线性判断分析(LDA)进行数据分析,两种分析方法都能较好的区分不同酵母对应的葡萄酒试样,表明酿酒酵母的产香能力具有菌株多态性,而电子鼻能够对其差异进行检测并加以区分。同时结合Loadings 分析方法得知,除7 号(对硫化物灵敏)、9 号(对有机硫化物灵敏)和10 号(对烷烃灵敏)外,其他的传感器在菌株的产香差异分析中起主要作用。     

Prediction of the curing time of Spanish hams using peptide profiles established by capillary zone electrophoresis

A capillary zone electrophoresis procedure for the prediction of curing time of Spanish hams using peptide profiles has been developed. Excellent resolution between the seven peptide peaks was achieved within 30 min analysis time with a BGE containing 60 mM MgSO₄ and 50 mM phosphate at pH 2.5. Using hams with curing times of 6, 8 and 12 months, both linear discriminant analysis (LDA) and multiple linear regression (MLR) models were constructed. In both cases, two different normalisation procedures of the peak areas were compared. Using LDA, all the ham samples corresponding to the three categories were correctly classified. Using MLR, the ham curing time could be predicted with average prediction errors below 2.5%.     

Classification of vegetable oils according to their botanical origin using amino acid profiles established by High Performance Liquid Chromatography with UV–vis detection: A first approach

A preliminary study using amino acid profiles to classify oils according to their botanical origin has been performed. Amino acid profiles were obtained from hydrolysis of proteins present in vegetable oils, and established by High Performance Liquid Chromatography (HPLC) with UV–vis detection. Proteins present in hazelnut, corn, soybean, olive, avocado, peanut and grapeseed oils were precipitated with acetone, and the residue was hydrolysed in acid medium. The amino acids obtained were derivatized with o-phthaldialdehyde and separated by HPLC. Peaks corresponding to 18 amino acids were observed using a C18 column and a gradient of acetonitrile–water in the presence of a 5 mM citric/citrate buffer at pH 6.5. The 16 peaks observed in each sample (arginine–serine and phenylalanine–leucine peaks appeared overlapped) were used to construct linear discriminant analysis (LDA) models. Ratios of the peak areas selected by pairs were used as predictors. With a LDA model, the oils were correctly classified with assignment probabilities higher than 99%.     

Polyphenolic Profile of Red Wines for the Discrimination of Controlled Designation of Origin

The aims were to determine the polyphenolic profile of red wines from Spanish Designation of Origin (DO) Rías Baixas and Ribeira Sacra and to evaluate the feasibility of using polyphenolic profiles and chemometric tools to classify wines for authentication purposes. Trans-resveratrol, oenin, malvin, (+)-catechin, (?)-epicatechin, quercetin and syringic acid were determined in 39 samples. Soft independent modeling of class analogy (SIMCA), linear discriminant analysis (LDA) and support vector machine classification (SVM) were used to classify. For resveratrol, Ribeira Sacra red wines showed higher values than Rías Baixas wines (4.60 and 3.36 mg L^?1, respectively). SVM classification was adequate for 100 % classification of wines by their polyphenolic profile. SIMCA classification was also adequate for wine classification of DO Rías Baixas and Ribeira Sacra wines. As conclusion, polyphenolic profile can be used for classification of DOs. The models can discriminate outside wines. Thus, this approach can be useful for authentication purposes.     

Monitoring the molecular changes by front face fluorescence spectroscopy throughout ripening of a semi-hard cheese

Twenty four semi-hard cheeses produced during autumn (n = 12) and summer (n = 12) periods were manufactured and ripened at an industrial scale. Tryptophan and vitamin A fluorescence spectra were scanned on the 24 cheeses at 2, 30 and 60 days of ripening. Principal component analysis (PCA) and factorial discriminant analysis (FDA) were applied on the spectral data sets. The first five principal components (PCs) of the PCA extracted from each data set (tryptophan or vitamin A) of cheeses produced during autumn or summer period were pooled into a single matrix and analysed by FDA. Regarding cheeses produced during the autumn period, the percentage of samples correctly classified was 95.8% and 86.1% for the calibration and validation samples, respectively. Similar results were obtained from cheeses produced during the summer period. Finally, concatenation technique was applied to the tryptophan and vitamin A spectra recorded on cheeses independently of their production seasons. Correct classification was observed for 87.5% and 80.6% for the calibration and validation samples, respectively. Although this statistical technique did not allow 100% correct classification for all groups, the results obtained were promising considering the significant effect of the season on the cheese characteristics.     

Anthocyanins, flavonols and hydroxycinnamates: an attempt to use them to discriminate Vitis vinifera L. cv ‘Barbera’ clones

Anthocyanin and flavonol profiles of Vitis vinifera berry skin have been diffusely studied in past years to identify the effects exerted by climate, environment and cultural practices on their biosynthesis. They have also been used for chemotaxonomic purposes with the aim of classifying grape varieties. Hydroxycinnamates and phenolic acids are the most important group of non-flavonoid phenols in grapes and wines. In the present work six ‘Barbera’ clones were grown in the same site to evaluate the influence of two seasons on the accumulation of flavonoids and hydroxycinnamates at maturity. Berry skins were extracted in an ethanolic buffer and flavonoids and hydroxycinnamates were separated by HPLC. Two principal component analysis (PCA) models were built to identify phenolic parameters exploitable to classify clones. The PCA scores were taken further to perform discriminant analysis to evaluate the degree of classification possible. A significant seasonal variability was observed for most phenolic features, whereas some parameters such as total anthocyanin expressed on a per berry basis, the sum of tri-hydroxylated anthocyanin percentages, the percentages of kaempferol glucuronide and the total hydroxycinnamate content were stable over the seasons. The percentage of individual anthocyanin alone, not associated with maturity data, was not effective in classifying clones; in association with maturity data it allowed to discriminate clones, similarly to what it was previously assessed for classifying varieties. The results indicated that LDA models developed on the PCA scores including maturity data correctly classified 75% of clones.     

Histamine production by bacilli bacteria, acetic bacteria and yeast isolated from fruit wines

Eleven red wines imported from foreign country and 40 domestic fruit wines, including 15 red wines, 4 white wines, 7 plum wines, and 14 other fruit wines, sold in the supermarkets in Taiwan were purchased and tested to determine the occurrence of biogenic amines and histamine-forming bacteria. The levels of pH, total soluble solids (TSS), titratable acidity (TA), reducing sugar (RS), total sugar (TS), sulphites, methanol (milligram per liter of pure ethanol), ethanol and Pb in all samples ranged from 3.0 to 4.1, 6.8 to 24.4 °Brix, 0.3 to 1.7 g/100 mL, 0.2 to 17.6 g/100 mL, 1.6 to 28.4 g/100 mL, <2 to 260.5 mg/L, <1 to 2559 mg/L, 5.0 to 15.6 g/100 mL and <1 to 46.2 μg/L, respectively. The levels of TSS, TA, RS, and TS in plum wine samples were significantly higher than those of the other wines samples, whereas the pH value in plum wine samples was lower than that of the other wines samples. The average content for each of the nine biogenic amines in all samples was less than 5.2 mg/L. However, higher levels of histamine and spermine were detected in domestic fruit wine samples than the imported red wine samples. Five histamine-forming isolates isolated from domestic red wine and jackfruit wine, capable of producing 13.0 mg/L to 69.1 mg/L of histamine in trypticase soy broth (TSB) supplemented with 2 g/100 mL l-histidine (TSBH) or MRS broth supplemented with 2 g/100 mL l-histidine (MRSH), were identified as Bacillus pumilus (one strain), Bacillus sp. (two strains) and Acetobacter pasteurianus (one strain) by 16S rDNA sequencing with PCR amplification, and Zygoascus hellenicus var. hellenicus (one strain) by internal transcribed spacer sequencing with PCR amplification. To our knowledge, this is the first report to demonstrate the occurrence of histamine-forming bacilli bacteria, acetic bacteria and yeast in fruit wine.     

A market study on the quality characteristics of eggs from different housing systems

To study the differences among commercial eggs from four housing systems i.e. cage, barn, free range and organic, 41 physical and chemical parameters were evaluated on 28 fresh egg samples from the Italian market. The univariate statistic analysis evidenced that organic eggs had the highest whipping capacity and foam consistency but the lowest freshness (the highest air cell height) and albumen quality (the lowest Haugh Unit); cage eggs presented instead the lowest whipping capacity and the highest shell resistance to breaking. The multivariate technique discriminant partial least-squares regression was unable to correctly classify the eggs from the four housing systems but successfully differentiated cage eggs from alternative (organic + barn + free range) eggs. The variables with the most discriminant power were shell breaking resistance, overrun, protein content, and shell thickness.