An overview of multivariate qualitative methods for food fraud detection

Multi-level data fusion strategies for modeling three-way electrophoresis capillary and fluorescence arrays enhancing geographical and grape variety classification of wines

Capillary electrophoresis with diode array detection (CE-DAD) and multidimensional fluorescence spectroscopy (EEM) second-order data were fused and chemometrically processed for geographical and grape variety classification of wines. Multi-levels data fusion strategies on three-way data were evaluated and compared revealing their advantages/disadvantages in the classification context. Straightforward approaches based on a series of data preprocessing and feature extraction steps were developed for each studied level. Partial least square discriminant analysis (PLS-DA) and its multi-way extension (NPLS-DA) were applied to CE-DAD, EEM and fused data matrices structured as two-way and three-way arrays, respectively. Classification results achieved on each model were evaluated through global indices such as average sensitivity non-error rate and average precision. Different degrees of improvement were observed comparing the fused matrix results with those obtained using a single one, clear benefits have been demonstrated when level of data fusion increases, achieving with the high-level strategy the best classification results.

The Sample, the Spectra and the Maths-The Critical Pillars in the Development of Robust and Sound Applications of Vibrational Spectroscopy

The last two decades have witnessed an increasing interest in the use of the so-called rapid analytical methods or high throughput techniques. Most of these applications reported the use of vibrational spectroscopy methods (near infrared (NIR), mid infrared (MIR), and Raman) in a wide range of samples (e.g., food ingredients and natural products). In these applications, the analytical method is integrated with a wide range of multivariate data analysis (MVA) techniques (e.g., pattern recognition, modelling techniques, calibration, etc.) to develop the target application. The availability of modern and inexpensive instrumentation together with the access to easy to use software is determining a steady growth in the number of uses of these technologies. This paper underlines and briefly discusses the three critical pillars-the sample (e.g., sampling, variability, etc.), the spectra and the mathematics (e.g., algorithms, pre-processing, data interpretation, etc.)-that support the development and implementation of vibrational spectroscopy applications.

Fraud in Animal Origin Food Products: Advances in Emerging Spectroscopic Detection Methods over the Past Five Years

Animal origin food products, including fish and seafood, meat and poultry, milk and dairy foods, and other related products play significant roles in human nutrition. However, fraud in this food sector frequently occurs, leading to negative economic impacts on consumers and potential risks to public health and the environment. Therefore, the development of analytical techniques that can rapidly detect fraud and verify the authenticity of such products is of paramount importance. Traditionally, a wide variety of targeted approaches, such as chemical, chromatographic, molecular, and protein-based techniques, among others, have been frequently used to identify animal species, production methods, provenance, and processing of food products. Although these conventional methods are accurate and reliable, they are destructive, time-consuming, and can only be employed at the laboratory scale. On the contrary, alternative methods based mainly on spectroscopy have emerged in recent years as invaluable tools to overcome most of the limitations associated with traditional measurements. The number of scientific studies reporting on various authenticity issues investigated by vibrational spectroscopy, nuclear magnetic resonance, and fluorescence spectroscopy has increased substantially over the past few years, indicating the tremendous potential of these techniques in the fight against food fraud. It is the aim of the present manuscript to review the state-of-the-art research advances since 2015 regarding the use of analytical methods applied to detect fraud in food products of animal origin, with particular attention paid to spectroscopic measurements coupled with chemometric analysis. The opportunities and challenges surrounding the use of spectroscopic techniques and possible future directions will also be discussed.

Classification of proanthocyanidin profiles using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) spectra data combined with multivariate analysis

Proanthocyanidin (PAC) profiles of apples (a-PAC), cranberries (c-PAC), and peanut skins (p-PAC) were determined by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). Deconvolution of overlapping isotopic patterns indicated that in apples, only 5% of the PAC oligomers contain one or more A-type bonds, whereas in cranberries and peanut skins, 96% of the PAC oligomers contain one or more A-type bonds. MALDI-TOF MS data combined with multivariate analysis, such as principal component analysis (PCA) and linear discriminant analysis (LDA), were used to differentiate and discriminate a-PAC, c-PAC, and p-PAC from one another. Mixtures of c-PAC with either a-PAC or p-PAC at different w/w ratios were evaluated by LDA modeling. The LDA model classified the training, testing, and validation sets with 99.4%, 100%, and 94.2% accuracy. Results suggest that MALDI-TOF MS and multivariate analysis are useful in determining authenticity of PAC from different sources and mixtures of PAC sources.

Chemometric analysis reveals influences of hot air drying on the degradation of polyphenols in red radish

Hot air drying is a commonly used technology in the preservation of red radish. This study was designed to investigate the correlations among total polyphenol content, total flavonoid content, antioxidant activities and polyphenol compounds in hot air dried red radish via chemometric analysis. UHPLC-QqQ-MS/MS analysis detected nine non-anthocyanin polyphenols and one anthocyanin in fresh and dried red radish samples, and found that hot air drying at 80 °C caused an increase in the p-coumaric acid and ferulic acid content of the red radish. The integral effect of hot air drying on the polyphenol profile of red radish was analyzed by principle component analysis, while sparse partial least squares-discriminant analysis showed that hot air drying induced changes mainly in the contents of poncirin, naringenin, phloetin and cyanidin-3-glucoside. These polyphenol degradations occurred as non-spontaneous and endothermic reactions during the hot air drying process, following first-order reaction kinetics.

Development of an FTIR based chemometric model for the qualitative and quantitative evaluation of cane sugar as an added sugar adulterant in apple fruit juices

A Fourier Transform Infrared Spectroscopy based chemometric model was evaluated for the rapid identification and estimation of cane sugar as an added sugar adulterant in apple fruit juices. For all the ninety samples, spectra were acquired in the mid-infrared range (4000 cm^-1-400 cm^-1). The spectral analysis provided information regarding the distinctive variable region, which lies in the range of 1200cm^-1 to 900cm^-1, designated as fingerprint region for the carbohydrates. A specific peak in the fingerprint region was observed at 997cm^-1 in all the adulterated samples and was undetectable in pure samples. Based on different levels of cane sugar adulteration (5, 10, 15, and 20%), principal component analysis showed the clustering of samples and further helped us in compression of data by selecting wavenumbers with maximum variability based on the loading line plot. Supervised classification methods (SIMCA and LDA) were evaluated based on their classification efficiencies for a test set. Though SIMCA showed 100% classification efficiency (Raw data set), LDA was able to classify the test set with an accuracy of only 96.67% (Raw as well as Transformed data set) between pure and 5% adulterated samples. For the quantitative estimation, calibration models were developed using partial least square regression (PLS-R) and principal component regression method (PCR) methods. PLS-1^st derivative showed a maximum coefficient of determination (R²) with a value of 0.991 for calibration and 0.992 for prediction. The RMSECV, RMSEP, LOD and LOQ observed for PLS-1^st derivative model were 0.75% w/v, 0.61% w/v, 1.28%w/v and 3.88%w/v, respectively. The coefficient of variation as a measure of precision (repeatability) was also determined for all models, and it ranged from 0.23% to 1.83% (interday), and 0.25% to 1.43% (intraday).

Differentiation of fresh and thawed Atlantic salmon using NMR metabolomics

NMR metabolomics approach was used to distinguish fresh and thawed Atlantic salmon. Statistical analysis revealed significant differences in the concentration of some metabolites in reference and frozen-thawed fish during its storage. It was found that salmon freezing/thawing caused a significant increase in the concentration of fumarate and phenylalanine in stored salmon muscle. The concentration of fumarate increased until the 3rd-5th day after thawing and then gradually decreased, reaching zero after two weeks of storage. The concentration of phenylalanine was constantly increased during the storage time. Furthermore, it was detected that aspartate was formed in the flesh of only thawed fish after the second day of storage. Its concentration followed the same trend as fumarate reaching its maximal concentration on the 3rd-5th day after thawing (up to 3.8 mg in 100 g of muscle) and gradually decreased to zero. Aspartate formation was influenced by storage time after thawing and not by the time after slaughter. We propose to use the formation of aspartate in stored salmon flesh as a marker of salmon freezing/thawing.

Volatile and sensory analysis to discriminate meat from lambs fed different concentrate-based diets

Context Diet is one the most important pre-slaughter factors that potentially influences meat quality, but its effect on flavour quality remains equivocal. Aim The aim of the present study was to investigate the effect of diet composition on the flavour and sensory quality of meat from Texel × Scottish Blackface ram lambs. Methods Groups of 11 lambs were assigned to one of the following four dietary treatments for 54 days before slaughter: a concentrate containing barley, maize and soybean (C treatment); C supplemented with a saturated fat source (Megalac®); C supplemented with protected linseed oil; a by-product-based diet containing citrus pulp, distillers grain and soybean. Samples of cooked M. longissimus thoracis et lumborum were subjected to volatile analysis involving solid-phase microextraction followed by gas chromatography–mass spectrometry and to sensory analysis performed by a trained panel. Key results Univariate analysis of volatile data and sensory data showed few differences due to dietary treatments. However, multivariate analysis of the volatile data, and to a lesser extent the sensory profile data, showed potential to discriminate between lamb meat samples, on the basis of the different dietary treatments. Conclusions The inclusion of certain dietary ingredients in the diets of lambs to enhance the nutritional profile of lamb meat (through increasing n-3 fatty acid content) or to reduce feed-formulation costs (through the use of by-products) has minor effects on sensory quality but permits some discrimination between dietary treatments following the application of multivariate analysis. Implications The application of the findings is in allowing lamb producers to use alternative feed types without affecting the sensory quality of lamb negatively, but with the potential to discriminate lamb meat on the basis of its dietary background.

Differential diagnosis of hereditary hemolytic anemias in a single multiscreening test by TGA/chemometrics

A multi-screening test based on the coupling of thermogravimetry and chemometrics was optimized for the differential diagnosis of hereditary hemolytic anemias.

Application of Spectroscopic UV-Vis and FT-IR Screening Techniques Coupled with Multivariate Statistical Analysis for Red Wine Authentication: Varietal and Vintage Year Discrimination

One of the most important issues in the wine sector and prevention of adulterations of wines are discrimination of grape varieties, geographical origin of wine, and year of vintage. In this experimental research study, UV-Vis and FT-IR spectroscopic screening analytical approaches together with chemometric pattern recognition techniques were applied and compared in addressing two wine authentication problems: discrimination of (i) varietal and (ii) year of vintage of red wines produced in the same oenological region. UV-Vis and FT-IR spectra of red wines were registered for all the samples and the principal features related to chemical composition of the samples were identified. Furthermore, for the discrimination and classification of red wines a multivariate data analysis was developed. Spectral UV-Vis and FT-IR data were reduced to a small number of principal components (PCs) using principal component analysis (PCA) and then partial least squares discriminant analysis (PLS-DA) and linear discriminant analysis (LDA) were performed in order to develop qualitative classification and regression models. The first three PCs used to build the models explained 89% of the total variance in the case of UV-Vis data and 98% of the total variance for FR-IR data. PLS-DA results show that acceptable linear regression fits were observed for the varietal classification of wines based on FT-IR data. According to the obtained LDA classification rates, it can be affirmed that UV-Vis spectroscopy works better than FT-IR spectroscopy for the discrimination of red wines according to the grape variety, while classification of wines according to year of vintage was better for the LDA based FT-IR data model. A clear discrimination of aged wines (over six years) was observed. The proposed methodologies can be used as accessible tools for the wine identity assurance without the need for costly and laborious chemical analysis, which makes them more accessible to many laboratories.

A review of Fourier Transform Infrared (FTIR) spectroscopy used in food adulteration and authenticity investigations

The increasing demand for food and the globalisation of the supply chain have resulted in a rise in food fraud, and recent high profile cases, such as the Chinese milk scandal in 2008 and the EU horsemeat scandal in 2013 have emphasised the vulnerability of the food supply system to adulteration and authenticity frauds. Fourier Transform Infrared (FTIR) spectroscopy is routinely used in cases of suspected food fraud as it offers a rapid, easy and reliable detection method for these investigations. In this review, we first present a brief summary of the concepts of food adulteration and authenticity as well as a discussion of the current legislation regarding these crimes. Thereafter, we give an extensive overview of FTIR as an analytical technique and the different foods where FTIR analysis has been employed for food fraud investigations as well as the subsequent multivariate data analyses that have been applied successfully to investigate the case of adulteration or authenticity. Finally, we give a critical discussion of the applications and limitations of FTIR, either as a standalone technique or incorporated in a test battery, in the fight against food fraud.

A Bottom-Up Approach for Data Mining in Bioaromatization of Beers Using Flow-Modulated Comprehensive Two-Dimensional Gas Chromatography/Mass Spectrometry

In this study, we report the combination of comprehensive two-dimensional gas chromatography (GC×GC) with multivariate pattern recognition through template matching for the assignment of the contribution of Brazilian Ale 02 yeast strain to the aroma profile of beer compared with the traditional Nottingham yeast. Volatile organic compounds (VOC) from two beer samples, which were fermented with these yeast strains were sampled using headspace solid-phase microextraction (HS-SPME). The aroma profiles from both beer samples were obtained using GC×GC coupled to a fast scanning quadrupole mass spectrometer. Data processing performed through multiway principal components analysis succeeded in separating both beer samples based on yeast strain. The execution of a simple and reliable procedure succeeded and identified 46 compounds as relevant for sample classification. Furthermore, the bottom-up approach spotted compounds found exclusively in the beer sample fermented with the Brazilian yeast, highlighting the bioaromatization properties introduced to the aroma profile by this yeast strain.

Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry Typings of Edible Oils through Spectral Networking of Triacylglycerol Fingerprints

Adulteration of edible oils by the manufacturers has been found frequently in modern societies. Due to the complexity of the chemical contents in edible oils, it is challenging to quantitatively determine the extent of adulteration and prove the authenticity of edible oils. In this study, a robust and simple MALDI-TOF-MS platform for rapid fingerprinting of triacylglycerols (TAGs) in edible oils was developed, where spectral similarity analysis was performed to quantitatively reveal correlations among edible oils in the chemical level. Specifically, we proposed oil networking, a spectral similarity-based illustration, which enabled reliable classifications of tens of commercial edible oils from vegetable and animal origins. The strategy was superior to traditional multivariate statistics due to its high sensitivity in probing subtle changes in TAG profiles, as further demonstrated by the success in determination of the adulterated lard in a food fraud in Taiwan. Finally, we showed that the platform allowed quantitative assessment of the binary mixture of olive oil and canola oil, which is a common type of olive oil adulteration in the market. Overall, these results suggested a novel strategy for chemical fingerprint-based quality control and authentication of oils in the food industry.

Opportunities and Challenges Using Non-targeted Methods for Food Fraud Detection

In recent years, non-targeted methods have been a popular "buzz" phrase in food fraud detection. Using analytical instrumentation techniques, non-targeted methods have been developed and applied in many food and agricultural situations. However, confusion and misstatements remain regarding how the methods are used. This perspective will discuss the definitions related to non-targeted testing, the procedure of developing and validating methods, the techniques and data analysis, and opportunities and challenges regarding the use of this class of analytical methods. The perspective seeks to provide readers with the latest information regarding recent advances in the use of non-targeted methods.

Alternative data mining/machine learning methods for the analytical evaluation of food quality and authenticity - A review

In recent years, the variety and volume of data acquired by modern analytical instruments in order to conduct a better authentication of food has dramatically increased. Several pattern recognition tools have been developed to deal with the large volume and complexity of available trial data. The most widely used methods are principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), soft independent modelling by class analogy (SIMCA), k-nearest neighbours (kNN), parallel factor analysis (PARAFAC), and multivariate curve resolution-alternating least squares (MCR-ALS). Nevertheless, there are alternative data treatment methods, such as support vector machine (SVM), classification and regression tree (CART) and random forest (RF), that show a great potential and more advantages compared to conventional ones. In this paper, we explain the background of these methods and review and discuss the reported studies in which these three methods have been applied in the area of food quality and authenticity. In addition, we clarify the technical terminology used in this particular area of research.

Multivariate Analysis Illuminates the Effects of Vacuum Drying on the Extractable and Nonextractable Polyphenols Profile of Loquat Fruit

The current study evaluated the effects of vacuum drying on the whole polyphenol profile of loquat fruit, including extractive and nonextractive polyphenols. Absorbance analysis determined that total polyphenol content and antioxidant levels were higher in loquat fruit vacuum dried at 140 °C than in loquat fruit vacuum dried at 70 °C. The results of ultra-HPLC-triple quadruple mass spectrum analysis showed that 15 phenolic acids and 17 flavonoids were found in dried loquat fruit. Multivariate integrative (MINT) sparse partial least square-discriminant analysis showed that vacuum drying affects the polyphenol profile of loquat fruit. Co-analysis of principal component analysis, partial least square-discriminant analysis, and orthometric partial least square-discriminant analysis revealed that vacuum drying mainly changed the content of chlorogenic acid, cryptochlorogenic acid, protocatechuic acid, phloretin, and hesperidin in loquat fruit. Chlorogenic acid (12.020 to 39.153 µg/g d.b. [dried base weight]), the main polyphenol in dried loquat fruit, was degraded to caffeic acid (0.028 to 2.365 µg/g d.b.) and protocatechuic acid (0.014 to 18.285 µg/g d.b.) during vacuum drying. Moreover, vacuum drying also induced the isomerization of chlorogenic acid into cryptochlorogenic acid (1.628 to 12.737 µg/g d.b.). These results might be used to develop dried loquat fruit with high levels of polyphenols and antioxidant activity. PRACTICAL APPLICATION: Interests in polyphenols of loquat fruit had increased greatly because of their possible role in health benefits. This work provided a holistic insight in the effects of vacuum drying on polyphenols profile of loquat fruit. Current results have contributed to the development of vacuum-drying method, which produced loquat fruit rich in polyphenols. Furthermore, it also suggested that multivariate analysis was a feasible method to reveal the important changes of polyphenols profile during food processing.

An Innovative Metabolomic Approach for Golden Rum Classification Combining Ultrahigh-Performance Liquid Chromatography-Orbitrap Mass Spectrometry and Chemometric Strategies

A comprehensive fingerprinting strategy for golden rum classification considering different categories such as fermentation barrel, raw material, and aging is provided, using a metabolomic fingerprinting approach. A nontarget fingerprinting of 30 different rums using liquid chromatography coupled to high-resolution mass spectrometry (Exactive Orbitrap mass analyzer, LC-HRMS) was applied. Principal component analysis (PCA) was used to assess the overall structure of the data and to identify potential outliers. Different chemometric analyses such as partial least-squares discriminant analysis (PLS-DA) were used. A variable importance in projection (VIP) selection method was applied to identify the most significant markers that allow group separation. Compounds related to aging and fermentation processes such as furfural derivates (e.g., hydroxymethylfurfural) and sugars (e.g., glucose, mannitol) were found as the most discriminant compounds (VIP threshold value >1.5). Suitable separation according to selected categories was achieved, and a classification ability of the models of close to 100% was achieved.

Electrostatic Spray Ionization-Mass Spectrometry for Direct and Fast Wine Characterization

Due to the globally existed and economically motivated adulteration including mislabeling and/or blending, fast wine characterization is important in wine industry. Herein, we developed an electrostatic spray ionization-mass spectrometry (ESTASI-MS)-based method to distinguish wines. Wine samples were directly analyzed by ESTASI-MS without any pretreatment. Microdroplets of wine were deposited on a plastic plate for analysis. The collection of each mass spectrometric datum can be accomplished in 1-2 min without any need of pretreatment to the sample, followed by principle component analysis to discriminate wines with different labels and vintages. Long-term storage of wine was simulated and characterized by utilizing the method. High-performance liquid chromatography-MS was further applied to identify the distinctive compounds in wines to indicate their difference. We found that the method can offer a strategy for quick wine analysis, which is of practical value in wine industry for wine classification and aging control.

Chromatographic Data Segmentation Method: A Hybrid Analytical Approach for the Investigation of Antiviral Substances in Medicinal Plant Extracts

The methodology described in this article will significantly reduce the time required for understanding the relations between chromatographic data and bioactivity assays. The methodology is a hybrid of hypothesis-based and data-driven scientific approaches. In this work, a novel chromatographic data segmentation method is proposed, which demonstrates the capability of finding what volatile substances are responsible for antiviral and cytotoxic effects in the medicinal plant extracts. Up until now, the full potential of the separation methods has not been exploited in the life sciences. This was due to the lack of data ordering methods capable of adequately preparing the chromatographic information. Furthermore, the data analysis methods suffer from multidimensionality, requiring a large number of investigated data points. A new method is described for processing any chromatographic information into a vector. The obtained vectors of highly complex and different origin samples can be compared mathematically. The proposed method, efficient with relatively small sized data sets, does not suffer from multidimensionality. In this novel analytical approach, the samples did not need fractionation and purification, which is typically used in hypothesis-based scientific research. All investigations were performed using crude extracts possessing hundreds of phyto-substances. The antiviral properties of medicinal plant extracts were investigated using gas chromatography-mass spectrometry, antiviral tests, and proposed data analysis methods. The findings suggested that (i) β- cis-caryophyllene, linalool, and eucalyptol possess antiviral activity, while (ii) thujones do not, and (iii) α-thujone, β-thujone, cis- p-menthan-3-one, and estragole show cytotoxic effects.

Geographical Classification of Tannat Wines Based on Support Vector Machines and Feature Selection

Geographical product recognition has become an issue for researchers and food industries. One way to obtain useful information about the fingerprint of wines is by examining that fingerprint’s chemical components. In this paper, we present a data mining and predictive analysis to classify Brazilian and Uruguayan Tannat wines from the South region using the support vector machine (SVM) classification algorithm with the radial basis kernel function and the F-score feature selection method. A total of 37 Tannat wines differing in geographical origin (9 Brazilian samples and 28 Uruguayan samples) were analyzed. We concluded that given the use of at least one anthocyanin (peon-3-glu) and the radical scavenging activity (DPPH), the Tannat wines can be classified with 94.64% accuracy and 0.90 Matthew’s correlation coefficient (MCC). Furthermore, the combination of SVM and feature selection proved useful for determining the main chemical parameters that discriminate with regard to the origin of Tannat wines and classifying them with a high degree of accuracy. Additionally, to our knowledge, this is the first study to classify the Tannat wine variety in the context of two countries in South America.