Variable selection for multivariate classification aiming to detect individual adulterants and their blends in grape nectars

Discrimination of whole grape juice using fluorescence spectroscopy data with linear discriminant analysis coupled to genetic and ant colony optimisation algorithms

In this study, a new approach was developed for classifying grape juices produced in Brazil using unfolded excitation-emission matrix (EEM) fluorescence spectroscopy and chemometrics, with respect to the agricultural production system, namely the conventional or organic agricultural one. Linear discriminant analysis (LDA) coupled to ant colony optimisation (ACO) and the genetic algorithm (GA) were used to select a more effective subset of variables to discriminate grape juice samples. The best results demonstrated highly efficient classification of grape juice samples according to a conventional or organic production process with an accuracy rate of up to 97% for the models and 94% in the prediction of these classes for samples external to the model. The models showed high selectivity and sensitivity with a rate of up to 100% for the training and test datasets, in addition to determining the most significant variables that explain the separation of classes. The proposed method proves to be viable, as it is fast and requires minimal sample preparation, allowing quality control in the food industry.

Variable selection in the chemometric treatment of food data: A tutorial review

Food analysis covers aspects of quality and detection of possible frauds to ensure the integrity of the food. The arsenal of analytical instruments available for food analysis is broad and allows the generation of a large volume of information per sample. But this instrumental information may not yet give the desired answer; it must be processed to provide a final answer for decision making. The possibility of discarding non-informative and/or redundant signals can lead to models of better accuracy, robustness, and chemical interpretability, in line with the principle of parsimony. Thus, in this tutorial review, we cover aspects of variable selection in food analysis, including definitions, theoretical aspects of variable selection, and case studies showing the advantages of variable selection-based models concerning the use of a wide range of non-informative and redundant instrumental information in the analysis of food matrices.

Rapid detection of apple juice concentrate adulteration with date concentrate, fructose and glucose syrup using HPLC-RID incorporated with chemometric tools

The present study investigates the substitute of apple juice concentrate with some cheap sweeteners including glucose syrup, fructose syrup, and date concentrate, as the most common adulterants. For this purpose, pure and authenticated apple juice concentrate was individually adulterated with 10% to 50% of glucose syrup, fructose syrup, and date concentrate. High-performance liquid chromatography coupled with a refractive index detector (HPLC-RID) was applied to determine the carbohydrates profile of samples. The results of HPLC-RID were subjected to multivariate statistical analysis, namely principal component analysis (PCA) and linear discriminant analysis (LDA). The results showed that the glucose/fructose ratio and maltose content were the best indicators to detect adulteration of apple juice concentrate. A set of glucose, sorbitol, sucrose, maltose, and glucose/fructose ratio was used as a discriminating factor. Using this approach, adulteration of apple juice concentrate with cheaper sweeteners was detected at a limit of 10%, depending on the adulterant.

Selected Instrumental Techniques Applied in Food and Feed: Quality, Safety and Adulteration Analysis

This review presents an overall glance at selected instrumental analytical techniques and methods used in food analysis, focusing on their primary food science research applications. The methods described represent approaches that have already been developed or are currently being implemented in our laboratories. Some techniques are widespread and well known and hence we will focus only in very specific examples, whilst the relatively less common techniques applied in food science are covered in a wider fashion. We made a particular emphasis on the works published on this topic in the last five years. When appropriate, we referred the reader to specialized reports highlighting each technique's principle and focused on said technologies' applications in the food analysis field. Each example forwarded will consider the advantages and limitations of the application. Certain study cases will typify that several of the techniques mentioned are used simultaneously to resolve an issue, support novel data, or gather further information from the food sample.

Detection of camellia oil adulteration using chemometrics based on fatty acids GC fingerprints and phytosterols GC-MS fingerprints

The fatty acid, squalene, and phytosterols, coupled to chemometrics were utilized to detect the adulteration of camellia oil (CAO) with palm superolein (PAO), refined olive oil (ROO), high oleic- sunflower oil (HO-SUO), sunflower oil (SUO), corn oil (COO), rice bran oil (RBO), rice oil (RIO), peanut oil (PEO), sesame oil (SEO), soybean oil (SOO), and rapeseed oil (RAO). CAO was characterized with higher triterpene alcohols, thus differentiated from other vegetable oils in principle component analysis (PCA). Using partial least squares-discriminant analysis (PLS-DA), CAO adulterated with PAO, ROO, HO-SUO, SUO, COO, RBO, RIO, PEO, SEO, SOO, RAO (5%-100%, w/w), could be classified, especially higher than 92.31% of the total discrimination accuracy, at an adulterated ratio above 30%. With less than 22 potential key markers selected by the variable importance in projection (VIP), the optimized PLS models were confirmed to be accurate for the adulterated level prediction in CAO.

ROC curves for the optimization of one-class model parameters. A case study: Authenticating extra virgin olive oil from a Catalan protected designation of origin

This paper proposes a ROC curve-based methodology to find optimal classification model parameters. ROC curves are implemented to set the optimal number of PCs to build a one-class SIMCA model and to set the threshold class value that optimizes both the sensitivity and specificity of the model. The authentication of the geographical origin of extra-virgin olive oils of Arbequina botanical variety is presented. The model was developed for samples from Les Garrigues, target class, Samples from Siurana were used as the non-target class. Samples were measured by FT-Raman with no pretreatment. PCA was used as exploratory technique. Spectra underwent pre-treatment and variables were selected based on their VIP score values. ROC curve and others already known criteria were applied to set the threshold class value. The results were better when the ROC curve was used, obtaining performance values higher than 82%, 75% and 77% for sensitivity, specificity and efficiency, respectively.

Circulating miR-27a-3p as a candidate for a biomarker of whole grain diets for lipid metabolism

Circulating miR-27a-3p was involved in the process of lipid synthesis under the dietary patterns of whole grain diets, and the expression of miR-27a-3p was decreased in serum, while it was elevated both in liver and ileum.

Rapid Detection of Adulterants in Whey Protein Supplement by Raman Spectroscopy Combined with Multivariate Analysis

The growing demand for whey protein supplements has made them the target of adulteration with cheap substances. Therefore, Raman spectroscopy in tandem with chemometrics was proposed to simultaneously detect and quantify three common adulterants (creatine, l-glutamine and taurine) in whey protein concentrate (WPC) powder. Soft independent modeling class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA) models were built based on two spectral regions (400-1800 cm-1 and 500-1100 cm-1) to classify different types of adulterated samples. The most effective was the SIMCA model in 500-1100 cm-1 with an accuracy of 96.9% and an error rate of 5%. Partial least squares regression (PLSR) models for each adulterant were developed using two different Raman spectral ranges (400-1800 cm-1 and selected specific region) and data pretreatment methods. The determination coefficients (R2) of all models were higher than 0.96. PLSR models based on typical Raman regions (500-1100 cm-1 for creatine and taurine, the combination of range 800-1000 cm-1 and 1300-1500 cm-1 for glutamine) were superior to models in the full spectrum. The lowest root mean squared error of prediction (RMSEP) was 0.21%, 0.33%, 0.42% for creatine, taurine and glutamine, and the corresponding limit of detection (LOD) values for them were 0.53%, 0.71% and 1.13%, respectively. This proves that Raman spectroscopy with the help of multivariate approaches is a powerful method to detect adulterants in WPC.