Correlation of sensory bitterness in dairy protein hydrolysates: Comparison of prediction models built using sensory, chromatographic and electronic tongue data

Techno-Functions and Safety Concerns of Plant-Based Peptides in Food Matrices

Plant-based peptides (PBPs) benefit functional food development and environmental sustainability. Proteolysis remains the primary method of peptide production because it is a mild and nontoxic technique. However, potential safety concerns still emanate from toxic or allergenic sequences, amino acid racemization, iso-peptide bond formation, Maillard reaction, dose usage, and frequency. The main aim of this review is to investigate the techno-functions of PBPs in food matrices, as well as their safety concerns. The distinctive characteristics of PBPs exhibit their techno-functions for improving food quality and functionality by contributing to several crucial food formulations and processing. The techno-functions of PBPs include solubility, hydrophobicity, bitterness, foaming, oil-binding, and water-holding capacities, which subsequently affect food matrices. The safety and quality of foodstuff containing PBPs depend on the proper source of plant proteins, the selection of processing approaches, and compliance with legal regulations for allergen labeling and safety evaluations. The safety concerns in allergenicity and toxicity were discussed. The conclusion is that food technologists must apply safe limits and consider potential allergenic components generated during the development of food products with PBPs. Therefore, functional food products containing PBPs can be a promising strategy to provide consumers with wholesome health benefits.

Comparison of prediction models for soy protein isolate hydrolysates bitterness built using sensory, spectrofluorometric and chromatographic data from varying enzymes and degree of hydrolysis

The bitterness of soy protein isolate hydrolysates prepared using five proteases at varying degree of hydrolysis (DH) and its relation to physicochemical properties, i.e., surface hydrophobicity (H0), relative hydrophobicity (RH), and molecular weight (MW), were studied and developed for predictive modelling using machine learning. Bitter scores were collected from sensory analysis and assigned as the target, while the physicochemical properties were assigned as the features. The modelling involved data pre-processing with local outlier factor; model development with support vector machine, linear regression, adaptive boosting, and K-nearest neighbors algorithms; and performance evaluation by 10-fold stratified cross-validation. The results indicated that alcalase hydrolysates were the most bitter, followed by protamex, flavorzyme, papain, and bromelain. Distinctive correlation results were found among the physicochemical properties, influenced by the disparity of each protease. Among the features, the combination of RH-MW fitted various classification models and resulted in the best prediction performance.

Bitter Peptides in Fermented Soybean Foods - A Review

Fermented soybean foods with a long history are popular worldwide because of rich nutrition. However, many traditional fermented soybean foods have unacceptable bitterness, which mostly comes from the bitter peptides produced from the hydrolysis of soybean proteins. In this review, the bitter peptides in fermented soybean foods is briefly reviewed. The structural properties of bitter receptors and bitter peptides were reviewed. Bitterness is perceived through the binding between bitter compounds and specific sites of bitter receptors (25 hTAS2Rs), which further activate the downstream signal pathway mediated by G-protein. And it converts chemical signals into electrical signals, and transmit them to the brain. In addition, the influencing factors of bitter peptides in fermented soybean foods were summarized. The bitterness of fermented soybean foods primarily results from the raw materials, microbial metabolism during fermentation, unique techniques, and interactions of various flavor compounds. Moreover, the structure-bitterness relationship of bitter peptides was also discussed in this review. The bitterness degree of the bitter peptide is related to the polypeptide hydrophobicity, amino acids in the peptide, peptide molecular weight and polypeptide spatial structure. Studying the bitter peptides and their bitter characteristics in fermented soybean foods is beneficial for improving the sensory quality of fermented soybean foods and prompting more consumers accept them.

Characteristics of the enzyme-induced release of bitter peptides from wheat gluten hydrolysates

Bitter peptides in the enzymatic hydrolysates were prepared and purified from wheat gluten using aqueous ethanol solutions and macroporous resin, which has opened a new road for the extraction and separation of bitter peptides. This report contains the release regularity of bitter peptides and the factors affecting the change of bitter intensity during enzymatic hydrolysis, providing a scientific basis for the research on debitterizing method. In this study, the effects of different degrees of hydrolysis (DH) and enzyme active sites on the bitter peptide content and bitter taste thresholds were discussed. The relationship between amino acid composition, molecular weight distribution, surface hydrophobicity and bitter taste thresholds was extensively researched. The results showed the exposure of hydrophobic amino acids and the bitterness intensity of the hydrolysates increased as the DH increased, and the bitterness of wheat gluten hydrolysates (WGHs) hydrolyzed by Alcalase was stronger than that of Trypsin. According to correlation analysis, the proportion of total hydrophobic amino acid is the first factor that affects the sensory properties of bitter peptide, and the release content of bitter peptides and the content of total bitter amino acids are the second, following by the content of peptide in the molecular weight range of 500–1,000 Da and the surface hydrophobicity. The amino acid sequence of bitter peptides from WGHs were identified and predicted using high performance liquid chromatography-mass spectrometry (HPLC-MS/MS) and bioinformatics. It was found that the molecular weight of most of the peptides was below 1,500 Da, and the Q value was higher than 5.86 kJ/mol.

Detecting the Bitterness of Milk-Protein-Derived Peptides Using an Electronic Tongue

Bitterness is a considerable limiting factor for the application of bioactive peptides in the food industry. The objective of this study was to compare the level of bitterness of milk-protein-derived peptides using an electronic tongue (E-tongue). Liquid milk protein concentrate (LMPC) was prepared from ultra-heat-treated skimmed cow’s milk. It was initially hydrolyzed with different concentrations of trypsin, namely, 0.008 g·L−1, 0.016 g·L−1 and 0.032 g·L−1. In a later exercise, tryptic-hydrolyzed LMPC (LMPC-T) was further hydrolyzed using Lactobacillus bulgaricus and Streptococcus thermophilus. The effect of glucose in microbial hydrolysis was studied. The bitterness of peptides was evaluated with respect to quinine, a standard bittering agent. The level of bitterness of the peptides after microbial hydrolysis of LMPC-T (LMPC-T-F and LMPC-T-FG) was evaluated using a potentiometric E-tongue equipped with a sensor array that had seven chemically modified field-effect transistor sensors. The results of the measurements were evaluated using principal component analysis (PCA), and subsequently, a classification of the models was built using the linear discriminant analysis (LDA) method. The bitterness of peptides in LMPC-T-F and LMPC-T-FG was increased with the increase in the concentration of trypsin. The bitterness of peptides was reduced in LMPC-T-FG compared with LMPC-T-F. The potential application of the E-tongue using a standard model solution with quinine was shown to follow the bitterness of peptides.

A comprehensive review on the glucoregulatory properties of food-derived bioactive peptides

Diabetes mellitus, a group of metabolic disorders characterized by persistent hyperglycemia, affects millions of people worldwide and is on the rise. Dietary proteins, from a wide range of food sources, are rich in bioactive peptides with antidiabetic properties. Notable examples include AGFAGDDAPR, a black tea-derived peptide, VRIRLLQRFNKRS, a β-conglycinin-derived peptide, and milk-derived peptide VPP, which have shown antidiabetic effects in diabetic rodent models through variety of pathways including improving beta-cells function, suppression of alpha-cells proliferation, inhibiting food intake, increasing portal cholecystokinin concentration, enhancing insulin signaling and glucose uptake, and ameliorating adipose tissue inflammation. Despite the immense research on glucoregulatory properties of bioactive peptides, incorporation of these bioactive peptides in functional foods or nutraceuticals is widely limited due to the existence of several challenges in the field of peptide research and commercialization. Ongoing research in this field, however, is fundamental to pave the road for this purpose.

Taste and stability characteristics of two key umami peptides from pufferfish (Takifugu obscurus)

Takifugu obscurus (T. obscurus) is known for its umami taste. Two taste-active peptides, Pro-Val-Ala-Arg-Met-Cys-Arg (PR-7) and Tyr-Gly-Gly-Thr-Pro-Pro-Phe-Val (YV-8), were proved as key compounds that contributed to the typical taste of T. obscurus. However, whether these peptides have the potential as umami supplements is unknown. The purpose of this study was to investigate the taste characteristics of PR-7 and YV-8, as well as stability at different pH values by sensory evaluation, instrumental analysis and quantum chemical calculation. The results indicated that PR-7 and YV-8 presented umami taste at near neutral pH (6.5-8.0) and had umami-enhancing effects. PR-7 also exhibited significant kokumi activity. Additionally, two peptides showed remarkable stability after different pH treatments, especially YV-8; this may be related to its stable structural property. All the results suggest that both peptides have great potential to be applied in complex foods to provide desirable taste, and act as a feasible alternative to monosodium l-glutamate.

Ultrasound-Assisted Multi-Enzymatic System for the Preparation of ACE Inhibitory Peptides with Low Bitterness from Corn Gluten Meal

The promising angiotensin converting enzyme (ACE) inhibitory peptides derived from corn protein usually have strong bitterness and thus limit their use among consumers. To prepare ACE inhibitory peptides with low bitterness, two energy-efficient types of ultrasound pretreatment were introduced into the multi-enzymatic system of corn gluten meal. The results showed that Flavourzyme–Protamex sequential enzymolysis produced the peptides with high ACE inhibitory activity and the lowest bitterness compared with other enzymolysis conditions. During the optimized sequential enzymolysis, the divergent ultrasound pretreatment with a frequency of 40 kHz for 60 min exhibited higher ACE inhibitory activity (65.36%, accounting for 73.84% of the highest ACE inhibitory activity) and lower bitterness intensity of peptides, compared with an energy-gathered ultrasound. The results of the study showed that, on the one hand, divergent ultrasound pretreatment induced the highest intrinsic fluorescence of protein, with more hydrophobic amino acid residues exposed for cleavage by exopeptidases, which leads to a reduction in bitterness. On the other hand, the amino acid composition analysis proved that more Tyr, Ile, and Val moieties, instead of Leu (bitterest substance), and more peptide fractions with a molecular weight >1000 Da should be the structural features of high ACE inhibitory peptides.

Manufacturing of Plant-Based Bioactive Peptides Using Enzymatic Methods to Meet Health and Sustainability Targets of the Sustainable Development Goals

Due to the rapid growth in the global population, the consumption of animal-based food products/food compounds has been associated with negative implications for food sustainability/security. As a result, there is an increasing demand for the development of plant-based food and compounds as alternatives. Meanwhile, a growing number of studies report the health benefits of food protein-based peptides prepared via enzymatic hydrolysis and exhibiting biological properties such as antioxidant, antihypertensive, anti-thrombotic, and antidiabetic activities. However, the inherent bitterness of some peptides hinders their application in food products as ingredients. This article aims to provide the latest findings on plant-based bioactive peptides, particularly their health benefits, manufacturing methods, detection and qualification of their bitterness properties, as well as debittering methods to reduce or eliminate this negative sensory characteristic. However, there is still a paucity of research on the biological property of debittered peptides. Therefore, the role of plant protein-derived bioactive peptides to meet the health targets of the Sustainable Development Goals can only be realised if advances are made in the industrial-scale bioprocessing and debittering of these peptides.

Sensory evaluation, chemical structures, and threshold concentrations of bitter-tasting compounds in common foodstuffs derived from plants and maillard reaction: A review

The bitterness of foodstuffs is often associated with toxicity, which negatively influences product acceptability. However, bitter compounds have many benefits, and a slight bitter taste is sometimes favored. In this review, we summarize the methods used to isolate and evaluate the taste of bitter compounds in different foods. The chemical structures and threshold concentrations of these compounds are also recapped. Although the structures and thresholds of many bitter compounds have been confirmed, further studies are needed to develop detailed bitter-masking strategies and establish the relation between functional groups (hetero-cyclic substituents and bonding types) and taste quality. Furthermore, a comprehensive bitterness database and chemometric data must be provided in order to quickly assess the bitterness of unfamiliar products.

Analysis of umami taste substances of morel mushroom (Morchella sextelata) hydrolysates derived from different enzymatic systems

Seven enzyme groups were applied to hydrolyze broken fruiting bodies of morel mushroom (Morchella sextelata) to extract umami substances. Physical-chemical properties, as well as compositions and concentrations of quintessential umami compounds of morel hydrolysates were analyzed. Electronic tongue and electronic nose were used to evaluate the sensory characteristics. The results suggested that peptides below 3 kDa showed the highest correlation with umami taste. Morel hydrolysate obtained from Neutrase-Flavourzyme (NF) combination contained the most contents of small peptides (<3 kDa), free amino acids (224.83 ± 0.87 mg/g), as well as flavor 5'-nucleotides (4.84 ± 0.32 mg/g), giving the best overall flavor properties. The reaction conditions of NF were optimized by response surface methodology (RSM). The highest degree of hydrolysis (DH) was up to 36.64%. An enzymatic hydrolysis approach was established to develop novel flavor products with high umami and low bitter taste from morel mushroom.

The use of electronic tongue and sensory panel on taste evaluation of pediatric medicines: a systematic review

The palatability of medications is an essential factor for children's adherence to drug treatment. Several methods for drug taste assessment have been developed. The aim of this review is to explore the literature reports of the main methods for the evaluation of medicines taste, named electronic tongue (e-tongue, in vitro) and human sensory panel. A systematic search was performed up to March 2020 and a total of 88 articles were selected. The e-tongue (57.5%) has been more frequently described than the sensory panel (10.3%), while some articles (32.2%) used both techniques. 74.7% of the articles mentioned 'pediatric', 'paediatric' or 'children' in the text, but only 19.5% developed formulations targeting pediatric audience and sensory testing in children is rarely seen. The e-tongue has predominance of use in the taste evaluation of pediatric medicines probably since it is fast, easy to perform and risk free, besides presenting less imprecise data and no fatigue. The human panel is more realistic, despite its intrinsic variability. In this sense, it is proposed the use of e-tongue as a fast way to select the most promising sample(s) and, after that, the sensory panel should be applied in order to confirm the taste masking.

Principal Component Analysis from Mass Spectrometry Data Combined to a Sensory Evaluation as a Suitable Method for Assessing Bitterness of Enzymatic Hydrolysates Produced from Micellar Casein Proteins

Enzymatic hydrolysis of food proteins generally changes the techno-functional, nutritional, and organoleptic properties of hydrolyzed proteins. As a result, protein hydrolysates have an important interest in the food industries. However, they tend to be characterized by a bitter taste and some off-flavors, which limit their use in the food industry. These tastes and aromas come from peptides, amino acids, and volatile compounds generated during hydrolysis. In this article, sixteen more or less bitter enzymatic hydrolysates produced from a milk protein liquid fraction enriched in micellar caseins using commercially available, food-grade proteases were subjected to a sensory analysis using a trained and validated sensory panel combined to a peptidomics approach based on the peptide characterization by reverse-phase high-performance liquid chromatography, high-resolution mass spectrometry, and bioinformatics software. The comparison between the sensory characteristics and the principal components of the principal component analysis (PCA) of mass spectrometry data reveals that peptidomics constitutes a convenient, valuable, fast, and economic intermediate method to evaluating the bitterness of enzymatic hydrolysates, as a trained sensory panel can do it.

Effect of enzyme-assisted hydrolysis on protein pattern, technofunctional, and sensory properties of lupin protein isolates using enzyme combinations

The modification of lupin protein isolates (LPI) by means of enzymatic hydrolysis (Lupinus angustifolius cultivar Boregine) was performed with four enzyme preparations (Alcalase 2.4 L, Papain, Corolase 7089, and Neutrase 0.8 L) in a one- and two-step process to determine the efficacy for the destruction of major IgE-reactive polypeptides and the evaluation of the technofunctional and sensory properties of lupin protein hydrolysates. Combinations of Alcalase 2.4 L and Papain were most effective in the degradation of polypeptides in L. angustifolius as measured by sodium dodecylsulfate-polyacrylamide gel electrophoresis. The enzymatic hydrolysis of the LPI increased their technofunctional properties such as protein solubility, foam activity, and emulsifying capacity almost independently of the enzyme preparation used. The sensory results showed a significant increase in bitterness from 1.9 for LPI to 5.7 for the combination of Alcalase 2.4 L and Papain in one-step process. The aroma attributes of the hydrolysates were very similar to untreated LPI. The results of this study show the possibility of enzymatic hydrolysis of LPI to destroy the major IgE-reactive polypeptides that increase the technofunctional properties of the isolates and thus their use in human nutrition as food ingredients.

Influence of storage time and elevated ripening temperature on the chemical and sensory properties of white Cheddar cheese

Aged cheese is an increasingly popular dairy product. One approach to reduce Cheddar cheese maturation time is by utilizing elevated temperature, despite potential problems including development of imbalanced or off-flavors and negative changes in texture. Therefore, the objective of this study was to evaluate the influence of elevated ripening temperature on chemical and sensory properties of aged white Cheddar cheese. White Cheddar cheese was aged at 7.2, 10, or 12.8 °C for 12 months, with samples evaluated at 2, 5, 8, 10, 11, and 12 months by a trained sensory panel (n = 10). Two consumer sensory panels (n = 120) assessed 8- and 12-month aged cheese for comparison to a commercially available reference sample of the same cheese, aged for 12 months. An electronic tongue methodology was developed for analysis of nonvolatile compounds. Trained panel results showed that 2-month cheeses were described by milkfat flavor and sweet taste, 5-month cheeses were described by nutty aroma and white color, and 8-, 10-, 11-, and 12-month cheeses developed aged characteristics, such as umami and bitter tastes, brothy aroma, and aged flavor. Consumer panel results showed similar overall liking scores for the reference cheese and cheeses aged at 10 or 12.8 °C for both evaluations. The electronic tongue could classify samples according to aging month with a validity value of 92.59%. In conclusion, the electronic tongue served as a valid method of instrumental analysis for Cheddar cheese samples throughout maturation. This study demonstrated that aging white Cheddar cheese for 8 months at an elevated storage temperature of 10 °C produced cheese similar in consumer acceptance to that aged at 7 °C for 12 months. PRACTICAL APPLICATION: This study showed that aged white Cheddar storage at a higher temperature was perceived similarly by consumers as one stored for 1 year at a slightly lower temperature. This may be useful to those in the dairy industry exploring ways to accelerate aging, reducing devoted resources, while still producing an acceptable product. Also, the electronic tongue was effective at distinguishing among aged white Cheddar cheese samples showing another application for this technology.

Enzymatic hydrolysis of lupin protein isolates-Changes in the molecular weight distribution, technofunctional characteristics, and sensory attributes

Enzymatic hydrolysis of lupin protein isolates (LPI; Lupinus angustifolius L.) was performed with nine different protease preparations to investigate their effect on technofunctionality, sensory properties, and the integrity of the proteins to estimate the reduction of the immunoreactivity. Alcalase 2.4 L, papain, and pepsin were most effective in the degradation of the α- and β-conglutin examined by SDS-PAGE analysis, although the degree of hydrolysis only slightly increased. The technofunctional properties of LPI-solubility, emulsifying, and foaming activity-were improved by most of the proteolytic enzymes with the most impressive increase from 980% foam activity for LPI up to 3,614% foam activity for pepsin hydrolysate. The formation of bitterness, most likely linked to generation of bitter peptides, was pronounced in the Alcalase hydrolysate, while the other hydrolysates did not show an extensive increase in bitterness compared to the LPI. Other sensory attributes of the hydrolysates-with the exception of Alcalase treatment-were also very similar to the LPI. The results of this study show the potential of enzymatic degradation of LPI to modify the IgE-reacting polypeptides and to improve the technofunctionality of the isolates and therefore their use as food ingredients.

Peptidomic screening of bitter and nonbitter casein hydrolysate fractions for insulinogenic peptides

Sodium caseinate hydrolysates (NaCaH) contain biologically active peptides that can positively influence human health. However, their intense bitterness hinders their inclusion in food products. To our knowledge, no studies have investigated whether a correlation between bitterness and bioactivity exists in NaCaH, so it is not yet known what effect selective removal of bitterness has on NaCaH bioactivity. A deeper understanding of the physicochemical characteristics affecting both bitterness and bioactivity is therefore needed. The aim of this study was to use in silico analysis to elucidate the relationship between bitterness and bioactivity of the insulinogenic NaCaH. The NaCaH fractions were generated by membrane filtration and flash chromatography and were subsequently evaluated for bitterness by a sensory panel. In this present study, peptidomic and bioinformatic processing of these NaCaH fractions allowed for the identification of insulinogenic peptides as well as other literature-identified peptides in each of the fractions. The results showed that the most bitter fraction contained the highest abundance of insulinogenic peptides, whereas another bitter fraction contained the highest abundance of other literature-identified bioactive peptides exhibiting angiotensin-converting enzyme-inhibition activity. Although some bioactive peptides were identified in the least bitter fractions, the abundance of these peptides was very low. These observations show a correlation between bitter taste and bioactivity, highlighting potential complications in removing bitterness while maintaining bioactivity. However, as the most bitter fraction contained the highest abundance of insulinogenic peptides, there is potential for using a lower dose of this enriched bioactive fraction to exert health benefits. The second most bitter fraction contained a very low abundance of insulinogenic peptides and other bioactive peptides. Therefore, removal of this fraction could reduce the NaCaH product's bitterness without significantly altering overall bioactive potential.

Discriminating aging and protein-to-fat ratio in Cheddar cheese using sensory analysis and a potentiometric electronic tongue

The objectives of this study were to evaluate the flavor and taste attributes of full-fat Cheddar cheeses with different protein-to-fat ratios (PFR) over aging time using a descriptive sensory analysis panel and a consumer panel, and to correlate these attributes with instrumental parameters obtained by the potentiometric electronic tongue. Three Cheddar cheese formulations (PFR of 0.74, 0.85, and 1.01) were produced in triplicate and composition was verified. Cheese was aged at 7.2°C and evaluated at 2, 5, 8, 10, 11, and 12 mo by a trained panel (n = 10) for 8 flavor and 5 taste attributes and using an electronic tongue for 7 nonvolatile taste attributes. Cheese aged for 12 mo was also evaluated by a consumer sensory panel for liking and intensity attributes. Principal component analysis was performed to discriminate cheese based on aging time and PFR, whereas correlation between sensory and instrumental attributes was assessed using partial least squares regression. Descriptive sensory analysis of flavor and taste attributes differentiated Cheddar cheeses over aging time, but not among PFR formulations. The electronic tongue distinguished changes among cheese samples due to PFR formulation and aging time. The electronic tongue proved successful in characterizing the nonvolatile flavor components in Cheddar cheese and correlated with taste perceptions measured by descriptive sensory analysis. Consumer evaluations showed distinctive attribute profiles for the 3 PFR Cheddar cheese formulations. Overall, higher fat content was associated with increased flavor intensities in Cheddar cheese and drove consumer acceptability and purchase intent ratings. The electronic tongue detected smaller changes in tastes (bitter, metallic, salty, sour, spicy, sweet, and umami) of the 3 PFR formulations over time when compared with the trained panelists, who detected no differences, suggesting that the electronic tongue may be more sensitive to tastants than humans and may have the capability for early detection or identification of problems in a batch of cheese during aging. Results suggest taste quality of cheese may be monitored using the electronic tongue with greater sensitivity than a trained panel, and may be more objective, rapid, and cost effective than human panelists.

Validation of a paper-disk approach to facilitate the sensory evaluation of bitterness in dairy protein hydrolysates from a newly developed food-grade fractionation system

Casein-hydrolysates (NaCaH) are desirable functional ingredients, but their bitterness impedes usage in foods. This study sought to validate a paper-disk approach to help evaluate bitterness in NaCaHs and to develop a food-grade approach to separate a NaCaH into distinct fractions, which could be evaluated by a sensory panel. Membrane filtration generated <0.2-μm and <3-kDa permeates. Further fractionation of the <3-kDa permeate by flash-chromatography generated four fractions using ethanol (EtOH) concentrations of 5, 10, 30 and 50%. As some fractions were poorly soluble in water, the fractions were resolubilzed in EtOH and impregnated into paper-disks for sensory evaluation. Bitterness differences observed in the membrane fractions using this sensory evaluation approach reflected those observed for the same fractions presented as a liquid. The flash-chromatography fractions increased in bitterness with an increase in hydrophobicity, except for the 50% EtOH fraction which had little bitterness. Amino acid analysis of the fractions showed enrichment of different essential amino acids in both the bitter and less bitter fractions.

Practical Applications

The developed food-grade fractionation system, allowed for a simple and reasonably scaled approach to separating a NaCaH, into physicochemically different fractions that could be evaluated by a sensory panel. The method of sensory evaluation used in this study, in which NaCaH samples are impregnated into paper-disks, provided potential solutions for issues such as sample insolubility and limited quantities of sample. As the impregnated paper-disk samples were dehydrated, their long storage life could also be suitable for sensory evaluations distributed by mail for large consumer studies. The research, in this study, allowed for a greater understanding of the physicochemical basis for bitterness in this NaCaH. As some essential amino acids were enriched in the less bitter fractions, selective removal of bitter fractions could allow for the incorporation of the less bitter NaCaH fractions into food products for added nutritional value, without negatively impacting sensory properties. There is potential for this approach to be applied to other food ingredients with undesirable tastes, such as polyphenols.

Optimization and validation of the protocol used to analyze the taste of traditional Chinese medicines using an electronic tongue

Tools to define the active ingredients and flavors of Traditional Chinese Medicines (TCMs) are limited by long analysis times, complex sample preparation and a lack of multiplexed analysis. The aim of the present study was to optimize and validate an electronic tongue (E-tongue) methodology to analyze the bitterness of TCMs. To test the protocol, 35 different TCM concoctions were measured using an E-tongue, and seven replicate measurements of each sample were taken to evaluate reproducibility and precision. E-tongue sensor information was identified and classified using analysis approaches including least squares support vector machine (LS-SVM), support vector machine (SVM), discriminant analysis (DA) and partial least squares (PLS). A benefit of this analytical protocol was that the analysis of a single sample took <15 min for all seven sensors. The results identified that the LS-SVM approach provided the best bitterness classification accuracy (binary classification accuracy, 100%; ternary classification accuracy, 89.66%). The E-tongue protocol developed showed good reproducibility and high precision within a 6 h measurement cycle. To the best of our knowledge, this is the first study of an E-tongue being applied to assay the bitterness of TCMs. This approach could be applied in the classification of the taste of TCMs, and serve important roles in other fields, including foods and beverages.

Evaluation of the Bitterness of Traditional Chinese Medicines using an E-Tongue Coupled with a Robust Partial Least Squares Regression Method

To accurately, safely, and efficiently evaluate the bitterness of Traditional Chinese Medicines (TCMs), a robust predictor was developed using robust partial least squares (RPLS) regression method based on data obtained from an electronic tongue (e-tongue) system. The data quality was verified by the Grubb’s test. Moreover, potential outliers were detected based on both the standardized residual and score distance calculated for each sample. The performance of RPLS on the dataset before and after outlier detection was compared to other state-of-the-art methods including multivariate linear regression, least squares support vector machine, and the plain partial least squares regression. Both R² and root-mean-squares error (RMSE) of cross-validation (CV) were recorded for each model. With four latent variables, a robust RMSECV value of 0.3916 with bitterness values ranging from 0.63 to 4.78 were obtained for the RPLS model that was constructed based on the dataset including outliers. Meanwhile, the RMSECV, which was calculated using the models constructed by other methods, was larger than that of the RPLS model. After six outliers were excluded, the performance of all benchmark methods markedly improved, but the difference between the RPLS model constructed before and after outlier exclusion was negligible. In conclusion, the bitterness of TCM decoctions can be accurately evaluated with the RPLS model constructed using e-tongue data.

Assessment of taste attributes of peanut meal enzymatic-hydrolysis hydrolysates using an electronic tongue

Peanut meal is the byproduct of high-temperature peanut oil extraction; it is mainly composed of proteins, which have complex tastes after enzymatic hydrolysis to free amino acids and small peptides. The enzymatic hydrolysis method was adopted by using two compound proteases of trypsin and flavorzyme to hydrolyze peanut meal aiming to provide a flavor base. Hence, it is necessary to assess the taste attributes and assign definite taste scores of peanut meal double enzymatic hydrolysis hydrolysates (DEH). Conventionally, sensory analysis is used to assess taste intensity in DEH. However, it has disadvantages because it is expensive and laborious. Hence, in this study, both taste attributes and taste scores of peanut meal DEH were evaluated using an electronic tongue. In this regard, the response characteristics of the electronic tongue to the DEH samples and standard five taste samples were researched to qualitatively assess the taste attributes using PCA and DFA. PLS and RBF neural network (RBFNN) quantitative prediction models were employed to compare predictive abilities and to correlate results obtained from the electronic tongue and sensory analysis, respectively. The results showed that all prediction models had good correlations between the predicted scores from electronic tongue and those obtained from sensory analysis. The PLS and RBFNN prediction models constructed using the voltage response values from the sensors exhibited higher correlation and prediction ability than that of principal components. As compared with the taste performance by PLS model, that of RBFNN models was better. This study exhibits potential advantages and a concise objective taste assessment tool using the electronic tongue in the assessment of DEH taste attributes in the food industry.