Exploratory data fusion of untargeted multimodal LC-HRMS with annotation by LCMS-TOF-ion mobility: White wine case study.

Applied sciences have increased focus on omics studies which merge data science with analytical tools. These studies often result in large amounts of data produced and the objective is to generate meaningful interpretations from them. This can sometimes mean combining and integrating different datasets through data fusion techniques. The most strategic course of action when dealing with products of unknown profile is to use exploratory approaches. For omics, this means using untargeted analytical methods and exploratory data analysis techniques. The current study aimed to perform data fusion on untargeted multimodal (negative and positive mode) liquid chromatography-high-resolution mass spectrometry data using multiple factor analysis. The data fusion results were interpreted using agglomerative hierarchical clustering on biplot projections. The study reduced the thousands of spectral signals processed to less than a hundred features (a primary parameter combination of retention time and mass-to-charge ratios, RT_m/z). The correlations between cluster members (samples and features from) were calculated and the top 10% highly correlated features were identified for each cluster. These features were then tentatively identified using secondary parameters (drift time, ion mobility constant and collision cross-section values) from the ion mobility spectra. These ion mobility (secondary) parameters can be used for future studies in wine chemical analysis and added to the growing list of annotated chemical signals in applied sciences.

Chemometric and sensometric techniques in enological data analysis.

Enological evaluations capture the chemical and sensory space of wine using different techniques; many sensory methods as well as a variety of analytical chemistry techniques contribute to the amount of information generated. Data fusion, especially integrating data sets, is important when working with complex systems. The success reported when trying to integrate different modalities is generally low and has been attributed to the lack of statistically considerate strategies focusing on the data handling process. Multiple stages of data handling must be carefully considered when dealing with multi-modal data. In this review, the different stages in the data analysis process were examined. The study revealed misconceptions surrounding the process and elucidated rules for purpose-driven approaches by examining the complexities of each stage and the impact the decisions made at each stage have on the resulting models. The two major modeling approaches are either supervised (discrimination, classification, prediction) or unsupervised (exploration). Supervised approaches were emphatic on the pre-processing steps and prioritized increasing performance. Unsupervised approaches were mostly used for preliminary steps. The review found aspects often neglected when it came to the data collection and capturing which in the end contributed to the low success in combining sensory and chemistry data.

Relationship between anthocyanins, proanthocyanidins, and cell wall polysaccharides in grapes and red wines. A current state-of-art review.

Numerous research studies have evaluated factors influencing the nature and levels of phenolics and polysaccharides in food matrices. However, in grape and wines most of these works have approach these classes of compounds individually. In recent years, the number of publications interconnecting classes have increased dramatically. The present review relates the last decade's findings on the relationship between phenolics and polysaccharides from grapes, throughout the entire winemaking process up to evaluating the impact of their relationship on the red wine sensory perception. The combination and interconnection of the most recent research studies, from single interactions in model wines to the investigation of the formation of complex macromolecules, brings the perfect story line to relate the relationship between phenolics and polysaccharides from the vineyard to the glass. Grape pectin is highly reactive toward grape and grape derived phenolics. Differences between grape cultivars or changes during grape ripeness will affect the extractability of these compounds into the wines. Therefore, the nature of the grape components will be crucial to understand the subsequent reactions occurring between phenolics and polysaccharide of the corresponding wines. It has been demonstrated that they can form very complex macromolecules which affect wine color, stability and sensory properties.

Untangling the impact of red wine maceration times on wine ageing. A multidisciplinary approach focusing on extended maceration in Shiraz wines.

Phenolic composition of young red wines has been shown to play an important role in their ageing potential. Therefore, the modulation of phenolic extraction during maceration may influence the subsequent phenolic evolution of these wines. The present work aimed to evaluate the impact of three different maceration times on the phenolic levels and evolution observed over time, using spectrophotometric and chromatography methods, and the effect on the aroma, taste, and mouthfeel sensory properties using Projective Mapping. Additionally, grape cell wall deconstruction was monitored during the extended maceration phase by GC-MS and Comprehensive Comprehensive Microarray Polymer Profiling (CoMPP). Our findings demonstrated that longer maceration times did not always correspond to an increase in wine phenolic concentration, although the level of complexity of these molecules seemed to be higher. Additionally, continuous depectination and possible solubilisation of the pectin is observed during the extended maceration which may be influencing the sensory perception of these wines. Maceration time was also shown to influence the evolution of the polymeric fraction and sensory perception of the wines.

Exploring the Typicality, Sensory Space, and Chemical Composition of Swedish Solaris Wines.

The Swedish wine industry has exponentially grown in the last decade. However, Swedish wines remain largely unknown internationally. In this study, the typicality and sensory space of a set of twelve wines, including five Swedish Solaris wines, was evaluated blind by Swedish wine experts. The aim of the work was to evaluate whether the Swedish wine experts have a common concept of what a typical Solaris wines should smell and taste like or not and, also, to bring out more information about the sensory space and chemical composition of Solaris wines. The results showed a lack of agreement among the wine experts regarding the typicality of Solaris wines. This, together with the results from the sensory evaluation, could suggest the possibility of different wine styles for Solaris wines. From a chemical perspective, the global volatile profile showed a larger variability between individual wines than between Solaris and non-Solaris. However, 4MMP, ethyl propionate, ethyl 2-Methyl propanoate, and diethyl succinate were significantly higher in Solaris wines. Concerning non-volatile compounds, the results showed a significant discrimination between Solaris and non-Solaris wines, the former being characterized by higher ethanol %, Mg, succinic acid, tartaric acid, and sucrose levels.

Wine Quality Drivers: A Case Study on South African Chenin Blanc and Pinotage Wines.

The aim of the study was to propose a methodology for the elucidation of sensory and chemical wine quality drivers. The winners of the 2018 Top 10 Chenin Blanc and Top 10 Pinotage challenges and additional lower scoring wines for each cultivar were evaluated. The two sets underwent sensory profiling by Check-All-That-Apply (CATA) and a 20-point quality rating by industry experts in non-competition conditions and chemical fingerprinting by Liquid Chromatography-High Resolution Mass Spectrometry (LC-HRMS). Data were submitted to Correspondence Analysis (CA) and Principal Component Analysis (PCA) for sensory and chemistry, respectively, from which the standardised deviates were correlated to quality scores to identify the quality drivers. The results illustrated the possibility to determine positive and negative sensory quality drivers (attributes), while the identification of drivers for chemistry (ions) was challenging due to the number of signals generated by the fingerprinting technique. The configurations of the sensory and chemical spaces were compared, but the similarities were relatively low as measured by Regression Vector (RV) coefficients, 0.437 and 0.505 for Pinotage and Chenin Blanc, respectively. The proposed methodology can also be used to explore the sensory space of wine sample sets with the added dimension of the quality drivers which, in turn, highlight the experts' opinions on what makes a winning wine.

Comparison of an Offline SPE-GC-MS and Online HS-SPME-GC-MS Method for the Analysis of Volatile Terpenoids in Wine.

The aroma profile is an important marker for wine quality. Various classes of compounds are responsible for the aroma of wine, and one such class is terpenoids. In the context of this work, a validated gas chromatography-mass spectrometry (GC-MS) method for the quantitation of terpenoids in red and white wine using headspace solid-phase microextraction (HS-SPME) and solid-phase extraction (SPE) was established. Calibrations were performed in the respective base wine using both sample preparation methods. The linearity, precision and accuracy evaluated for the respective matrices were excellent for both sample preparations. However, the HS-SPME approach was more sensitive and more accurate. For both sample preparations, the quantification limits were lower than the odor thresholds in wine. The terpenoid concentrations (µg/L) were evaluated for 13 white wines using both sample preparation methods. Importantly, the online HS-SPME approach was more sensitive than the offline SPE method. The major terpenoids identified in the white wines evaluated were linalool (0.2-63 µg/L), geraniol (nd-66 µg/L) and α-terpineol (nd-85 µg/L).

Investigation of olfactory interactions of low levels of five off-flavour causing compounds in a red wine matrix.

The qualitative sensory perception of individual and of complex mixtures of five compounds, guaiacol ('burnt note'), o-cresol ('phenolic/tar'), 4-ethylphenol (4-EP, 'leather/barnyard'), 2-iso-butyl-3-methoxypyrazine (IBMP, 'green pepper/herbaceous'), and 2,4,6-trichloroanisole (TCA, 'cork taint/ mouldy') were tested in a partially de-aromatised red wine matrix using descriptive analysis by a trained panel of eleven judges. Compounds were characterised at peri- and sub-threshold concentrations using a partial D-optimal statistical design and response surface methodology. Results indicated that complex mixtures in red wine elicit an olfactory response that could not be predicted from the attributes or descriptors of single compounds. Positive sweet/fruity attributes were more intense in solutions containing fewer off-flavour compounds. Novel findings of this study include that IBMP at sub- and peri-threshold levels shows perceptual interaction with volatile phenols at the same levels, and samples containing combinations of these compounds manifested herbaceous and burnt characteristics. Olfactory interactions of this many off-flavour compounds have not been investigated previously in one study. The findings have direct implications for wines made from cultivars that are known to contain these compounds, and add to the understanding of the behaviour and impact of very low levels (peri- and sub-threshold) of volatile phenols, IBMP, and TCA derived from various sources during winemaking.

Viability of IR spectroscopy for the accurate measurement of yeast assimilable nitrogen content of grape juice.

Up to date, there have been only a few reports on the measurement of YAN and/or its components using IR spectroscopy, suffering from various limitations (number of samples, validation strategies, etc.). In this work, three IR spectral instruments measuring in different modes and ranges of the IR spectrum (FT-IR, FT-NIR, and ATR-MIR), were compared and evaluated for their accuracy to measure both total YAN as well as the components, FAN and ammonia separately, using over 900 grape juice samples from 28 cultivars over three seasons. The global and vintage-based models were evaluated using R2CAL/VAL, RMSEC/P, and RPDCAL/VAL. Randomization tests were used for pair-wise comparison of models. FT-IR and FT-NIR instruments gave the best results, while ATR-MIR can be used for screening purposes. Considering the accuracy, robustness, high throughput, and cost-effective nature, the models produced by both FT-IR and FT-NIR spectroscopy can provide winemakers with the opportunity to make timelier and more informed nutrient supplementation decisions, facilitating the achievement of their desired wine style and quality.

A multivariate approach to evaluating the chemical and sensorial evolution of South African Sauvignon Blanc and Chenin Blanc wines under different bottle storage conditions.

Volatile compound composition contributes to the aroma profile of wine and is susceptible to change due to oxidation which may occur during storage and transportation, especially at high temperatures. Changes in sensory attributes may also occur, altering the sensory profile of wine. Classical univariate analysis only looks at the deviations for one factor at a time and may overlook the overall effect of treatments. In this study, changes in South African Sauvignon Blanc and Chenin Blanc wine sensory profile, volatile and antioxidant-related parameters resulting from storage under different temperatures (room temperature, 15 °C and 25 °C) and durations (0, 3 and 9 months) were investigated using a multivariate approach. Bottled, unwooded wines of both cultivars from six wineries were used. As expected, the chemical evolution of the wines was characterised by increases in absorbance at 420 nm (browning), colour density and hue with prolonged storage at high temperatures. To be able to compare the evolution of the sample sets regardless of the initial (T0/control) wine profile and composition, multivariate regression analysis in the form of regression vector (RV) coefficients were used to assess the correlations in the sensory and chemical changes relative to the control in each set. Using Pivot©Profile for the first time in this type of stability assessment and applying a new algorithm for data handling in addition to the classical one, this study showed that prolonged exposure to higher temperatures resulted in the change from fruity to toasted aroma attributes.

Matrix effects influencing the perception of 3-mercaptohexan-1-ol (3MH) and 3-mercaptohexyl acetate (3MHA) in different Chenin Blanc wines by Projective Mapping (PM) with Ultra Flash profiling (UFP) intensity ratings.

Interaction studies are used in sensory evaluation to elucidate the complex influences (additive, masking, synergistic) that various compounds have on the perception of wine aroma. To these, the interactions between the matrix and the compounds of interest add another layer of complexity. Unlike previous interaction studies, the current work used a rapid method, Projective Mapping (PM) coupled with intensity of attributes, to evaluate the interaction effects two thiols (3-mercaptohexan-1-ol, 3MH, and 3-mercaptohexyl acetate, 3MHA) have in various matrices. The matrices used were increasingly complex, from model wine to partially dearomatized neutral Chenin Blanc to commercial wines. The results reflected the effect of the thiols interaction with the matrix more than that between compounds. The methodology proposed for the data handling highlighted possible advantages and shortcomings of the Projective Mapping with intensity approach. The choice of base matrix as well as the sensory method are relevant when studying interaction effects, and are dependent on the desired outcome of the experiment.

A statistical exploration of data to identify the role of cultivar and origin in the concentration and composition of yeast assimilable nitrogen.

The study was undertaken to gain insight into the nitrogen status of grape juices currently used to make commercial wines in South Africa. This was done as yeast assimilable nitrogen (YAN) is most often suspected as the cause for problematic fermentations and has major implications for the organoleptic qualities of the final product. Using exploratory statistical methods, this study explored the possibility of identifying the role of cultivar and grape-growing district in the determination of the concentration and composition of YAN. However, as the dataset was found to be non-parametric and heteroscedastic, paired with unequal sample sizes, data analysis was approached with caution. Through the use of various suitable statistical analyses, cultivar was shown to play the more important role in determining the concentration and composition of YAN.

Chitinases and thaumatin-like proteins in Sauvignon Blanc and Chardonnay musts during alcoholic fermentation.

Protein precipitation, also referred to as protein instability, may lead to haziness in bottled wines and result in significant commercial losses. To avoid problems of this nature, fining finished wines with clay (bentonite) is the most commonly applied methodology. However, bentonite fining reduces yield and may affect wine quality. Protein haze has been primarily linked to grape pathogenesis-related proteins, in particular chitinases and thaumatin-like proteins. To better understand the persistence of these proteins during fermentation, reverse phase chromatography was used to monitor the evolution of total grape proteins as well as of chitinases and thaumatin-like proteins during alcoholic fermentation. The data confirm a previously reported significant decrease in total protein content during fermentation. This reduction in total protein levels was observed throughout fermentation, and was affected by factors such as fermentation temperature, yeast strain or grape cultivar. However, significant changes in the concentration of free chitinases were observed in a yeast strain-dependent manner. The data thus confirm the correlation between the levels of yeast cell wall chitin and changes in chitinase concentration, and suggest that it is primarily the amount of lateral chitin, and not the chitin in bud scars, that is responsible for this activity.

Investigating the relationship between cell wall polysaccharide composition and the extractability of grape phenolic compounds into Shiraz wines. Part II: Extractability during fermentation into wines made from grapes of different ripeness levels.

This study evaluated the relationship between cell wall breakdown, from Shiraz grapes harvested at three different ripeness levels and the colour and phenolics extracted during alcoholic fermentation into wines. Phenolic differences between the ripeness treatments were minimal after » of the fermentation was completed. However, colour and phenolic content were significantly higher in finished wines made from 25°Brix grapes compared to those from 21°Brix and 23°Brix. Levels of grape cell wall polysaccharide deconstruction during fermentation was a determining correlative factor in relation to phenolic extractability. In this context, the de-pectination observed during ripening was found to enhance this deconstruction or "opening-up" of the grape pomace during fermentation, thus increasing the differential extraction of specific polyphenols, especially polymeric polyphenols, into the wines. Additionally, the degree of cell wall deconstruction seemed to play a role in the possible retention and extraction of specific grape proanthocyanidins, depending on their nature and polymer length.

Investigating the relationship between grape cell wall polysaccharide composition and the extractability of phenolic compounds into Shiraz wines. Part I: Vintage and ripeness effects.

Phenolic compounds play an important role in colour stability and sensory properties of red wine. This study evaluated berry skin cell wall composition and how this influences grape and wine phenolics at different ripeness levels (21°Brix, 23°Brix, and 25°Brix) over two consecutive vintages. The vintage effect was highly significant, especially in the pectin fraction of the grape cell walls and affected the concentrations of certain phenolics extracted. The climatic variance between the seasons might have influenced the differences observed in the grape cell wall compositions. Firstly, a higher grape and wine phenolic content, especially in polymeric phenols, was found in 2015 wines. Additionally, grape berry cell walls, especially at the earliest stages of ripening, were found to be more intact in 2015 than in 2016. Thus, a possible relationship was found between the degree of berry intactness, especially for pectin-rich components, and the corresponding phenolic extractability during the winemaking.

Testing the Sensitivity of Potential Panelists for Wine Taint Compounds Using a Simplified Sensory Strategy.

The odor detection threshold (ODT) of a compound is the lowest concentration at which individuals can reliably perceive a difference between a sample and its corresponding control, with 50% performance above chance. Wine is a complex matrix, and ODTs used in studies on wine can be based on inappropriate matrices and informal sensory methodologies. Formal studies confirming ODTs in wine are relatively scarce in the literature, and are complex and expensive to carry out. In this study, the sensitivity of panelists to previously published ODTs for five compounds: Guaiacol, o-cresol and 4-ethyl phenol, 3-isobutyl-2-methoxypyrazine (IBMP), and 2,4,6-trichloroanisole (TCA) associated with off-flavor/taint issues in wine, was investigated. The study was carried out in partially de-aromatized young Shiraz wine (unwooded) using a simplified version of the formal sensory approach. A triangle test in triplicate was carried out with 34 panelists, at the ODT for each compound, in one day. The study explored whether previous training affected panelists' sensitivity for threshold differences. Results showed that samples spiked with volatile phenols were significantly different (p = 0.01) to controls. The spiked TCA and IBMP samples were not significantly different from the control in either case. Judges were better able to detect compounds if they had prior experience or training in wine evaluation. Despite some limitations, this pragmatic approach may be useful when carrying out sensory studies with fairly limited resources and within tight timelines, as it provides helpful information on panel members and detection thresholds for a specific matrix.

Measuring Thiols in Single Cultivar South African Red Wines Using 4,4-Dithiodipyridine (DTDP) Derivatization and Ultraperformance Convergence Chromatography-Tandem Mass Spectrometry.

Wine varietal thiols are important contributors to wine aroma. The chemical nature of thiols makes them difficult to measure due to low concentrations, high sensitivity to oxidation, and low ionization. Methods for the measurement of thiols usually consist of multiple steps of sample preparation followed by instrumental measurement. Studies have collected large datasets of thiols in white wine but not in red wine, due to the lack of availability of suitable methods. In this study, for the first time, convergence chromatography was used to measure thiols in red wine at ultratrace levels with improved sensitivity compared to previous methods. Performance parameters (selectivity, linearity, limits of detection, precision, accuracy) were tested to demonstrate the suitability of the method for the proposed application. Red wine thiols were measured in South African Pinotage, Shiraz, and Cabernet Sauvignon wines (n = 16 each). Cultivar differentiation using the thiol profile was demonstrated.

Indications of the prominent role of elemental sulfur in the formation of the varietal thiol 3-mercaptohexanol in Sauvignon blanc wine.

Elemental sulfur is a fungicide traditionally used to control Powdery Mildew in the production of grapes. The presence of sulfur residues in grape juice has been associated with increased production of hydrogen sulfide during fermentation, which could take part in the formation of the varietal thiol 3-mercaptohexanol. This work examines whether elemental sulfur additions to Sauvignon blanc juice can increase the levels of sought-after varietal thiols. Initial trials were performed in South Africa and indicated a positive impact of sulfur on the levels of thiols. Further experiments were then carried out with New Zealand Sauvignon blanc and confirmed a positive relationship between elemental sulfur additions and wine varietal thiols. The formation of hydrogen sulfide was observed when the addition of elemental sulfur was made to clarified juice, along with an increase in further reductive sulfur compounds. When the addition of sulfur was made to pressed juice, prior to clarification, the production of reductive sulfur compounds was drastically decreased. Some mechanistic considerations are also presented, involving the reduction of sulfur to hydrogen sulfide prior to fermentation.

Spectrophotometric Analysis of Phenolic Compounds in Grapes and Wines.

Phenolic compounds are of crucial importance for red wine color and mouthfeel attributes. A large number of enzymatic and chemical reactions involving phenolic compounds take place during winemaking and aging. Despite the large number of published analytical methods for phenolic analyses, the values obtained may vary considerably. In addition, the existing scientific knowledge needs to be updated, but also critically evaluated and simplified for newcomers and wine industry partners. The most used and widely cited spectrophotometric methods for grape and wine phenolic analysis were identified through a bibliometric search using the Science Citation Index-Expanded (SCIE) database accessed through the Web of Science (WOS) platform from Thompson Reuters. The selection of spectrophotometry was based on its ease of use as a routine analytical technique. On the basis of the number of citations, as well as the advantages and disadvantages reported, the modified Somers assay appears as a multistep, simple, and robust procedure that provides a good estimation of the state of the anthocyanins equilibria. Precipitation methods for total tannin levels have also been identified as preferred protocols for these types of compounds. Good reported correlations between methods (methylcellulose precipitable vs bovine serum albumin) and between these and perceived red wine astringency, in combination with the adaptation to high-throughput format, make them suitable for routine analysis. The bovine serum albumin tannin assay also allows for the estimation of the anthocyanins content with the measurement of small and large polymeric pigments. Finally, the measurement of wine color using the CIELab space approach is also suggested as the protocol of choice as it provides good insight into the wine's color properties.

Inactive dry yeast application on grapes modify Sauvignon Blanc wine aroma.

This study investigated the potential to improve wine aroma by applying two inactive dry yeast products (IDYs) at the onset of ripening on Sauvignon Blanc grapes. Both products led to increased reduced glutathione concentrations in the grape juice and corresponding wines, as well as differences in individual higher alcohol acetates (HAAs) and ethyl esters of straight chain fatty acids (EEFAs) at the end of fermentation. After two months of storage, a significantly slower decrease of EEFAs and to a lesser extent of HAAs was found for wines made from grapes with IDY applications. These wines also resulted in significantly slower synthesis of ethyl esters of branched acids, whereas varietal thiols were altered in a product-specific manner. The modifications in the wine chemical composition were also sensorially corroborated. This study showed that vineyard additions of IDY products directly on the grapes at the onset of ripening have a subsequent benefit to the production and preservation of aroma in wines.