Wine Metabolomics: Objective Measures of Sensory Properties of Semillon from GC-MS Profiles

Abstract

The contribution of volatile aroma compounds to the overall composition and sensory perception of wine is well recognized. The classical targeted measurement of volatile compounds in wine using GC-MS is laborious and only a limited number of compounds can be quantified at any time. Application of an automated multivariate curve resolution technique to nontargeted GC-MS analysis of wine makes it possible to detect several hundred compounds within a single analytical run. Hunter Valley Semillon (HVS) is recognized as a world class wine with a range of styles. Subtle characters reliant upon the development of bottle maturation characteristics are a feature of highly esteemed HVS. In this investigation a metabolomic approach to wine analysis, using multivariate curve resolution techniques applied to GC-MS profiles coupled with full descriptive sensory analysis, was used to determine the objective composition of various styles of HVS. Over 250 GC-MS peaks were extracted from the wine profiles. Sensory scores were analyzed using PARAFAC prior to development of predictive models of sensory features from the extracted GC-MS peak table using PLS regression. Good predictive models of the sensorial attributes honey, toast, orange marmalade, and sweetness, the defining traits for HVS, could be determined from the extracted peak tables. Compound identification for these rated attributes indicated the importance of a range of ethyl esters, aliphatic alcohols and acids, ketones, aldehydes, furanic derivatives, and norisoprenoids in the development of HVS and styles. The development of automated metabolomic data analysis of GC-MS profiles of wines will assist in the development of wine styles for specific consumer segments and enhance understanding of production processes on the ultimate sensory profiles of the product.

Clonal differences and impact of defoliation on Sauvignon blanc (Vitis vinifera L.) wines: a chemical and sensory investigation

“…Logarithmic scaling was considered appropriate since concentrations vary by large orders of magnitude across the different compounds and scaling using the Pareto equation was used to ensure that weighting of variables with large concentrations did not dominant compounds with small to medium concentrations. 34 Orthogonalisation was performed to improve model interpretation by removing variance not associated with the experimental factors by projecting the data matrix onto a vector of sample class (clone or defoliation) and removing a maximum of one latent variable prior to determining the final models. OPLS-DA was conducted using external parameter orthogonalisation in the PLS toolbox version 7.8 (Eigenvector Research Inc, Wenatchee, WA, USA) within Matlab version 7.14.0.739 (The Mathworks, Natick, MA, USA).…”

“…Prior to OPLS‐DA compound concentrations were preprocessed by Log10 scaling, mean centre and standardisation of variance using a Pareto equation. Logarithmic scaling was considered appropriate since concentrations vary by large orders of magnitude across the different compounds and scaling using the Pareto equation was used to ensure that weighting of variables with large concentrations did not dominant compounds with small to medium concentrations . Orthogonalisation was performed to improve model interpretation by removing variance not associated with the experimental factors by projecting the data matrix onto a vector of sample class (clone or defoliation) and removing a maximum of one latent variable prior to determining the final models.…”

Wine science in the metabolomics era

“…Recently, multivariate curve resolution techniques applied to non-targeted GC-MS profiles of wine coupled with full descriptive sensory analysis allowed predictive models using partial least-squares regression [77]. Good predictive models of the sensorial attributes of Semillon wines were achieved thanks to the development of automated metabolomic GC-MS data.…”

Evaluation of changes induced in rice metabolome by Cd and Cu exposure using LC-MS with XCMS and MCR-ALS data analysis strategies

“…This robustness is demonstrated by the fact that satisfactory results can also be obtained when dealing with lower resolution data. For instance, MCR-ALS has successfully overcome inherent difficulties of GC-MS and CE-MS data analysis, such as multiple MS signals for the same metabolite due to derivatization in GC-MS and large migration time shifts between samples in CE-MS [33,34].…”