Investigation of Components that Affect Flavors and Visualizing Differences in Tastes

Importance of the Topic

Food flavor evaluation traditionally relies on sensory panels using the five senses. However, subjective bias and limited reproducibility have driven demand for complementary instrumental methods to identify and quantify chemical components that underlie specific taste attributes. In beverage research and quality control, linking objective chemical profiles to sensory perceptions enables more precise flavor design and faster product development cycles.

Objectives and Study Overview

This study aimed to establish a systematic process flow for identifying components responsible for the ambiguous sake flavor attribute “fukurami,” the sensation of flavor expansion in the mouth. By combining sensory rankings with comprehensive chemical analysis and multivariate modeling, researchers sought to isolate key flavor and aroma markers and validate an analytical model capable of predicting fukurami from chemical data.

Methods and Instrumentation

The research involved eight sake samples standardized to 15 % alcohol, differing in rice polishing ratios and yeast strains. Sensory analysis by trained assessors ranked samples by perceived fukurami and grouped them into “with” or “without” categories. Instrumental measurements included:

• Liquid chromatography–tandem mass spectrometry (LC–MS/MS) using a Shimadzu LCMS-8060 with the Ion-pair-free Method Package for Primary Metabolites to quantify 153 polar metabolites (amino acids, organic acids, nucleotides, saccharides).
• Headspace gas chromatography–mass spectrometry (HS-GC/MS) on a Shimadzu GC-2010 Plus and QP2020 to identify 21 volatile aroma compounds per NIST20 and Wiley libraries.

Main Results and Discussion

Principal component analysis (PCA) of flavor and aroma data initially grouped samples by yeast strain rather than fukurami perception. Partial least squares–discriminant analysis (PLS-DA) integrating sensory categories revealed key markers distinguishing “with” versus “without” fukurami:

• Higher relative levels of disaccharides and monosaccharides were associated with fukurami, indicating enhanced sweetness.
• Lower proportions of organic acids such as malic acid correlated with fukurami, suggesting reduced acidity.
• Distinct levels of nucleobases (cytidine) and amino acids (ornithine) contributed to differentiation.
• In aroma data, fusel alcohols and esters—particularly isobutanol, isobutyl acetate, and ethyl acetate—were enriched in samples exhibiting fukurami.

Percentage composition plots confirmed that samples with fukurami had a higher share of sweet components and aroma-active esters. Using only the selected markers, support vector machine models achieved accurate classification of unknown samples previously uncharacterizable by human panel data, demonstrating the predictive power of the analytical workflow.

Benefits and Practical Applications

The established process flow enables:

• Objective identification of chemical drivers behind complex sensory attributes.
• Predictive models to assist brewers in selecting raw materials and fermentation conditions for desired flavor profiles.
• Enhanced quality control by monitoring key markers to ensure batch‐to‐batch consistency.

Future Trends and Potential Uses

Future developments may include:

• Extension of the workflow to other ambiguous taste or mouthfeel attributes in food and beverages.
• Integration with metabolomics and machine learning platforms for real‐time prediction and process control.
• Expansion of marker libraries to capture microbial and enzymatic contributions in fermentation.

Conclusion

This study demonstrates a robust analytical‐sensory integration workflow that successfully identifies and validates chemical markers for the sake of fukurami. Multivariate models based on LC/MS and GC/MS data can predict sensory outcomes, offering a valuable tool for flavor optimization and quality assurance.

References

• Ozaki K. et al. Journal of the Brewing Society of Japan, 103(3):150–162 (2008).
• Furukawa H. et al. Journal of the Brewing Society of Japan, 78(6):419–422 (1983).
• Yoshizawa K., Koizumi T. Journal of the Brewing Society of Japan, 92(3):217–223 (1997).
• Yoshizawa K. Journal of the Brewing Society of Japan, 75(6):451–457 (1980).