Electronic tongue-based discrimination of Korean rice wines (makgeolli) including prediction of sensory evaluation and instrumental measurements

A commercial electronic tongue was used to discriminate Korean rice wines (makgeolli) brewed from nine cultivars of rice with different amino acid and fatty acid compositions. The E-tongue was applied to establish prediction models with sensory evaluation or LC–MS/MS by partial least squares regression (PLSR). All makgeollis were classified into three groups by principal components analysis, and the separation pattern was affected by rice qualities and yeast fermentation. Makgeolli taste changed from the complicated comprising sweetness, saltiness, and umami to the uncomplicated, such as bitterness and then, sourness, with a decrease of amino acids and fatty acids in the rice. The quantitative correlation between E-tongue and sensory scores or LC–MS/MS by PLSR demonstrated that E-tongue could well predict most of the sensory attributes with relatively acceptable r2, except for bitterness, but could not predict most of the chemical compounds responsible for taste attributes, except for ribose, lactate, succinate, and tryptophan.