Decoding and reconstructing yeast protein flavor based on an integrated sensory-omics approach

To enhance the sensory quality of yeast proteins and facilitate their broader application in food products, its key aroma compounds were systematically characterized through sensory evaluation and GC × GC-TOF-MS. The combined use of three extraction techniques (SAFE, SDE, and SPE) enabled the identification of 104 odor-active compounds. Sensory profiling revealed the roasted, sour, sweet, and almond-like were dominant. Among the three extraction methods, SAFE and SDE were optimal for bulk recovery, while SPE targeted organic acids. The aroma recombination experiments confirmed the recombination model based on the three methods demonstrates a relatively high similarity (90%) with the protein sample. The omission experiments further confirmed that benzaldehyde, n -decanoic acid, hexanoic acid, methyl benzoate, geranylacetone, 1-nonanol, 2-decanol, 2-pentadecanone, farnesol, γ -undecalactone, and octanal were identified as key aroma compounds in yeast proteins. The work establishes a robust analytical framework for yeast protein flavor characterization, aiding product development and quality control.