GC × GC-ToF-MS combined with multivariate statistical methods to explore the effects of L. paracasei fermentation on bread flavor characteristics.

GC × GC-ToF-MS is increasingly used to analyze complex food flavors due to its high resolution and sensitivity, but few studies have used the method to identify aroma components of bread. For the first time, this study combines GC × GC-ToF-MS and multivariate statistical methods to explore the effects of L. paracasei fermentation on bread flavor characteristics. A total of 1534 volatile organic compounds were identified, of which 447 were obtained by metabolome normalization. Based on the variable importance for the projection and p values, 82 different compounds were screened in L. paracasei bread compared with yeast bread, and the total relative content was 1.52 times higher than that of yeast bread. 2-Furancarboxaldehyde, 5-methyl-, pentanoic acid, 2-hydroxy-4-methyl-, ethyl ester, pyrazine, 2,5-dimethyl- and γ-terpinene are aroma-presenting substances specific to L. paracasei bread that could be potential identification compounds. This study provides a new techno-theoretical approach for the characterization and discrimination of LAB bread flavors.