highly significantP < 0.001
The significance of omission experiments was according to sensory analysis dictionary: 9 or 10 of ten panelists could distinguish the omission model correctly was defined as very highly significant ( P < 0.001); 8 and 7 of ten panelists could distinguish the omission model correctly were defined as highly significant ( P < 0.01) and significant ( P < 0.05), respectively; 6 or less panelists could distinguish the omission model correctly was defined as not significant ( P > 0.05). 2.12 Statistical analysis All experiments and samples were performed in triplicate, and the results were presented as the mean ± SD (standard deviation).