Quantitative Assessment of Sourness Intensity in purpleback squid (Symlectoteuthis oualaniensis) from Indian Ocean based on PCA and MVDA models

Abstract

The occurrence of sour taste with off-flavor in Symlectoteuthis oualaniensis during storage and transportation is a common issue. This current study aimed to quantitatively identify the intensity of sour taste in freeze-stored S. oualaniensis by employing partial least squares regression (PLSR) and Bayesian discriminant model for the relationship between chemically analyzed sourness-related substances and sour taste perception. Principal component analysis was used to classify all tested S. oualaniensis into three groups, accounting for over 90% of the overall variance. It was found that opines, specifically β-alanopine, tauropine, and meso-alanopine, were the main contributors to the sour taste of S. oualaniensis. Discriminant models were successfully developed for the quantification of sour taste with the explanation of 85.0%−103% to the sour taste in squid. The recognition accuracies of self and cross validation results were 100% and 81.0%, respectively. The overall recognition accuracy of the training sample set was 83.3%. These findings suggest that the discriminant approach, combining principal component analysis and multivariate discriminant analysis, can serve as a valuable tool for both qualitative and quantitative analysis of the sour taste intensity in S. oualaniensis from Indian Ocean.