Partial Least Squares Regression for Studying the Texture Characteristics of Tibetan Chicken Breast Meat

In order to reveal the unique texture characteristics of Tibetan chicken breast meat and their correlation with chemical composition and microstructure, this study tested the shear force, texture profile analysis (TPA) parameters and stress relaxation characteristics of breast meat from Tibetan chicken, and Xueshan chicken, spent hen and white feathered broilers with the same age or body mass as Tibetan chicken, and it also analyzed the correlation between the texture properties and chemical properties (moisture content, fat content, protein content, collagen content, and pH) as well as histological characteristics (muscle fiber diameter and density). Meanwhile, partial least squares regression (PLSR) was used to explore the significant factors that affect the texture characteristics of chicken meat. Our results indicated that PLSR could quantify various chemical indices related to chicken meat texture as comprehensive quality indices. Shear force was more suitable to establish a model to predict chicken meat texture properties. The R_X² and R_Y² of the proposed model were 0.621 and 0.739, respectively, suggesting good fitness. The Q_cum² of the model was above 0.6, showing acceptable prediction accuracy. The loading plot showed that shear force was positively correlated with pH, total collagen content and muscle fiber diameter significantly (P < 0.05), but was negatively correlated with fat content, muscle fiber density, soluble collagen content and collagen solubility significantly (P < 0.05). Collagen solubility, muscle fiber density, soluble collagen content and muscle fiber diameter all contributed significantly to the model. These findings provide data support for revealing the unique “chewiness” of Tibetan chicken meat from the perspective of its texture, and a theoretical basis to enrich research on meat mechanical properties.