Predictive model for resistant starch (RS) in rice (Oryza sativa L.) based on digestibility factors

Altering the digestibility of starch to enhance the resistance to digestion, specifically targeting the creation of resistant starch (RS), holds significant importance in the fields of agriculture, food, and nutrition. This modification not only restricts the amplitude of glycemic response but also promotes gut health. Traditionally, quantifying RS has relied on complex, time-consuming, and costly human digestion simulation assays. Acknowledging the association between starch digestibility and various factors, such as microstructure, gelatinization temperature, total starch (TS), total amylose, and amylopectin, this study aimed to establish a fundamental relationship among these explanatory variables and RS through the development of a forecasting model known as the starch quality matrix (SQM). Constructed using Pearson's correlation, the SQM proved to be significant based on model statistics, and the regression model's adequacy was confirmed through residual diagnosis. Notably, both TS and total amylose content exhibited a significantly positive impact on RS, with coefficients of 0.030 and 0.024, respectively. Model validation utilized root mean square error and mean absolute error. The correlation between RS and inherent glycemic potential was further verified through in-house developed in-vitro starch hydrolyzation kinetics. This study unveils, for the first time, a perspective on the relationship influencing starch digestibility and introduces the SQM tool. This tool is poised to facilitate future efforts in breeding high-RS rice varieties with a low glycemic index.