Exploring bioactive components of sugar bean with some data mining algorithms: relationships between antioxidants, enzyme inhibitors, and minerals

Sugar bean (Phaseolus vulgaris L.) is a legume of notable nutritional and functional value, and its genotypes have substantial variation in bioactive compounds, antioxidant activity, mineral composition, and enzyme inhibition properties. This study aimed to comprehensively evaluate the biochemical and functional diversity of 10 sugar bean genotypes using data mining algorithms to identify superior candidates for functional food and nutraceutical development. Associations between the phytochemical profiles of sugar bean genotypes and their bioactivity, antioxidant capacity, and enzyme inhibition were modeled with classification and regression trees (CART), random forests (RF), and multivariate adaptive regression splines (MARS), while one-way ANOVA quantified varietal effects on these responses. Data mining algorithms, particularly RF and MARS, provided accurate and robust predictions for the relationships between biological, antioxidant, and phytochemical properties of genotypes; however, the predictive power for enzyme inhibition was limited. Statistically significant differences (p < 0.01) were observed among genotypes in extract yield (5.92–7.92%), protein ratio (23.92– 28.70%), total phenolic content (up to 12.28 mg GAE/g), and flavonoid content (up to 0.48 mg QE/g). The Mor Toparlak genotype had superior total phenolic content, antioxidant activity, and acetylcholinesterase inhibitory effects. The Mor Toparlak, Asmaçam, Akçabelen, and Beyşehir genotypes showed the highest α-amylase and α-glucosidase inhibitory activities. The findings showed the significant diversity in nutritional and functional properties among sugar bean genotypes and emphasized which phytochemical traits and metabolic factors can be associated with biological activity, antioxidant activity, and enzyme inhibition. Sugar bean genotypes with high antioxidant capacity, rich mineral content, significant protein levels, and notable enzyme inhibition potential are of great importance for human nutrition and health. In particular, the Mor Toparlak, Kozak, Asmaçam, Akçabelen, and Beyşehir genotypes stand out as valuable raw materials for the development of functional foods and breeding programs aimed at enhancing health benefits, as well as for the creation of health-promoting, palatable, and high-quality genotypes.