Correlation of structure and function in native starches assisted by principal component analysis.

BACKGROUND: Understanding how properties such as amylose/amylopectin ratio, granule morphology, crystallinity, molecular weight, and branching degree dictate functionality is essential for optimizing starch applications in food and biomaterials. Principal component analysis (PCA) is a multivariate tool that integrates physicochemical, structural, and functional variables, reducing dimensionality while preserving variability and enabling visualization of similarities among starches. RESULTS: This work applies PCA to characterize native starches from six botanical origins - corn, cassava, plantain, avocado, achira, and balú - using integrated physicochemical, thermal, and rheological analyses. Structural data revealed A- and C-type crystallinity in corn and plantain, and B-type patterns in achira, avocado, and balú, influenced by the interplay between molecular weight and branching degree. Thermal and pasting profiles indicated that no single factor governs starch performance, which depends instead on composition, morphology, and crystalline organization. PCA integrated 24 variables, explaining over 70% of the total variance and distinguishing clusters associated with specific applications from strong gels (achira, balú) to flexible, low-viscosity systems (avocado). CONCLUSION: These findings validate PCA as a predictive tool for linking starch structure to functionality and highlight the potential of underutilized sources for sustainable food and material applications. The resulting clustering reflected meaningful functional differences associated with botanical origin and technological performance, providing an objective framework for classifying native starches according to their application potential. © 2026 Society of Chemical Industry.