Unraveling the physicochemical differences among Osborne protein classes via bioinformatics and AI

Authors: KWON, HYUKJIN - Kansas State University; Xu, Yixiang; XU, XUAN - Kansas State University; LI, YONGHUI - Kansas State University

Submitted to: Food Research International
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/10/2025
Publication Date: 8/13/2025
Citation: Kwon, H., Xu, Y., Xu, X., Li, Y. 2025. Unraveling the physicochemical differences among Osborne protein classes via bioinformatics and AI. Food Research International. 221. Article 117322. https://doi.org/10.1016/j.foodres.2025.117322.
DOI: https://doi.org/10.1016/j.foodres.2025.117322

Interpretive Summary: Seed storage proteins (SSPs) play a pivotal role in seed development and serve as key dietary protein sources for humans. A comprehensive understanding of their structures and molecular characteristics is essential, not only for elucidating their biological functions but also for optimizing protein extraction, processing, and functionality in food applications. Currently, the most widely used classification system for SSPs is solubility-based, introduced by Osborne. However, this method has limitations, particularly in its inability to capture the full structural and functional diversity of SSPs. In the past five years, significant advancements in artificial intelligence (AI) and computational biology have led to the development of new algorithms and technologies. Therefore, the USDA scientist collaborating with the researchers at Kansas State University utilized AI and bioinformatics techniques to integrate sequence, structural, and physicochemical data to refine our understanding of SSP classification. A dataset of 1,039 SSPs from 215 species was used, and structural-based features were extracted and compared to identify class-specific characteristics. Accurate identifying distinctive physicochemical features defining each class not only to enhance our knowledge of SSP classification but also to improve its application in food processing and functional optimization. This finding will therefore benefit protein scientists, food processors, consumers who are interested in plant-based food products with desired functional properties.

Technical Abstract: Osborne fractionation remains a cornerstone in food science for categorizing seed storage proteins (SSPs), yet molecular distinctions among the classes remain unclear. This study employs a computational framework integrating structural modeling, AI (artificial intelligence)-driven classification, and molecular dynamics (MD) simulations to elucidate these underlying physicochemical differences. Using a dataset of 1,039 SSPs from 215 species, sequence and structural-based features were extracted and compared to identify class-specific characteristics, such as low abundance of charged residues in prolamins. Machine learning classifiers, including binary support vector machines and graph convolutional networks were trained on these features, achieving validation and test accuracies ranging from 96.0% to 100.0%. This study integrates modern computational approaches to provide molecular insights into the distinct properties of each Osborne class and their solubility under different solvent conditions. The proposed framework and key findings not only advance our understanding of seed storage protein behavior but also have valuable implications for food science.