Key points are not available for this paper at this time.
Over the past few decades, significant progress has been made in the research of polysaccharides extracted from natural resources, which are often used in functional foods, medicines, cosmetics, and biomedical materials. However, traditional research heavily relies on trial-and-error screening, which is limited by challenges in elucidating structure-function relationships, low preparation efficiency, and poor application adaptability. The integration of artificial intelligence (AI) has provided a critical pathway to overcome these constraints. This review outlines recent AI applications in polysaccharide research, discusses current challenges, and identifies future trends. For polysaccharide extraction, AI employs models such as artificial neural networks and genetic algorithm-backpropagation to optimize processing conditions. Its prediction accuracy often reaches above 0.95, significantly higher than the 0.7-0.8 of traditional response surface methodology models. In practical applications, AI integrates multi-omics data to support personalized polysaccharide scheme design. For instance, graph convolutional networks can correlate structural features with biological activities (e.g., tumor cell inhibition rates and immune cell activation), thereby promoting the development of personalized functional products. However, the field still faces challenges such as inconsistent data quality, limited model interpretability, and difficulties in cross-disciplinary collaboration. Solving these problems is key to advancing AI from a supportive tool to a central driver of innovation in polysaccharide research, with potential impacts on precision medicine, functional foods, and advanced biomaterials.
Zhang et al. (Sun,) studied this question.