基于专利数据挖掘的糖尿病药食同源方组方规律研究OA

Objective To analyze the formulation and medication rules of domestic patent formulas for diabetic medicinal diets using data mining technology,and to provide evidence for the development of food-medicine homologous products for diabetes.Methods Relevant invention patents were retrieved from the patent database of the China National Intellectual Property Administration.Frequency statistics,association rule analysis,and cluster analysis were employed to mine the data on the frequency of TCM usage,their properties,flavors,meridian affinity,and core combinations.Results A total of 110 valid patents were included,involving 234 TCMs.The high-frequency core medicines were Shanyao(Dioscoreae Rhizoma,47.27％),Gegen(Puerariae Lobatae Radix,47.27％),Gouqizi(Lycii Fructus,36.36％),Fuling(Poria,32.73％),Huangjing(Polygonati Rhizoma,32.73％)and Sangye(Mori Folium).The dominant medicinal property was neutral(348 occurrences),and the dominant flavor was sweet(788 occurrences).The primary meridian affinity were the lung(516 occurrences),spleen(498 occurrences),and kidney(371 occurrences)meridians.Association rule analysis revealed that"Shanyao(Dioscoreae Rhizoma)-Gegen(Puerariae Lobatae Radix)"were the core medicinal combinations.Cluster analysis categorized the high-frequency medicines into 5 groups.Conclusion The composition of the patented food-medicine homologous formulas for diabetes closely follows the core pathogenesis of Xiaoke disease(diabetes),namely"qi and yin deficiency with internal accumulation of dryness-heat,"and reflects the compatibility characteristic of balancing cold and heat.The findings of this study can provide a basis for evidence-based research and development in the food-medicine homologous industry for diabetes.