Study on the material basis of cold and hot properties of traditional Chinese medicines in Lamiaceae based on pattern recognition
10.16438/j.0513-4870.2022-1100
- VernacularTitle:基于模式识别的唇形科中药寒热药性物质基础研究
- Author:
Chuan-yao ZHANG
1
;
Qing ZHOU
2
;
Xiao WANG
1
;
Fei-fei BU
1
;
Peng WANG
3
Author Information
1. School of Traditional Chinese Medicine, Anhui University of Chinese Medicine, Hefei 230012, China
2. School of Medical Information Engineering, Anhui University of Chinese Medicine, Hefei 230012, China
3. Graduate School, Anhui University of Chinese Medicine, Hefei 230012, China
- Publication Type:Research Article
- Keywords:
pattern recognition;
Lamiaceae;
traditional Chinese medicine;
cold and hot property;
material basis;
identification model
- From:
Acta Pharmaceutica Sinica
2023;58(2):429-438
- CountryChina
- Language:Chinese
-
Abstract:
To study the material basis of cold and hot properties of traditional Chinese medicines (TCMs) in Lamiaceae and to establish a cold and hot properties identification model, a database of material components of TCMs in Lamiaceae was established. A three-level classification system of material components was used to obtain the material basis of cold and hot properties of the Lamiaceae family by using data mining methods such as frequency analysis, association rule analysis, logistic regression, and feature selection. Several identification models were established to recognize the cold and hot properties. The chi-square test results showed that the material composition ratios of cold and hot properties were significantly different at the first-level, second-level, and third-level classification (P < 0.05), and the differences varied as the levels of substance classification changed. The average coefficients of variation were 42.30%, 79.07%, and 91.51% at the first-level, second-level, and third-level classification levels, respectively. In other words, in terms of the percentage differences in material composition ratio, the first-level was smaller than the second-level, and the second-level was smaller than the third-level. The results of the association rule analysis showed that under the third-level classification, there were many effective association rules, and 27 core groups and 34 specific groups of chemical components were obtained based on these rules. 15 decisive groups were obtained from the feature selection results. Multinomial logistic regression analysis was used to successfully establish a cold and hot properties identification model with an overall accuracy of 89%. The material basis of cold and hot properties of TCMs in Lamiaceae is different and intersect with each other. Twenty-seven groups of chemical components, such as bicyclic diterpenes, are the core groups of cold and hot properties, of which 15 groups are the decisive groups. The cold and hot properties are often characterized by the interaction of multiple classes of substances, and a single class of substances often cannot be used to characterize the properties. The organic combination of multiple classes of substances is the material basis of cold and hot properties.
