Table 1.

The 18 articles focused on the relevance and discrimination between Chinese medicinal properties and chemical compositions.

First author	Year	Region	Research objects	Research methods	Main results or conclusions	Statistical methods
Chen and Chen [6]	1993	Zhejiang	414 CMM with various flavors and traits	Analyze the material basis and action principle of the functions for five flavors	The five flavors have an inevitable close relationship with the active ingredients of CMM	Intuitive analysis
Hu [7]	1996	Hunan	92 main active ingredients of CMM	Analyze the relationship between the CMP and the molecular weight of major components of CMM	The CMP change regularly with the molecular weight of the main contained active ingredients	Intuitive analysis
Fan and Wang [8]	2008	Beijing	247 vine Chinese medicine data	Correlation analysis of four gas properties and chemical components in Rattan herbs	There is a certain correlation between the four gas properties of Rattan herbs and the chemical constituents and drug efficacy contained in drugs	Correlation analysis
Yao et al. [9]	2008	Chengdu	Herbs for inducing resuscitation	Analyze the working mechanism of volatile components and their correlations with CMP	There is a correlation between the CMP of aromatic herbs and their material basis and pharmacological effects	Intuitive analysis
Yang [10]	2010	Shandong	The chemical compositions of 100 botanicals	Logistic regression equations were established at the four levels of organic composition categories, subcategories, substructural groups, and single compounds to discriminate the cold and heat properties of CMM	There is a correlation between the cold and hot CMP and the chemical constituents of CMM. The discriminative model of cold and hot CMP was established based on the chemical components. The cold and hot CMP can be predicted and discriminated	Logistic regression analysis, rank sum test, principal component analysis, cluster analysis, Fisher analysis, Bayesian analysis, and support vector machine
Shuai et al. [11]	2010	Shandong	The protein content of 50 Chinese herbs determined by the Coomassie brilliant blue method	Measuring the amount of protein in 25 cold, 25-flavored CMM	The average protein content of hot CMM was 2.37 times that of cold CMM, indicating that the protein content had a certain correlation with the cold and hot CMP	t-test
Zhou et al. [12]	2010	Shandong	The water-soluble sugars of 10 cold and 10 hot CMM determined by GC/MS fingerprinting	Fisher method to establish discriminant function	Water-soluble sugar is one of the material bases of cold and hot CMP	Fisher method
Long et al. [13]	2011	Tianjin	284 CMM with cold/hot properties and their chemical components	A novel strategy, weight center treatment was used to solve the problem that the chemical description was unable to be applied to CMM	The accuracy of 83.3% and 81.0% for the training and the test set indicate that this system is a useful tool to predict the property of unidentified folk herbs and foreign herbs	Support vector machine, descriptors generation, and weight center calculation
Liang et al. [14]	2013	Beijing	20 CMM with cold (10) and hot (10) properties and their active target proteins and main compounds	The active target proteins were analyzed to find out the property-related biological activities, and the main compounds were analyzed to decipher the properties	Cold-propertied CMM show intensive toxicity in the heart, which is likely to be correlated with the specific chemical fragments constructions, such as long chain alkenes, benzoheterocycle, and azotic heterocycle	Bioinformatics approach
Qin et al. [15]	2013	Guangxi	28 herbs that release the exterior and their volatile ingredients	Cross-train the data with support vector machines and establish a predictive model for the CMP	There is a high correlation between the volatile components and the cold and hot CMP	Support vector machines
Liu et al. [16]	2015	Nanjing	90 CMM with diuresis efficacy	Ancient and modern literature was employed to construct database. Data mining technology was used to analyze the intrinsic relation between CMP and effective components	The CMM attributed in upper Jiao major contain flavonoids and followed with terpenoids, in middle Jiao major contain terpenoids and followed with steroidal, and in lower Jiao major contain terpenoids and followed with flavonoids	Data mining, association rule
Yang et al. [17]	2015	Nanjing	34 kinds of volatile oils with transdermal penetration-promoting effects	Using frequency analysis, variable cross-tabulation for visual analysis and correlation analysis	There are clear correlation and regularity between the transdermal and permeation-promoting effects of essential oils and the CMP	Intuitive analysis, correlation analysis, and mathematical modeling
Jia et al. [18]	2015	Beijing	157 asteraceae CMM and their chemical constituents	Two-factor correlation analysis method	The main chemical constituents contained in cold, warm, and neutral CMM are volatile oils, flavonoids, terpenes, alkaloids, and acids	Correlation
Jiang et al. [19]	2016	Nanjing	18 kinds of volatile oil	Logistic regression analysis	There is a clear correlation and regularity between the CMP and the composition of volatile oils	Logistic regression analysis
Wang et al. [20]	2018	Shenyang	416 CMM attributed to heart, liver, and lung channel and their chemical constituent, pharmacological effect, and clinical application	Frequency analysis with 5% as the cut point. The chemical constituents, pharmacological effects, and clinical applications were analyzed with association methods, and the results were selected with support degree >5% and confidence >40%	It has explicit interrelation of channel tropism traditional Chinese medicine in chemical constituent, pharmacological effects, and clinical application	Frequency statistics, corresponding analysis, and association rules
Jiang et al. [21]	2019	Shenyang	11 Yang-tonifying herbs	The empirical regression equation was constructed to explore the tissue distribution of the receptors in the training set, and the criterion for determining whether herbs distribute to kidney meridian was established	This study explored a new method for judging whether CMM distributes to kidney meridian, established an effective criterion model, and verified the reliability of the new method	Empirical regression, neural network
Wenguo et al. [22]	2019	Nanjing	20 kinds of essential oils with different “four natures” drug properties	Essential oils were extracted by steam distillation and analyzed by GC-MS, and the skin resistance kinetic technology was used to investigate the abilities of penetration enhancement. The factors were selected by the stepwise discrimination and variance analysis method	There was a correlation among the “four natures” drug properties, the abilities of penetration enhancement, and chemical components of essential oils from pungent Chinese herbs	Stepwise discrimination analysis method, and variance analysis method
Dai and Xue [23]	2019	Leshan	9 kinds of essential oils with different “four natures” drug properties	The volatility method was used to analyze the different volatile oil components. The relationship between volatile oil and CMP was analyzed by logistic regression analysis	The different types of volatile oils are related to the four natures and five flavors of CMP. The transdermal penetration promoting effect of volatile oil has certain correlation and regularity	Logistic regression