Author: Yili Hu, Huyun Dong
I. Introduction
1.1. Objective
In this study, we used data from 188 Chinese herbs prescriptions for local prevention of COVID-19, SARS, and H1N1: three representative viral respiratory infectious diseases released or announced by National Administration of Traditional Chinese Medicine, provincial and municipal health commissions, some hospitals, and Chinese and American experts on traditional Chinese medicine from April 2003 to March 2020. We applied complex networks based on relevant statistical theories to explore respective core herbs and their association strengths of Chinese herbs for the prevention of the three types of typical viral respiratory infectious diseases including COVID-19, SARS, and H1N1. Meanwhile, the core prescriptions of Chinese herbs for the prevention of each type of viral respiratory infectious diseases were analyzed, and core prescriptions of the broad-spectrum Chinese herbs commonly used for the prevention of COVID-19, SARS and H1N1 and other viral respiratory infectious diseases were further analyzed, so as to provide guidance and data basis for the research and development of new products of traditional Chinese medicine for mass prevention of viral respiratory infectious diseases.
1.2 Data Description
Data was collected by China Changxing Weifan Traditional Chinese Medicine Technology and Culture Exchange Co., Ltd. To solve the problems of non-unified names of traditional Chinese medicine for various organizations, medical institutions, and traditional Chinese medicine experts, the standardization processing was carried out for the Chinese traditional medicine names in the related prescriptions, and the standard database for prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases was established in this study according to ChineseTraditionalMedicine and Materia Medica Subject Headings, Selected Edition of the Dictionary of Traditional Chinese Medicine Prescriptions, Chinese Pharmacopoeia, etc. Meanwhile, the data format of relevant traditional Chinese medicine was transformed into data that could be processed by software.
Based on the given standard database, we conducted descriptive analysis on the number of prescriptions of Chinese herbs and the number of Chinese herbs for the prevention of three types of viral respiratory infectious diseases including COVID-19, SARS and H1N1 using Python3.6 and Origin8.5, as shown in Figure 1.1 and 1.2:
The number of prescriptions for the prevention of COVID-19, H1N1, and SARS was 89, 46, and 53, respectively, which ensured sufficient data size in this study and laid a solid data foundation for subsequent study and analysis. The number of Chinese herbs for the prevention of COVID-19, H1N1, and SARS was 103, 68, and 57 respectively. It was shown in the study that the number of Chinese herbs used to treat COVID-19 was 51.47% higher than that of H1N1 and 80.7% higher than that of SARS. The number of Chinese herbs used for COVID-19, H1N1, and SARS was 807, 406, and 491 respectively. It could be concluded that a wide range of Chinese herbs was used for the prevention of COVID-19. Furthermore, It was shown that 2019-nCoV was more complex than SARS and H1N1 virus, with stronger virus variability and more serious harm to human beings, which was of more urgent and great significance to the exploration and research of effective prevention methods and the development of new traditional Chinese medicine products.
1.3 Method
(1) Construction of weighted association network for the prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases
First, the M single Chinese herbs involved in N prescriptions of Chinese herbs in the database were set as the network nodes, and the weight of two different Chinese herbs was connected to represent the medication frequency of the two Chinese herbs in multiple compounds. The weighted association network of prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases was constructed according to the nodes and their association of single Chinese herbs identified in all the prescriptions of Chinese herbs. The total number of edges connected between the nodes is called the strength of association, also known as the strength of nodes. The strength of the nodes is one of the judgment bases for describing the importance (centrality) of nodes for herbal medicines in a prescription in the network analysis. The higher the strength of the node is, the more compatibility the node has with other nodes; and the higher the strength centrality of the node is, the more important it is in the network.
(2) Analysis of the prescriptions of the core Chinese herbs for the prevention of viral respiratory infectious diseases
The prescription data in the database of prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases were analyzed based on the relevant theoretical knowledge such as complex network and biostatistics. The analysis results were visualized using Gephi 9.2, in which the nodes of the traditional Chinese medicine were represented by the names of the Chinese herbs, and the core Chinese herbs were located at the center of the Figure.
(3) Analysis of association degree of core Chinese herbs for the prevention of viral respiratory tract infectious diseases
In accordance with the complex network analysis, the compatibility between Chinese herbs was further analyzed. The weight of the edges connecting two different Chinese herbs indicated the strength of the simultaneous use of the two Chinese herbs in different prescriptions. The darker the color of the connected sideline was in the network diagram, the closer and stronger the association was between two different Chinese herbs.'
(4) Based on the analysis on the core Chinese herbs for the prevention of three types of representative viral respiratory infectious diseases including COVID-19, H1N1, and SARS, a broad-spectrum core prescription for the prevention of viral respiratory infectious diseases was finally obtained through the application of the relevant theoretical knowledge of statistics and the complex network.
II. Analysis of the Study of the Core Prescription for the Prevention of Typical Viral Respiratory Infectious Diseases
2.1 COVID-19
(1) Statistical analysis of the frequency of Chinese herbs
In this section, medication frequency of 103 Chinese herbs in 89 prescriptions of Chinese herbs for the prevention of COVID-19 was statistically analyzed, and the results were shown in Figure 2.1. The top 10 herbs among the 103 Chinese herbs were as follows: Astragali Radix, Licorice Root, TrolliusChinensis Bunge, Giant Knotweed, Reed Rhizome, Patchouli, Aged Tangerine Peel, Forsythia Fruit, CentellaeHerba and Saposhnikoviae Radix. Since each Chinese herb was used only once per prescription, approximately 25.24 % of the Chinese herbs could be seen in more than 10% of the prescriptions.
(2) Analysis of weighted association network of prescriptions of Chinese herbs for the prevention of COVID-19
Through the application of the complex network technology and standard data of prescriptions of Chinese herbs for the prevention of COVID-19, the weighted association network of the prescription of Chinese herbs for the prevention of COVID-19 was constructed and analyzed based on the method of constructing the weighted association network for the prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases in Section 1.3, as shown in Figure 2.2. The importance degree of the nodes (Chinese herbs) was represented by the color depth; the darker the color was, the more important it was. The strength of the association between Chinese herbs was represented by the thickness and color of the edges, which were positively correlated.
In order to ensure the accuracy of the core prescription extracted, the node strength of the weighted association network of the prescriptions of Chinese herbs for the prevention of COVID-19 was analyzed in this study, as shown in Figure 2.3. The strength of the nodes in this network complied with the power-law distribution, indicating that the nodes in this network conformed to the “80/20 Rule”. Namely, small number of Chinese herbs can be seen in most prescriptions, while most Chinese herbs can be observed in very few prescriptions. This suggested the presence of key core Chinese herbs in multiple prescriptions for the prevention of COVID-19, ensuring the effectiveness of the study. Next, we would study and analyze the strength of the association between the core prescription and the Chinese herbs.
1) Extraction of the core prescription
We used the strength of nodes to depict the importance degree of Chinese herbs in this study. Besides, based on this, we extracted the core Chinese herbs from the core prescription for the prevention of COVID-19. It could be concluded that the core Chinese herbs of the prescription were as follows: Astragali Radix, Licorice Root, Giant Knotweed, Trollius Chinensis Bunge, Patchouli, Reed Rhizome, Aged Tangerine Peel, Forsythia Fruit, Centellae Herba, Lonicerae Flos, Siberian Ginseng, Ganoderma, Underleaf Pearl, Saposhnikoviae Radix, Isatis Leaf and Platycodonis Radix.
2) Analysis on strength of association of Chinese herbs
We used the weight of connecting edges for the network between Chinese herbs to describe the strength of association (positive correlation) in the process of analysis on the strength of association between Chinese herbs, as shown in Figure 2.4. The thicker and darker the edge lines in the network diagram were, the closer the association degree between the two different Chinese herbs was. It can be seen that Astragali Radix and Licorice Root were most closely associated, followed by Centellae Herba and the Giant Knotweed, and Astragali Radix and the Saposhnikoviae Radix. The analysis results of compatibility of Chinese herbs with strength of association greater than 18 were shown in Table 2.1.
2.2 H1N1
(1) Statistical analysis of the frequency of Chinese herbs
In this section, statistical analysis was made on the medication frequency of 68 Chinese herbs in 46 prescriptions of Chinese herbs for the prevention of H1N1. The results were shown in Figure 2.5. The top 10 herbs among the 103 Chinese herbs were as follows: Isatis Leaf, Licorice Root, Forsythia Fruit, Woad Root, Trollius Chinensis Bunge, Astragali Radix, Lonicerae Flos, Aged Tangerine Peel, Reed Rhizome, and Patchouli. Since each Chinese herb was used only once per prescription, approximately 35.29% of the Chinese herbs could be seen in more than 10% of the prescriptions.
(2) Analysis of weighted association network of the prescription of Chinese herbs for the prevention of H1N1
Through the application of the complex network technique and standard data of prescriptions of Chinese herbs for the prevention of H1N1, the weighted association network of the prescription of Chinese herbs for the prevention of H1N1 was constructed and analyzed based on the method of constructing the weighted association network for the prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases in Section 1.3, as shown in Figure 2.6. The importance degree of the nodes (Chinese herbs) was represented by the color depth; the darker the color was, the more important it was. The strength of the association between Chinese herbs was represented by the thickness and color of the edges, which were positively correlated.
In order to ensure the accuracy and scientificity of the core prescription extracted, the node strength of the weighted association network of the prescriptions of Chinese herbs for the prevention of H1N1 was analyzed in this study, as shown in Figure 2.7. The strength of the nodes in this network complied with the power-law distribution, indicating that the nodes in this network conformed to the “80/20 Rule”. Namely, small number of Chinese herbs can be seen in most prescriptions, while most Chinese herbs can be observed in very few prescriptions. This suggested the presence of key core Chinese herbs in multiple prescriptions for the prevention of H1N1, ensuring the effectiveness of the study. Next, we would study and analyze the strength of the association between the core prescription and the Chinese herbs.
1) Extraction of the core prescription
We used the strength of nodes to depict the importance degree of Chinese herbs in this study. Besides, based on this, we extracted the core Chinese herbs from the core prescription for the prevention of H1N1. It could be concluded that the core Chinese herbs of the prescription were as follows: Licorice Root, Isatis Leaf, Forsythia Fruit, TrolliusChinensis Bunge, Woad Root, Astragali Radix, Aged Tangerine Peel, Lonicerae Flos, Patchouli, Reed Rhizome, Giant Knotweed, Siberian Ginseng, Ganoderma, ZingiberisRhizomaRecens, Underleaf Pearl, Centellae Herba, and Arctii Fructus.
2) Analysis on strength of association of Chinese herbs
We used the weight of connecting edges for the network between Chinese herbs to describe the strength of association (positive correlation) in the process of analysis on the strength of association between Chinese herbs, as shown in Figure 2.8. The thicker and darker of the edge lines in the network diagram were, the closer the association degree between the two different Chinese herbs was. It can be seen that Licorice RootandIsatisLeafwere most closely associated, followed by Woad Root and the Isatis Leaf, and Isatis Leaf and the Trollius Chinensis Bunge. The analysis results of compatibility of Chinese herbs with strength of association greater than 10 were shown in Table 2.2.
2.3 SARS
(1) Statistical analysis of the frequency of Chinese herbs
In this section, statistical analysis was made on the medication frequency of 57 Chinese herbs in 53 prescriptions of Chinese herbs for the prevention of SARS. The results were shown in Figure 2.9. The top 10 herbs among the 57 Chinese herbs were as follows: Woad Root, Licorice Root, Forsythia Fruit, Isatis Leaf, Lonicerae Flos, Giant Knotweed, Astragali Radix, Patchouli, TrolliusChinensis Bunge, and Aged Tangerine Peel. Since each Chinese herb was used only once per prescription, approximately 38.60% of the Chinese herbs could be seen in more than 10% of the prescriptions.
(2) Analysis of weighted association network of prescriptions of Chinese herbs for the prevention of SARS
Through the application of the complex network technology and standard data of prescriptions of Chinese herbs for the prevention of SARS, the weighted association network of the prescription of Chinese herbs for the prevention of SARS was constructed and analyzed based on the method of constructing the weighted association network for the prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases in Section 1.3, as shown in Figure 2.10. The importance degree of the nodes (Chinese herbs) was represented by the color depth; the darker the color was, the more important it was. The strength of the association between Chinese herbs was represented by the thickness and color of the edges, which were positively correlated.
In order to ensure the accuracy and scientificity of the core prescription extracted, the node strength of the weighted association network of the prescriptions of Chinese herbs for the prevention of SARS was analyzed in this study, as shown in Figure 2.11. The strength of the nodes in this network complied with the power-law distribution, indicating that the nodes in this network conformed to the “80/20 Rule”. Namely, small number of Chinese herbs can be seen in most prescriptions, while most Chinese herbs can be observed in very few prescriptions. This suggested the presence of key core Chinese herbs in multiple prescriptions for the prevention of SARS, ensuring the effectiveness of the study. Next, we would study and analyze the strength of the association between the core prescription and the Chinese herbs.
1) Extraction of the core prescription
We used the strength of nodes to depict the importance degree of Chinese herbs in this study. Besides, based on this, we extracted the core Chinese herbs from the core prescription for the prevention of SARS. It could be concluded that the core Chinese herbs of the prescription were as follows: Woad Root, Licorice Root, Forsythia Fruit, Isatis Leaf, LoniceraeFlos, Astragali Radix, Giant Knotweed, Patchouli, TrolliusChinensis Bunge, Reed Rhizome, Aged Tangerine Peel, Saposhnikoviae Radix, RhizomaAtractylodisMacrocephalae, Scutellariae Radix, Ganoderma, and Siberian Ginseng.
2) Analysis on strength of association of Chinese herbs
We used the weight of connecting edges for the network between Chinese herbs to describe the strength of association (positive correlation) in the process of analysis on the strength of association between Chinese herbs, as shown in Figure 2.12. The thicker and darker the edge lines in the network diagram were, the closer the association degree between the two different Chinese herbs was. It can be seen that Woad Root and Licorice Root were most closely associated, followed by Licorice Root and the Forsythia Fruit, and Woad Root and the LoniceraeFlos. The analysis results of compatibility of Chinese herbs with strength of association greater than 11 were shown in Table 2.3.
III. Analysis of the Study on Broad-spectrum Core Prescriptions for the Prevention of Viral Respiratory Infectious Diseases
3.1 Analysis of Broad-spectrum Core Prescriptions Based on the Single Disease
In this section, we further extracted abroad-spectrum core prescriptions for the prevention of viral respiratory infectious diseases based on the above data obtained from the analysis of COVID-19, H1N1, and SARS. The statistics on the number of types of core prescriptions for various Chinese herbs for the prevention of COVID-19, H1N1, and SARS was shown in Table 2.4. Based on this, Chinese herbal medicinal ingredients in the broad-spectrum core prescriptions for the prevention of viral respiratory infectious diseases were determined, and it could be concluded that the broad-spectrum core ingredients were 16 Chinese herbs in total including Astragali Radix, Ganoderma, Siberian Ginseng, LoniceraeFlos, TrolliusChinensis Bunge, Isatis Leaf, Patchouli, Forsythia Fruit, Aged Tangerine Peel, Giant Knotweed, Reed Rhizome, Licorice Root, Underleaf Pearl, Woad Root, and Centellae Herba.
3.2 Analysis of Broad-spectrum Core Prescriptions based on Combined Multi-diseases
(1) Statistical Analysis of the Frequency of Chinese Herbs for Combined Multi-diseases
In this section, statistical analysis was made on the medication frequency of 125 traditional Chinese herbs in 188 prescriptions of Chinese herbs for the prevention of viral respiratory tract infectious diseases, and the results were shown in Figure 2.13. The top 10 herbs among the 125 Chinese herbs are as follows: Licorice Root, Astragali Radix, Forsythia Fruit, Isatis Leaf, TrolliusChinensis Bunge, Giant Knotweed, Patchouli, LoniceraeFlos, Woad Root, Reed Rhizome, and Aged Tangerine Peel. Since each Chinese herb was used only once per prescription, approximately 20% of the Chinese herbs could be seen in more than 10% of the prescriptions.
(2) Analysis of weighted association network of prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases
We used the established standard database for prescriptions of Chinese herbs of viral respiratory infectious diseases to make a comprehensive analysis of viral respiratory infectious diseases. As shown in Figure 2.14, a weighted association network model of prescriptions of Chinese herbs for viral respiratory infectious diseases was constructed by taking prescriptions of Chinese herbs for the prevention of such diseases as the nodes, connecting edges based on whether two Chinese herbs were in the same prescription, and taking the medication frequency of two Chinese herbs in multiple prescriptions as the weight. The importance degree of the nodes (Chinese herbs) was represented by the color depth; the darker the color was, the more important it was. The strength of the association between Chinese herbs was represented by the thickness and color of the edges, which were positively correlated.
In order to ensure the accuracy and scientificity of the core prescription extracted, the node strength of the weighted association network of the prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases was analyzed in this study, as shown in Figure 2.15. The strength of the nodes in this network complied with the power-law distribution, indicating that the nodes in this network conformed to the “80/20 Rule”. Namely, small number of Chinese herbs can be seen in most prescriptions, while most Chinese herbs can be observed in very few prescriptions. This suggested the presence of key core Chinese herbs in multiple prescriptions for the prevention of viral respiratory infectious diseases, ensuring the effectiveness of the study. Next, we would study and analyze the strength of the association between the core prescriptions and traditional Chinese medicine.
1) Extraction of the core prescriptions
We used the strength of nodes to depict the importance degree of Chinese herbs in this study. Besides, based on this, we extracted the core Chinese herbs from the core prescription for the prevention of viral respiratory infectious diseases. It could be concluded that the core Chinese herbs of the prescription were as follows: Astragali Radix, Ganoderma, Siberian Ginseng, Underleaf Pearl, LoniceraeFlos, TrolliusChinensis Bunge, Isatis Leaf, Woad Root, Patchouli, Forsythia Fruit, CentellaeHerba, Aged Tangerine Peel, Giant Knotweed, Reed Rhizome, Licorice Root and Saposhnikoviae Radix.
2) Analysis on strength of association of Chinese herbs
We used the weight of connecting edges for the network between Chinese herbs to describe the strength of association (positive correlation) in the process of analysis on the strength of association between Chinese herbs, as shown in Figure 2.16. The thicker and darker the edge lines in the network diagram of the combined multi-diseases were, the closer the association degree between the two different Chinese herbs was. It can be seen that Forsythia Fruit and the Licorice Root were most closely associated, followed by Licorice Root and the Astragali Radix, and Licorice Root and the Isatis Leaf. The analysis results of compatibility of Chinese herbs with the strength of association greater than 38 were shown in Table 2.5.
Through study and analysis on the single disease and combined multi-diseases of three representative viral respiratory infectious diseases in this study from local and the global perspectives, the broad-spectrum core prescriptions for the prevention of viral respiratory infectious diseases were extracted respectively, as shown in Table 2.6. It was shown that the importance levels of some Chinese herbs in the broad-spectrum core prescriptions extracted by the two methods were different to some extent, but there was high consistency in Chinese herbal medicinal ingredients. Therefore, it indicates that the broad-spectrum core prescriptions for the prevention of viral respiratory infectious diseases extracted by complex network and biostatistics in the study are of high accuracy and scientificity, which provide scientific guidance and strong guarantee for the research and development of traditional Chinese medicine products for the prevention of mass viral respiratory infectious diseases.
IV. Conclusion
Furthermore, the strength of nodes (Chinese herbs) was selected as the index to study, analyze, and extract the core prescription. Meanwhile, the weights connecting edges of the network were selected as the index to study and analyze the strength of association between the Chinese herbs. Finally, the broad-spectrum core prescriptions of Chinese herbs for the prevention of viral respiratory infectious diseases were deeply explored through the application of the knowledge relevant to statistics, so as to provide data reference for the active prevention of viral respiratory infectious diseases. In this way, it can provide more scientific thinking and basis for TCM experts to make prescriptions of Chinese herbs for prevention, so as to better play the advantages of traditional Chinese medicine in the prevention of viral respiratory infectious diseases. Especially, they can be used as convenient and inexpensive new Chinese medicine products for the susceptible population during the epidemic or outbreak of viral respiratory tract infectious diseases.
There were only a few disease types in this study since study and analysis in this study were only based on the data of three representative diseases including COVID-19, SARS, and H1N1. In the follow-up study, it is expected that there will be large sample-size and high-quality data for further study and analysis to further confirm the results of this study. Meanwhile, the feasibility and accuracy of the results of this study will be further confirmed through the design of precise and scientific clinical trials.
Comentários