A network-based systematic study for the mechanism of the treatment of zhengs related to cough variant asthma

Abstract

Traditional Chinese medicine (TCM) has shown significant efficacy in the treatment of cough variant asthma (CVA), a special type of asthma. However, there is shortage of explanations for relevant mechanism of treatment. As Zhengs differentiation is a critical concept in TCM, it is necessary to explain the mechanism of treatment of Zhengs. Based on TCM clinical cases, this study illustrated the mechanism of the treatment of three remarkably relevant Zhengs for CVA: "FengXieFanFei," "FeiQiShiXuan", and "QiDaoLuanJi." To achieve this goal, five steps were carried out: (1) determining feature Zhengs and corresponding key herbs of CVA by analyses of clinical cases; (2) finding out potential targets of the key herbs and clustering them based on their functional annotations; (3) constructing an ingredient-herb network and an ingredient network; (4) identifying modules of the ingredient network; (5) illustrating the mechanism of the treatment by further mining the latent biological implications within each module. The systematic study reveals that the treatment of "FengXieFanFei," "FeiQiShiXuan," and "QiDaoLuanJi" has effects on the regulation of multiple bioprocesses by herbs containing different ingredients with functions of steroid metabolism regulation, airway inflammation, and ion conduction and transportation. This network-based systematic study will be a good way to boost the scientific understanding of mechanism of the treatment of Zhengs.

Figure 1

The flow diagram of the systematic study.

Figure 2

Count of Zhengs in the clinical cases. Among the clinical cases, there were 133 Zhengs in total. This histogram only shows the Zhengs with a count larger than 3. The horizontal axis represents different Zhengs, and the vertical axis represents the counts of one Zheng in the clinical cases.

Figure 3

Conditional probability of an herb given feature Zhengs. This figure shows the conditional probability of an herb given feature Zhengs (“FeiQiShiXuan,” “FengXieFanFei,” and “QiDaoLuanJi”). Herbs with probability less than 0.1 are ignored.

Figure 4

The interactions between potential targets and compounds. Yellow nodes represent the compounds, and the blue ones are potential targets.

Figure 5

Herb-ingredient network. In this network, circular nodes are ingredients, square nodes are herbs, and an edge represents that an ingredient is a composition of an herb. Different colors stand for different target clusters. The shades of the colors represent the relation intensity of an ingredient to a disease-related target cluster. To be more specific, if the number of targets in Cluster 4 is larger than targets in Cluster 5, the ingredient is assigned purple, and the ingredient is assigned blue if the number of targets in Cluster 5 is larger. The shade of the color is proportional to the absolute value (ranging from 1 to 7) of the difference of the number of targets in Cluster 4 and Cluster 5. The only ingredient, with an absolute value of zero, was colored as bluish violet. Ingredients with targets which are not in disease-related target clusters are yellow, and those ones without any targets are not shown in this network. As the ingredients were hardly collected, they are not the overall compositions, but they can elucidate the problem in some degree.

Figure 6

Ingredient network. The meanings of colors are the same as in Figure 5, the same in Table 3.

Figure 7

Modules of the ingredients network. The blue and purple nodes are in different modules in general, in accordance with the expectation that ingredients of different functions are mainly located in different modules. Then large modules should be in accordance with the main functions of key herbs. Module 1 is mainly composed of ingredients whose targets are not in the disease-related target clusters. Module 2 and Module 3 are composed of ingredients whose targets are mainly in the disease-related target Cluster 4 and Cluster 5, respectively.