3.1 Identification of the active compounds in DCHD
DCHD consists of Chaihu(Radix Bupleuri), Huangqin(Scutellariae Radix), Dahuang (Radix Rhei Et Rhizome), Baishao(Paeoniae Radix Alba), Banxia(Arum Ternatum Thunb), Zhishi(Aurantii Fructus Immaturus), Shengjiang(Zingiber Officinale Roscoe) and Dazao(Jujubae Fructus). Using the TCMSP database, the main active components of each herbal medicine should meet the requirements of OB value≥30% and DL value≥0.18. 140 active compounds were retrieved, including 18 in CH, 36 in HQ, 15 in DH, 13 in BS, 13 in BX, 22 in ZS, 5 in SJ, 29 in DZ. A total of 133 highly active compounds were obtained by eliminating 18 repetitive compounds, which are listed in Table 1.
3.2 Identification of targets of DCHD, PD and AHS
We screened out 1034 effective targets of 80 of 133 high active compounds, including 276 in CH, 242 in HQ, 48 in DH, 61 in BS, 73 in BX, 117 in ZS, 32 in SJ, and 185 in DZ. (Tables S1–S8). Simultaneously, We screened 3878 important gene targets related to PD and 2674 important gene targets related to AHS (Table S9-S10). In other words, 129 genetic symbols may be the key to DCHD 's treatment of PD and AHS.
3.3.1 Drug Target - Disease Target Analysis
We through the analysis of Venn's diagram can get the conclusion that there are 129 overlaps between 3878 PD related gene targets, 2674 AHS related gene targets and 1034 effective disease targets.(Figure2)
3.3.2 Drug Ingredients-Target Network Analysis
The drug composition target network consists of 293 nodes (80 compounds and 206 compound targets in DCFT) and 785 edges.As shown in Figure 3, A1 (kaempferol) is CH, and BS has common components;A2 (quercetin) is a common component of CH and DZ;A3 (stigmasterol) is a common component of HQ, BX, SJ, DZ;A4 (baicalein) is a common component of HQ, BX ;A5 (eriodyctiol (flavanone)) is HQ and ZS;A6 (beta sitosterol) is composed of DH, BS, BX and SJ;A7 (sitosterol) is HQ and BS;A8 ((-) - catechin) is the common component of DH, DZ;A9 (mairin,(+) - catechin) which is a common component of BS and DZ;The network shows that many compound targets can be adjusted by a variety of compounds. In addition, we can also roughly observe the relationship between the active compounds and their targets.
3.3.3 Herb Compounds - disease Target- disease Network Analysis
This network was built to show the relationship between eight herbs, compound, PD and AHS targets, PD and AHS(Figure 4). Analysis of this network revealed that the network map contained 201 nodes and 898 edges. The 201 nodes included 69 active components of DCHD, 8 herbs, 2 disease and 129 targets of PD and AHS targets. Figure 3 shows that DCHD may affect drug targets by controlling related proteins (compound targets).
3.4 Screening and analysis of key shared targets
3.4.1 Analyses of a PPI Network
Through the results of Venn graph, we get 129 repeat important targets. In order to illuminate the significance of degree in compound targets, we created a PPI network about the relationship of the common targets between compounds and PD and AHS. (Figure 5) This PPI network consisting of 129 nodes and 492 edges. Meanwhile, the PPI network was constructed via utilizing the network visualization software Cytoscape3.7.2. (Figure 5) In this network, the node size and color are used to reflect the number of combined targets (degree). As shown in Figure 6, the darker color and the larger circle indicates a higher degree. What’s more, the bar plot of the PPI network analysis results showed that AKT1(degree=31), JUN(degree=29), RELA(degree=28), IL6(degree=26), MAPK1(degree=26), APP(degree=20), MAPK14 (degree=20), EGFR(degree=19), MAPK8(degree=19),VEGFA(degree=19) were the pivotal targets in this network. (Figure 7) In this figure, the x-axis represents the number of neighboring proteins of the target protein. The y-axis represents the target protein. From the above figures, we indicating that there were 10 core targets obtained after PPI network analysis, as well as the critical role in the treatment of PD and AHS.
3.4.2 Network construction and topological analysis of "key target organ tissue"
Key shared targets AKT1, Jun, rela, IL6, mapk1, app, Mapk14, EGFR, Mapk8, VEGFA were imported into biogps to obtain the distribution in organs and tissues, and relevant information was imported into Cytoscape 3.72 to build a "key target organ tissue" network map and carry out topological analysis(Figure 8) . In the figure, the blue diamond represents the disease, and the pink triangle node is the key target, The red circle node is the organ tissue, and the circle size represents the size of the degree value. According to the degree value, the information of the first five distribution organs is as follows: CD33+_ Myeloid.2(degree = 4),Prostate.2(degree = 3),CD56+_ NKCells.1(degree = 3),Lung.2(degree = 3),CD56+_NKCells.2 (degree = 2),suggesting that the above organs and tissues play an important role in the treatment of PD and AHS.
3.5 Analyses of Enrichment of GO Pathways
Using the GO enrichment analysis function of the Bioconductor (R), We carried out GO enrichment analyses to further determine the functions of these shared targets from three aspects, and GO entries were determined using a false discovery rate (FDR) of <0.05. Through the analysis of GOBP, GOCC, GOMF, we obtained 2281 biological processes, 65 cell components and 142 molecular functions. As shown in Figure 9(a,b,c), Top 20 functional terms were enriched in the biological process category, such as response to lipopolysaccharide,response to oxidative stress,response to nutrient levels,reactive oxygen species metabolic process,response to steroid hormone,regulation of apoptotic signaling pathway and muscle cell proliferation.Top 20 functional terms were enriched in the cellular components’category, such as membrane raft,cytoplasmic vesicle lumen,secretory granule lumen, mitochondrial outer membrane,nuclear chromatin,platelet alpha granule, endoplasmic reticulum lumen,protein kinase complex,RNA polymerase II, transcription factor complex and serine/threonine protein kinase complex. Additionally, Top 20 functional terms were enriched in the molecular function category, such as cytokine activity,heme binding,cytokine receptor bindingtetrapyrrole binding,nuclear receptor activity,transcription factor activity, and direct ligand regulated sequence-specific DNA binding. Undoubtedly, these biological processes were all involved in the pathogenesis of PD and AHS, so they may serve as a potential therapeutic mechanism for PD and AHS.
3.6 Analyses of Enrichment of KEGG Pathways
Analyses of enrichment of the KEGG pathway were also carried out using Bioconductor (R) and ClueGO. (p< 0.01) The pathways of 129 proteins involved in PPI network were analyzed via pathway enrichment, and 161signaling pathways were obtained. Figure 10 shows the top 20 pathways for DCHD in the treatment of PD and AHS target. The functionally grouped network of enriched categories was generated for the target genes using ClueGO and CluePedia (Figure 11). In addition, we established a drug-target-path network ,which clearly indicates that DCHD may achieve its purpose of treating PD and AHS through multiple targets and pathways. Figure 13 shows the proportion of each group associated with 129 targets. The most significant KEGG terms of the target genes included Fluid shear stress and atherosclerosis Kaposi sarcoma-associated herpesvirus infection,AGE-RAGE signaling pathway in diabetic complications,Human cytomegalovirus infection,Hepatitis B,PI3K-Akt signaling pathway,TNF signaling pathway,Hepatitis C Prostate cancer,Influenza A,Epstein-Barr virus infection,Human papillomavirus infection,IL-17 signaling pathway, Proteoglycans in cancer,MAPK signaling pathway,MicroRNAs in cancer, Measles, Human immunodeficiency virus 1 infection,Salmonella infection,HIF-1 signaling pathway. Figure 13 shows the AGE-RAGE signaling pathway in diabetic complications ,PI3K-Akt signaling pathway,IL-17 signaling pathway and MAPK signaling pathway were discussed to illustrate the underlying therapeutic mechanisms of DCHD for MAPK signaling pathway treatment.