maximal clique centralitymcc bottle- neck BN EcCentricity close- ness radiality betweenness stress 11 MATLAB Zscore Zscore metalloproteinase-2 1

Size: px

Start display at page:

Download "maximal clique centralitymcc bottle- neck BN EcCentricity close- ness radiality betweenness stress 11 MATLAB Zscore Zscore metalloproteinase-2 1"

倾洪甘
5 years ago
Views:

1 * 张燕玲, 黄明峰, 乔延江 ( 北京中医药大学中药信息工程研究室, 北京 ) 11 - C γ % 79. 2% 81. 5% 78. 3% 79. 6% coronary heart disease CHD OMIM CHD Reactome 4 5 Cytoscape Plugins Advanced Network Merge 3 1 CHD CHD 1 D CHD E F G BAI51B * Tel net E F D 2359

2 maximal clique centralitymcc bottle- neck BN EcCentricity close- ness radiality betweenness stress 11 MATLAB Zscore Zscore metalloproteinase CTSS cathepsin S 7 GBE1 glycogen-branching enzyme 8 GALK1 galactose kinase1 Fig. 1 The schematic diagram of the relationship between the 9 AGRN agrin CHD networkthe action network of drugs and the action network of traditional Chinese medicine 10 RAD50 DNA repair protein RAD support vector mechine SVM 6 Intersection RBF D F C γ 2 C C E F γ parallel grid search C γ 2 C γ RMS 1. 1 C γ 358 C γ cytohubba degree 1. 3 edge percolated component EPC maximum neighborhood component MNC density of maximum neighborhood componentdmnc IL-2 IFN-γ CD Table 1 Information of some positive and negative targets No. 1 LPL lipoprotein lipase 2 JAK3 tyrosine-protein kinase JAK3 3 PPARG peroxisome proliferator-activated receptor gamma 4 Calmodulin calmodulin 5 MMP-1 MMP-2 matrix metalloproteinase-1 and matrix ~ 4 CD86 FOXO1 atherosclerosis As

As 5. 14-1. 11 B CD86 7 81. 6% 79. 2% 2 FOXO1 SIRT1 SE SP ACC 81. 5% 78. 3% 8 FOXO1 79.

3 As B CD % 79. 2% 2 FOXO1 SIRT1 SE SP ACC 81. 5% 78. 3% 8 FOXO % IFN-γ 9 ELISA IFN-γ IL IL-2 11 IL-2 IFN-γ 12 2 Table 2 Molecules in testing set 1 CD86 cluster of differentiation 86 2 FOXO1 forkhead box protein O1 3 IFN- γ interferon-gamma 4 IL-2 interleukin-2 5 CMPK1 cytidine monophosphate kinase 6 AK2 adenylate kinase 2 7 VKORC1 vitamin K epoxide reductase complexsubunit 1 8 CHEK2 checkpoint kinase 2 9 FGF10 fibroblast growth factor GUK1 guanylate kinase VKORC1 K 10 1 K 2 C = γ = C = γ = Fig. 2 SVM C γ Results of optimized C and γ in SVM model VKORC1 9 FGF C FGF10 γ C γ= % 75. 6% C = γ = Cγ =

4 222 cytohubba interleukin 1 IL-1 23 tumor necrosis factor TNF Calnexin - 24 Calnexin fibroblast growth factor FGF 15 E12 /E47 1 FGF Id SRRM1sacoplasmic reticulum-related matrix protein 1 mrna 17 SRRM1 18 SRRM S6 ribosomal S6 kinase RSK Raf-MEK-ERK RSK - Ser vasodilator-stimulated phosphorprotein VASP 3 VASP VASP 21 Death-associated protein kinase - DAPKs serine /threonine DAPK

5 D J IL-10 J J D J disease J. Am J Physiol Heart Circ Physiol H J Li SZhang ZWu Let al. Understanding ZHENG in traditional Chinese medicine in the context of neuro-endocrine-immune network J. IET Syst Biol Li SZhang BZhang N B. Network target for screening synergistic drug combinations with application to traditional Chinese 19. EVC-3-4 medicine J. BMC Syst Biol2011 5S101. D D ECV304 J J J Lin C YChin C HWu H Het al. Hubbahub objects analyzer-a framework of interactome hubs identification for network biology J. Nucl Acids Res Vapinkvn. The nature of statistical learning theory M. New YorkSpringer Verlag IL-4 IFN-γ CYP2C9 VKORC1 J Azfer ANiu JRogers L Met al. Activation of endoplasmic reticulum stress response during the development of ischemic heart β -6 J. 7. B γ 24. FGF21 J D SIRT Wagner SChiosea SNickerson J A. The spatial targeting and nuclear matrix binding domains of SRm160J. Proc Natl Acad Sci USA / J J α

6 Study on preventive and therapeutic mechanism of traditional Chinese medicine based on analysis on key network nodes ZHANG Yan-lingHUANG Ming-fengQIAO Yan-jiang * Research Center of Traditional Chinese Medicine Information EngineeringBeijing University of Chinese MedicineBeijing China AbstractTo analyze the intersections among the western medicine action network for preventing and treating coronary heart disease CHD as well as traditional Chinese medicine TCMaction network for activating blood and dissolving stasis. In this article11 characteristic parameter values of network nodes including connectivitybottleneckbetweennesswere calculated. The target identification model was established based on key node characteristic parameters in the CHD-western medicine intersection network with support vector machine. Its C and γ parameters were and respectivelywith the predicted accuracies for positive and negative samples of 81. 6% and 79. 2%. The predicted sensitivityspecificity and accuracy of the test set samples were 81. 5% 78. 3% and 79. 6% respectively. Besides the model was applied to predict potential action targets of the CHD-activating blood and dissolving stasis TCM intersection networkand 180 positive nodes and 42 negative nodes were obtained. In this article9 positive nodes including calnexin interleukin-1 tumor necrosis factorwere taken as examples to analyze the action mechanism of TCM for preventing and treating CHD. The results showed that the network potential key target analysis method was helpful to explore the potential action targets of activating blood and dissolving stasis TCMs for preventing and treating CHDmethodologically supportive to reveal the action mechanism of TCMs at molecular and systematic levelsand significant in guiding the research and development of new drugs. Key wordscoronary heart diseasetraditional Chinese medicine for activating blood and dissolving stasisbiological networktopological propertynetwork analysisaction mechanism doi /cjcmm

1 Fig. 1 Coronary heart disease related targets from literatures CHD GP IIb /IIIa β 14 CHD DrugBank Table 2 DrugBank Coronary heart dise

1 Fig. 1 Coronary heart disease related targets from literatures CHD GP IIb /IIIa β 14 CHD DrugBank Table 2 DrugBank Coronary heart dise * 张燕玲, 黄明峰, 乔延江 ( 北京中医药大学中药信息工程研究室, 北京 100102) 2 Intersection 15 221 31 177 15 073 32 376 2 IL-6 JAK /STAT 2 coronary heart disease CHD 1 % 390 1 CHD CHD 1 OMIM Table 1 Coronary heart disease related genes