ABSTRACT
Objective:To explore the pharmacological mechanism of Danhong injection (DHI) in the treatment of coronary heart disease with angina pectoris from the level of functional modules by modular pharmacological analysis framework. Method:The targets of drug components in the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) and the angina-related genes in DisGeNET, OMIM and CTD databases were combined to construct the target network of DHI for the treatment of coronary angina pectoris by STRING version 11.0. Functional modules were identified by the molecular complex detection (MCODE), Markov cluster (MCL) and GLay algorithms, and the results were optimized by the minimum network structure entropy algorithm. The Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis was performed on the modules by DAVID version 6.8 bioinformatics analysis platform. Result:By integrating 262 genes related to DHI and 192 genes related to angina pectoris, the target network of DHI for angina pectoris was constructed, including 414 nodes and 6 621 edges. After optimization of the minimum network structure entropy, 12 functional modules (number of nodes>3) were identified by MCODE algorithm, of which the largest module (module 1) has 47 nodes and 962 edges, MCODE score=41.826. KEGG pathway enrichment analysis was conducted on the gene network of DHI for angina pectoris and the modules divided by MCODE, and 37 and 58 KEGG signaling pathways were obtained respectively, with the coverage rate of 86.5%. The pathways enriched by the modules could be roughly divided into 11 categories, among which human diseases (45%), signal transduction (17%), and amino acid metabolism (14%) were involved in a large proportion. Module 1 was enriched into 39 pathways, which was signal transduction-related module. Module 3 was amino acid metabolism-related module. Conclusion:The therapeutic effect of DHI on coronary heart disease with angina pectoris is achieved through multiple modules, multiple pathways and multiple functions, mainly by regulating modules related to signal transduction, amino acid metabolism, neuroactive ligand-receptor interaction, Ca2+ and p53 signaling.
ABSTRACT
BACKGROUND: Communication between cells is indispensable in a multicellular organism society. Many cells communicate with each other, forming a complex cellular network structure. It is very important to measure and evaluate the relevant attribute values of cellular network structure. OBJECTIVE: To propose a metrics method of cellular network structure based on a complex network. METHODS: Based on literature research and practical application, the metrics framework of cellular network structure was established. The structure of cellular network was measured in the aspects of the degree of node, the degree distribution of cellular network, the average path length of cellular network and the cluster coefficient of cellular network. A small experiment was taken as an example to verify the validity of the method. RESULTS AND CONCLUSION: In the degree distribution of cellular networks, the degree values of most cell nodes are relatively small, and only a small number of cell nodes have higher degree values. The more obvious power-law distribution of the degree distribution P(k) of cell nodes indicates the more reasonable structure of the cellular network as well as the more normal cellular network. At the same time, many cellular network structures have smaller average path lengths. The larger cluster coefficient of the cellular network indicates the higher aggregation characteristics of the cellular network. Generally speaking, the tighter the cell network structure is, the more obvious the clustering characteristics of the cell network structure are, and the more normal the cellular network is.
ABSTRACT
Recent years, convolutional neural network (CNN) is a research hot spot in machine learning and has some application value in computer aided diagnosis. Firstly, this paper briefly introduces the basic principle of CNN. Secondly, it summarizes the improvement on network structure from two dimensions of model and structure optimization. In model structure, it summarizes eleven classical models about CNN in the past 60 years, and introduces its development process according to timeline. In structure optimization, the research progress is summarized from five aspects (input layer, convolution layer, down-sampling layer, full-connected layer and the whole network) of CNN. Thirdly, the learning algorithm is summarized from the optimization algorithm and fusion algorithm. In optimization algorithm, it combs the progress of the algorithm according to optimization purpose. In algorithm fusion, the improvement is summarized from five angles: input layer, convolution layer, down-sampling layer, full-connected layer and output layer. Finally, CNN is mapped into the medical image domain, and it is combined with computer aided diagnosis to explore its application in medical images. It is a good summary for CNN and has positive significance for the development of CNN.