Font Size: a A A

Research On Prediction Methods Of Virus-host Protein Interaction And Visualization Implementation

Posted on:2021-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:M Y GaoFull Text:PDF
GTID:2480306569493474Subject:Cyberspace security
Abstract/Summary:PDF Full Text Request
Global infectious diseases caused by viruses not only seriously endanger human health and life safety,but also cause serious economic losses.Recently,2019-n Co V had a great impact on social production and life order during the epidemic.By predicting potential virus-host interactions,we can understand the process of viral infection and thus promote the development of antiviral drugs.Understanding virus-host interactions and their networks not only contributes to a deeper understanding of the nature of life and the mechanisms of disease,but also promotes the research of target drugs.At present,predicting potential virus-host protein interaction has become an important direction of virus research.Common virus-host protein interaction prediction algorithm can be divided into two categories:one is based on the biological information of virus and the host protein,another is based on the topology of virus-host protein interaction network.The former is according to the virus and the host's biological information to find the potential relationship between virus and host,the latter is according to the topology of virus-host protein interaction network.On this basis,this paper proposes a virus-host protein interaction prediction method that fuses GO term similarity and common neighbor similarity.Firstly,the virus-host protein interaction network and host-host protein interaction network have been constructed.Then,host-host-protein interaction network is weighted according to the GO term of the host proteins and the common neighbor similarity.Finally,the potential virus-host protein interactions were predicted according to the functional similarity modules of the host proteins obtained from the clustering.The proposed algorithm fuses the topological characteristics of the interaction network and the genetic functional similarity of proteins to predict the virus-host protein interaction.The experimental results are compared with other recent prediction algorithms.The results show that the algorithm proposed in this topic has certain improvement in accuracy,precision,recall rate and F1 value,compared with other prediction algorithms.With the development of molecular biology,especially the development of yeast double hybridization,gene chip,molecular mass spectrometer and other highthroughput experimental techniques,a large number of experimental data and literatures related to the viral infection process have been produced.Therefore,this paper also built a new virus-host protein interaction database and its visualization system,this system has realized the browse,download,submit,query,visualization,and other functions,useing the network to the virus-host protein visualization.Researchers can easily retrieve needed information about virus-host interactions.
Keywords/Search Tags:Virus, Interaction prediction, GO term, Common neighbor, Visualization
PDF Full Text Request
Related items