Research On Phylogenetic Network Construction And Evolution Of Influenza Virus

Influenza viruses have multiple subtypes and cause great damage to humans.This is due to the fact that some reticulate evolution events such as recombination or gene reassortment make them capable of antigenic shift.Therefore,it is important to understand whether a new influenza virus is caused by some reticulate evolution events.The phylogenetic network can analyze the evolutionary history of these influenza viruses,which is important for studying how influenza viruses mutate in the natural environment.Another factor that makes influenza viruses so prevalent is their two surface proteins,hemagglutinin HA and neuraminidase NA,which accumulate small mutations causing antigenic drift.The most effective way to prevent and treat influenza viruses is by vaccination.The way to judge the effectiveness of an influenza vaccine is the antigen ic similarity between influenza and influenza vaccines.This paper starts with the antigenic shift of influenza virus,and studies how to construct a more accurate evolutionary history of influenza virus.Next,aiming at the antigenic drift of influenza vi rus,an antigenic prediction method using the similarity data of influenza virus and influenza vaccine is proposed.The main research contents and contributions of this paper are as follows:(1)When several reticulate events exist simultaneously,the performance of some existing phylogenetic network reconstruction method will decrease.Based on this,this paper proposes a new phylogenetic network reconstruction method QS-Net,which uses the information of the relationship between the six taxa.In order to evaluate the performance of QS-Net,experiments were performed from three artificial sequence data sets.Comparison with other methods shows that QS-Net is comparable to other methods in reconstructing the tree evolution history,and is superior to other m ethods in reconstructing reticulate events.In addition,various reconstruction methods were run in a bacterial taxonomic real data set consisting of 36 bacterial species and the whole genome sequences of 22 H7N9 influenza A viruses.The results indicate that QSNet is capable of inferring commonly believed bacterial taxonomy and influenza evolution as well as identifying novel reticulate events.(2)The hemagglutination inhibition HI test reflects the antigenic correlation between the tested influenza virus(antigen)and the reference influenza vaccine(antisera).Furthermore,antigen characterization is typically based on more than one HI data set.The combination of multiple data sets results in an incomplete high information matrix with many unobserved entries.Based on this,this paper proposes a new method NMFAS,which integrates the similarity data of influenza virus and influenza vaccine to predict the incomplete value of H3N2 seasonal influenza A virus data set.Compared to some previous methods,NMFAS reduced the predicted root mean square error RMSE in a 10-fold cross-validation analysis.Then,using multidimensional scaling MDS to construct antigenic cartography and genetic cartography of influenza viruses to comprehensively analyze their antigen and genetic evolution,the results show that 11 antigen clusters can be well distinguished.The obtained antigenic cartography can be used for identification of influenza virus variants,which can be used to promote influenza vaccine selection.