Epidemic Trend Prediction Of Influenza And Genetic Variation Analysis Of Influenza Virus In Jilin Province

Objective:The purpose of the study was to understand the epidemic trends in Jilin Province from May 2009 to December 2016,and to establish an ARIMA prediction model to predict the incidence of influenza outpatient clinics in Jilin Province from January to March in 2017?The predictive capacity of the ARIMA model was tested by the real outpatient morbidity rate of January to March in 2017.Studying the genetic variation of HA and NA genes of H3N2 subtype and Yamagata type B influenza virus in Jilin Province from 2013 to 2017,to understand the evolution of influenza virus in Jilin Province,and to provide scientific basis for the prevention and control of influenza in Jilin Province.Method:In this study,we collected the monitoring data of Influenza Surveillance System from May in 2009 to March of 2017 in Jilin Province,used Excel and SPSS 24.0software to collate the influenza surveillance data,and used the"tseries"and"forecast"packages of RStudio 3.4.3 to establish ARIMA prediction model.A total of46 strains of H3N2 subtype influenza virus and 24 strains of type B Yamagata virus were obtained from Center for Disease Control and Prevention of Jilin.The HA and NA gene fragments were obtained by RT-PCR and sequencing.Fragment splicing of genes was performed by Senquencher 5.3 software.MEGA6.0.6 software was used to analyze the genetic variation between the sequence of influenza strains and the sequence of selected reference strains.Result:1.The ARIMA model established by the influenza surveillance data of Jilin province was ARIMA?1,0,0??1,0,0?₁₂.This model was the best model for predicting the incidence of influenza in a short-term outpatient clinic in Jilin province.The actual value of influenza outbreaks was basically consistent with the predicted value from January to the March of 2017,and the actual incidence rate from January to the March was within the 95%confidence interval of the outpatient morbidity rate.The best ARIMA model for influenza A surveillance data is ARIMA?2,0,0??1,0,0?₁₂.The best ARIMA model for influenza B surveillance data is ARIMA?1,0,0??1,1,0?₁₂.2.The study found that amino acid in sites of 66,156,174,189,232,242,391,466 in HA gene of H3N2 subtype influenza virus were changed comparing with the reference strain A/New York/392/2004,of which the amino acids in 156 and 189 sites were located in the B region of antigenic determinant of H3N2 subtype influenza virus,and 174 site was located in the D region of antigenic determinant of H3N2subtype influenza virus.There were 7 common amino acid sites in which amino acids changed of influenza strains in 2015,2016,and 2017.Among them,the 160th and187th amino acid sites were located in the B region of antigenic determinant.The 158and 227 sites of amino acids which were located in B and C region of antigenic determinant of H3N2 subtype influenza virus were mutated too.3.Compared with the reference strain A/Victoria/224/2008,11 amino acids mutated in the NA protein sequences of the H3N2 subtype influenza virus,among which the amino acid mutation at position 221 may cause drug resistance changed of the virus.The 77th amino acid of the fourth,fifth,and sixth branches of the evolutionary tree in the phylogenetic tree were all mutated.4.Compared with the reference sequence of B/Victoria/224/2008,amino acid at the sites of 131,218,314,328 were all mutated.The 131th amino acid site was located in the antigenic determinant of B influenza virus.5.With reference to the protein sequence of the B/Victoria/224/2008 vaccine strain,a total of 24 amino acid sites were mutated in the NA gene of two influenza viruses in 2014,and the 198th amino acid change was related to the drug resistance of the virus.In the evolutionary tree of the NA gene of influenza B virus,the second branch and the third branch had the same four amino acid variation sites.Besides,two strains of 2015,amino acid at the site of 198 were mutated too.Conclusion:1.The ARIMA?1,0,0??1,0,0?₁₂ model can better predict the short-term development trend of influenza in Jilin Province.It can provide scientific basis for the prevention and control of influenza in Jilin Province.Influenza A prediction model ARIMA?2,0,0??1,0,0?₁₂ and influenza B predictive model ARIMA?1,0,0??1,1,0?₁₂have poor predictive effects.2.Variants of amino acids at multiple epitopes of H3N2 subtype influenza virus in Jilin Province suggesting that H3N2 subtype influenza virus may produced new variants,and a possible neuraminidase inhibitor resistance site at position 221changed in NA protein molecules may lead to resistance of influenza virus?3.In the viruses of type B influenza,amino acid mutations at position 131 of the antigen binding site was found.A mutation in the resistance site at position 198 of neuraminidase inhibitor was found in the NA protein molecule,and Influenza B virus might develop resistance.