Prediction of forthcoming influenza virus strain
Using information science, we estimate epidemics and develop bioinformatics technology prevent zoonosis. Our main focus is influenza, by analyzing massive genetic information from past outbreaks, we are developing tools to estimate future mutation.
The hemagglutinin (HA) of influenza are selected by immune pressure of human, and antigenicity continues to change. Vaccination of influenza are effective, however since various stains with amino acid substitution on HA are isolated every year, it is difficult to select the vaccine strain of the following year. In our research, we use massive HA genetic information to find the pattern of amino acid substitution, and with the pattern of mutation in the past, we developed a method to estimate future possible amino acid substitution. By visualizing the virus in multidimensional space, we found that between different HA there is a pattern in the relative distance. By retrospectively analyzing 12 years of mutation estimation of the following year, this method has shown that it can estimate the angiogenetic variation of the following year in high precision.