| Influenza viruses are the most widespread zoonotic acute infectious diseases onthe planet which is also the number one pathogenic viruses. Human beings is alwaysfight against this kind of virus. It is like an atomic bomb hiding in human society. Thebattle between human society and the viruses never stopped. The key point ofpreventing, diagnosing, curing and controlling the new influenza outbreak is to have acomprehensive understanding of the evolution, cross-species transmission, infectionand pathogenic mechanism of the influenza viruses. The complex interaction betweenvariability and host of the influenza virus made the research complicated andchallenging.The main idea of this paper is to study the host specificity of influenza virusesthrough influenza virus genome sequence and RNA structure by using statistics andmachine learning. This research is helpful to understand the evolution andpathogenicity of influenza viruses. In this paper, first, I have collected influenza Avirus subtype eight segments of gene sequences: H1N1, H3N2, H5N1, H7N9andH2N2. Its host is humans, birds and pigs respectively, screen and pretreated thedownload genomic sequence. Then, I calculated the frequency of different subtypes ofinfluenza virus in different host which reflect the specificity of the host and thesubtypes. It turned out that, subtype H5N1is not outstanding in humans, birds andpigs host specificity. However, other subtype influenza viruses have a stronger hostspecificity.In addition, I draw the conclusion by using AP cluster and Rand index measuringthe composition frequency of different nucleotides’ distinguish ability for differenthosts and different subtypes of influenza virus.On the other hand, RNA structure hasan important function in the virus life cycle of the influenza and it has enormouspressure on virus evolution.It is a relatively new starting point that studying the hostspecificity of influenza virus from the perspective of RNA structure in influenza virus. First we make use of the minimum free energy algorithm which based on dynamicprogramming predicting the secondary structure in each gene fragment of fiveinfluenza viruses subtypes, comparing differences between different subtypes ofinfluenza virus in a host of different secondary structure and minimum free energy ondifferent gene fragment, and studying the host specificity through its free energy andstructural level. |
