We propose a comprehensive analysis of all available influenza A gene sequences from isolates from avian hosts with the specific aims to (i) estimate rates of cross-hemisphere movement for all gene segments and subtypes, (ii) identify predictors of cross-hemisphere movement, (iii) analyze global reassortment rates, and (iv) analyze associations between gene segment movement and reassortment events.
NON-TECHNICAL SUMMARY: Phylogenetic trees of all avian influenza virus (AIV) gene segments show evidence for a division between the eastern and western hemisphere. This division has been interpreted as evidence for limited virus movement between the hemispheres. Recently, our research group found exceptions to this hemispheric division in the phylogenetic tree of the AIV H6 gene. These exceptions provided evidence for recent cross-hemisphere gene movement and reassortment. For this project we propose a comprehensive analysis of cross-hemisphere movement and reassortment of all AIV gene segments and subtypes. We will estimate segment and subtype specific cross-hemisphere movement and reassortment rates. In addition we will search gene sequences for markers of increased cross-hemispheric movement and reassortment. We expect this analysis to significantly contribute to our understanding of the factors that drive global AIV circulation. This work will allow us to better assess and mitigate the risk that highly pathogenic AIVs currently restricted to Eurasia could invade North America.
APPROACH: Cross-hemispheric movement rates will be estimated from published gene sequences using the Monte Carlo Markov Chain package MIGRATE 3.0.3. Amino acid markers for cross-hemisphere movement and reassortment will be determined by correlating changes on amino acid position with changes in phenotypic state (such as cross-hemisphere movement or reassortment) using sign tests on phylogenetically independent pairings. To estimate reassortment rates, phylogenetic trees with timed nodes will be estimated for each gene segment using BEAST 1.4. Next, the minimum number of reassortment events will be determined that are needed to resolve topological differences between any two phylogenetic trees from different gene segments. This will provide an estimate of the number of reassortment events per unit time for different segment pairs. Two reassortment models will be fitted to the data - one with different reassortment rates for invading and non-invading gene segments and one with equal rates - to determine whether cross-hemisphere invasion is associated with increased reassortment.