Subtle differences in receptor binding specificity and gene sequences of the 2009 pandemic H1N1 influenza virus

A recent phylogenetic inference indicated that the 2009 pandemic H1N1 strains circulating from March 2009 to September 2009 could be divided into two closely related but distinct clusters. Cluster one contained most strains from Mexico, Texas, and California, and cluster two had most strains from New York, both of which were reported to co-circulate in all continents. The same study further revealed nine nucleotide changes in six gene segments of the new virus specific for the two clusters. In the current study, the informational spectrum method (ISM), a bioinformatics technique, was employed to study the receptor binding patterns of the two clusters. It discovered that while both groups shared the same primary human binding affinity, their secondary binding preferences were different. Cluster one favored swine binding as its secondary binding pattern, whereas cluster two mostly exhibited the binding specificity of A/South Carolina/1/18 (H1N1) (one of the 1918 flu pandemic strains) as its secondary binding pattern. Besides all the nine nucleotide changes found in the previous study, Random Forests were applied to uncover several new nucleotide polymorphisms in 10 genes of the strains between the two clusters, and several amino acid changes in the HA protein that might be accountable for the discrepancy of the secondary receptor binding patterns of the two clusters. Finally, entropy analysis was conducted to present a global view of gene sequence variations between the two clusters, which illustrated that cluster one had much higher genetic divergence than cluster two. Furthermore, it suggested a significant overall correspondence between the nucleotide positions of high importance in differentiating the two clusters and nucleotide positions of high entropy in cluster one.