ABSTRACT
The winter of 2009 witnessed the concurrent spread of 2009 pandemic H1N1 with 2009 seasonal H1N1. It is clinically important to develop knowledge of the key features of these two different viruses that make them unique. A robust pattern recognition technique, Random Forests, was employed to uncover essential amino acid markers to differentiate the two viruses. Some of these markers were also part of the previously discovered genomic signature that separate avian or swine from human viruses. Much research to date in search of host markers in 2009 pandemic H1N1 has been primarily limited in the context of traditional markers of avian-human or swine-human host shifts. However, many of the molecular markers for adaptation to human hosts or to the emergence of a pandemic virus do not exist in 2009 pandemic H1N1, implying that other previously unrecognized molecular determinants are accountable for its capability to infect humans. The current study aimed to explore novel host markers in the proteins of 2009 pandemic H1N1 that were not present in those classical markers, thus providing fresh and unique insight into the adaptive genetic modifications that could lead to the generation of this new virus. Random Forests were used to find 18 such markers in HA, 15 in NA, 9 in PB2, 11 in PB1, 13 in PA, 10 in NS1, 1 in NS2, 11 in NP, 3 in M1, and 1 in M2. The amino acids at many of these novel sites in 2009 pandemic H1N1 were distinct from those in avian, human, and swine viruses that were identical at these positions, reflecting the uniqueness of these novel sites.