Speaker
Description
Influenza A viruses (IAVs) remain a significant public health threat due to their ability to jump between host species, as demonstrated by the H1N1 pandemic in 2009. Despite increased genomic surveillance, knowledge of the evolutionary dynamics allowing such zoonotic events is still limited, and the genetic markers that facilitate transmission between humans and swine remain unclear. To explore this gap, we analysed a large publicly available whole-genome dataset of human and swine IAV sequences, focusing on H1N1 and H3N2 subtypes. Using phylogenetic reconstructions combined with ancestral sequence and host state inferences, we generated datasets representing the genetic diversity and evolution of IAV proteins at the swine-human interface. We then employed statistical models to identify genetic markers associated with intra- and interspecies transmissions, including analysing mutation rates and selective pressures on the viral proteins and using a gradient-boosted decision tree machine learning approach to predict key amino acid positions critical for different types of transmission. Our analyses revealed complex mutational patterns within and across the viral proteins, but highlighted specific regions and amino acid positions, particularly in the internal gene segments, of potential importance for interspecies transmission. These findings demonstrate the utility of integrating phylogenetic analysis with machine learning to uncover genetic determinants of interspecies adaptations and represent an important step towards identifying key genetic signatures across the IAV proteins that drive host adaptation and zoonotic potential. Such insights offer valuable markers for early-warning genomic surveillance systems, ultimately enhancing human and animal health and minimising the potential for zoonotic transmission of IAV.
| Expedited Notification | No thanks, I do not require Expedited Notification |
|---|
