Machine learning
Moving away from genome scan methods used for human GWAS (ultimately inappropriate for the short highly polymorphic genomes of RNA viruses), the group has demonstrated the potential of multi-class machine learning algorithms in inferring the functional genetic changes associated with phenotypic change (e.g. a virus crossing the species barrier).
These genotype to phenotype (GP) methods allow to uncover a set of features and insights that ultimately could be quite relevant in understanding viral transmission across host species. They:
- Show that even distantly related viruses within a viral family share highly conserved genetic signatures of host specificity;
- Reinforce how fitness landscapes of host adaptation are shaped by host phylogeny;
- Highlight the evolutionary trajectories of RNA viruses in rapid expansion and under great evolutionary pressure.
These methods can serve as rigorous tools of emergence potential assessment, specifically in scenarios where rapid host classification of newly emerging viruses can be more important than identifying putative functional sites.