Potential use of host-derived genome signatures to root virus phylogenies

Two methods are presented that provide independent evidence with which to test virus-phylogeny roots. The methods may be applied to phages and other viruses in which at least one horizontal transfer between hosts is inferred across the virus phylogeny. The methods are based upon the inference that viral DNA sequences acquire similar genome signatures (or "genomic signatures") to those of their hosts. One of the two methods may be applied to horizontal virus transfers between three or more hosts. Both methods may potentially be extended to rooting plasmid and transposable-element phylogenies. The effect of using different word lengths (2-bp, 3-bp, or 4-bp) to calculate genome signatures, which host genomes (of bacteria) or genome regions (of eukaryotes) were used, which virus sequences were used, and whether distance, counts, or Z-scores are applied were examined using empirical datasets.