How many distinct genomes are there that we include under the label "smallpox"? It's also a great deal more complicated than simply storing the genome - much as gene expression in humans is more complicated than simply "this is what the DNA says", it's also more complicated than that for viruses. We're finally to the point that we can simulate interactions, but it's -absolutely- not a trivial thing to do (supercomputer shit) and simply saving the source material prevents any risk of loss of information.