In the book Genome, author Matt Ridley starts chapter four like so:
Open any catalogue of the human genome and you will be confronted not with a list of human potentialities, but a list of diseases, mostly ones named after pairs of obscure central-European doctors. This gene causes Niemann-Pick disease; that one causes Wolf-Hirschhorn syndrome. The impression given is that genes are there to cause disease.
As you might suspect, Ridley is at pains to point out that genes do not cause disease. At least they don’t cause disease any more than, say, hearts cause heart attacks. But we tend to notice genes when they fail in spectacular ways.
When it comes to viruses, we have to admit that they often do cause disease. But of all the viruses out there, very few are interested in making us sick. And since they surround us in such a thick cloud, they can perform an under-appreciated role: gene libraries. They are busy little day traders, moving their stock of DNA and RNA in and out of cells all day long. As such, they’re in a good position to acquire, store, and transfer useful genetic knowledge that bigger folk might have written off. Like bacteria, they take a beating for causing various diseases without getting any compensating credit for their health-giving talents and, I think you’ll agree, good looks.
One of the fun things about viruses is that, being small, it’s relatively easy to take their (virtual) snapshots. Virusworld, based at the University of Wisconsin, keeps a regular family photo album of virus pictures. There are some real lookers in the bunch. Check out the handsome reovirus core. And the Norwalk virus that ruined your aunt’s last cruise is a charmer up close. We have such a comprehensive three-dimensional understanding of some viruses by now that we can print out solid copies of them, like this fist-sized version of the pariacoto virus. I find it amusing that this virus has caused us to print out giant versions of itself.
Rambles Blog at Star Chamber 
Why are viruses so symmetrical? They’re sequences of random-looking basepairs, so shouldn’t they look like jumbled string, rather than something out of a 3D kaleidoscope?
anon raises an interesting point about literacy, any thoughts Alan?
The gene sequences of viruses are random in the same way that poetry written in Urdu looks random if you don’t speak Urdu. The symmetry you refer to is the “meaning” of the sequence: it’s the evidence that the sequence isn’t random. Just like an icosahedron is a sphere-like shape assembled from a crowd of small triangles, these viruses encode little chunks that swarm together to make pretty little beach balls that, from the host cell’s point of view, are absolutely terrifying.
Well said, Ned. In fact, when I look at a viral genome, I see restriction sites and open reading frames and codons, etc., and I can use simple online programs that can read even more information out of it. It’s the same for Urdu poetry, and even more the same for binary — how can a sequence of random-looking ones and zeroes produce an application?