Yes, quite a lot. Most interestingly for me, the notion that 4 base pairs is, in some manner, the "most efficient" way to represent data. I have been an avid collector of such arguments ever since I was in college, and I can attest that the matter is hardly settled. The most entertaining paper on this subject that I know of, dates back to the 70's, coming out of the memory-i/o chip size reduction races, and postulates that the perfect representational base to compromise the cost of storage with the cost of transmission, is e (2.71828....).
Which makes balanced trinary, not quat, the most efficient realizable representational base. This is a pretty open question today, but I have never seen quat even put on the table as a possible answer.
There's a preprint from "Complexity International" that claims to have an argument that four is the proper number. I don't remember the guy's name, but I think the journal is available on the net. (There haven't been new issues for several years though.)