[stevewillie (OP)]

Since we don't know what 'junk' DNA codes for, is it possible that whole structures, like wings, could be encoded in animals that don't have wings? This would help explain sudden, rather than gradual, structural changes in the fossil record. Genetic regulatory mechanisms might derepress "junk" genes and allow their expression.

[JimBob]

It is possible. Our understanding of genetics is rather incomplete - to say the least!

[RD]

Like legs on a whale...

Probably the most well known case of atavism is found in the whales. According to the standard phylogenetic tree, whales are known to be the descendants of terrestrial mammals that had hindlimbs. Thus, we expect the possibility that rare mutant whales might occasionally develop atavistic hindlimbs. In fact, there are many cases where whales have been found with rudimentary atavistic hindlimbs in the wild (see Figure 2.2.1; for reviews see Berzin 1972, pp. 65-67 and Hall 1984, pp. 90-93). Hindlimbs have been found in baleen whales (Sleptsov 1939), humpback whales (Andrews 1921) and in many specimens of sperm whales (Abel 1908; Berzin 1972, p. 66; Nemoto 1963; Ogawa and Kamiya 1957; Zembskii and Berzin 1961). Most of these examples are of whales with femurs, tibia, and fibulae; however, some even include feet with complete digits.

http://www.talkorigins.org/faqs/comdesc/section2.html#atavisms_ex1

[LeeE]

All vertebrate animals have the same basic structure - wings are just different arms (on bipeds) and different forelegs (on quadrapeds).  The next major difference is with the cartilegenous animals i.e. sharks.  Then there's a huge gap to the next family.

It is very likely that 'junk' DNA codes for something, although it may not code in the same straightforward and relatively simple ways that we understand so far.

There are two aspects of DNA coding - the code itself, and the rules for the coding, which may be several layers deep.  We've only just scratched the surface and know just one, or perhaps two of the rules, but that doesn't mean that there are not more coding rules embedded in there.

[blakestyger]

Quite a lot of the so-called junk DNA is Viral in origin.

It has been discovered so far that the genomes of every living organism (or those investigated) is laden with the remains of ancient viral infections - mostly retroviruses. These are RNA viruses that, after infecting a cell, convert their genome into DNA and then incorporate it into the host's genome. If they become permanent they become endogenous retroviruses (ERVs); and human DNA is full of them: 8% has been clearly identified and 40 - 50 % looks very like ERVs and much of the rest consists of DNA elements that multiply and spread in virus-like ways.

This amounts to about 90% of the human genome consisting of virus-like genes. [:0]

There's a lot more about this in last week's New Scientist article Welcome to the Virosphere, p38-41.