About ten years ago I heard Fred Hoyle give a talk where he argues that "junk" DNA segments in fact must code for something else - his particular conjecture was that they coded for structural instructions (the example he used was the shape of leaves).
It was intriguing, there is a lot of junk DNA in some genomes, but on the other hand we understand how it comes about as a result of transcription errors and mutations - genes are truncated or erroneously partly duplicated, or skipped over, leaving randomly mutating junk which is both added to and deleted under weak secondary selection.
Further, the shape function Hoyle discussed was very simple, and most likely due to very simple local iterated chemical gradient competition.
It was however testable: mutate the junk where you think shape is coded, and see if the shape changes - I suggested that, at Fred told me to do it. I didn't (duh!) and don't know that anyone else bothered.
Now the Quantum Pontiff points us at a different but related issue - a possible code in the structural assembly of DNA which may control gene activation.
Article is in Nature, shows up in their "advanced online publication", text seems to be inaccessible;
by Segal et al.
Comments on QP's site worth reading also.
FWIW, I'm in the "could be a big deal" camp, although it is not entirely surprising.
Then some of the most interesting stuff is in the "of course" category after someone has the data to elevate the conjecture above the pitcher-of-beer discussion level.
- Log in to post comments