Just like in our DNA...

[Michael Vanier]

Philip Lijnzaad <lijnzaad at ebi.ac.uk> writes:

> > It seems to me
> > (but I am not a biologist!) that it would be extremely unlikely, that
> > without a "Divine Planner", DNA strands would evolve in a way so as to
> > be devoid of "Junk" sections.... that's just asking too much, to ask
> > not just for a molecule that "does the job" of carrying the genetic
> > information, but that does it with 100% efficiency.  There's not
> > sufficient selective pressure for the DNA to be forced to be efficient
> > (as opposed to, for example, the sunlight-collecting efficiency of a
> > leaf, where efficiency matters).  The code-bases are small, molecules
> > are cheap, and there's plenty of room for some extra unused bits in
> > there.
> 
> I'm not going into religuous debates here, but current understanding is that
> this is very achievable. For small bacteria and especially viruses, the junk
> is very costly indeed, and getting rid it offers a selective advantage (and
> they they have little or no junk DNA). 
> 
> What's more (and this I think is exceeddingly interesting and has been
> pointed out many times before): some viruses (I thnk phiX 174 is one) have
> double coding. To explain: each triplet of nucleotides (DNA building blocks;
> there are 4 different ones) translates to one amino acid (protein building
> blocks; of these there are 20 different ones). Now 4x4x4 gives 64 different
> triplets, coding (with a lot of redundancy) for 20 differnt amino acids. This
> is called the genetic code, and there are largely the same throughout all the
> kingdoms (there are a few variations, but let's gloss over that) Now to
> translate DNA into protein, it is critical that you now where the triplets
> start; CGT TTG AAC CCC specifies a differnt piece of protein than 
>     [C] GTT TGA ACC [CC], which is exactly the same DNA but the translation
> was started shifted one nucleotide to the right. In most cases, this protein
> is just nonsense, but some viruses actually make use of this: they use the
> same stretch of DNA to specify different stretches of protein, basically
> exploiting the redundancy in the genetic code. Very cool, and I've been told
> that in the early days of computing and especially games programming, similar
> tricks with the instructino sets of of early microprocessors were done.
> 

It's also important to realize that bacteria and viruses have much tighter
space constraints than eukaryotic (e.g. mammalian) cells, which are much
larger and have *much* more DNA.  In the latter case, there may be little
or no advantage to getting rid of all untranscribed (I dislike the term
"junk") DNA.  In fact, given the way that chromosomes cross over at meiosis
(to produce sperm or eggs), it may be advantageous to have some
untranscribed DNA between transcribed regions, so that crossovers recombine
existing genes without disrupting them.  There may also be structural
reasons for having untranscribed DNA as well.  In fact, it may be important
to have particular AT/GC ratios in untranscribed DNA even if the particular
sequence is of little importance, because this ratio may affect how easy it
is to uncoil the DNA prior to RNA transcription.  And there may be other
uses for untranscribed DNA as well -- who knows?

But back to python... :-)

Mike

--------------------------------------------------------------
Mike Vanier	mvanier at bbb.caltech.edu
Department of Computation and Neural Systems, Caltech 216-76
GNU/Linux: We can't lose; we're on a mission from God.