Non-Coding Repeats:
Several years ago, when the effort was being made to complete an analysis of an entire human genome, there was a strategic question.  Genes were believed to contain the information for making proteins.  Proteins were responsible for much of the complexity of life.  So it was reasoned that if one understood genes one would understand life.  In those days it was rather popular to exclaim over how much alike all life forms were.  We shared such a proportion of our genes with asparagus, a higher proportion with lizards, more yet with dogs and so forth, and the proportions always seemed wonderfully high.  All life was alike.

It failed to occur to those making the most noise that we don’t look or act much like asparagus, and maybe there was more to the story than the genes.  In those days, the way to prosper in genetics was to find a gene, characterize it and then gloat over how much this would add to human wellbeing.  Everything that wasn’t a gene in the DNA was shamelessly dubbed “junk” DNA.  But while those making the most noise, as those applying for the most genes, were content that genes were the key to everything, those making the decisions took seriously the proposition that maybe we didn’t know all there was to know anyway.

So the choice was made to analyze the whole human genome, all the DNA, not just the genes.  I remember myself at the time thinking that this seemed noble but extravagant.  I am glad the decision was not left up to me.  I might have made the wrong choice.

They had broken the “DNA triplet code” years before.  You could look at a length of DNA and see what amino acids is specified in what order, and knew those amino acids would be strung into large molecules called polypeptides, which would in turn … (mumble, mumble) … PROTEINS.  There is now a great deal of interest in the question of just how the polypeptide chains are folded into useful proteins.  Well and good.

But the “junk” DNA was doing either nothing or doing something else.  One of the things that seemed to exasperate a lot of people was an abundance of “non-coding repeats.”  There were lengths of DNA that could not be interpreted by the triplet code and in which short sequences of this would repeat again and again and again and again and again … again and ….  How is a guy supposed to get a grant proposal out of that?

Some time back I offered to give out copies of my computer model of the genetics of a population evolving over time.  There were no takers.  Had you been reading at that time and had taken the trouble to ask for one and then to play around with it, you will notice that the program asks you what the impact of a single detuning mutation you would like to have in terms of the chance of losing one offspring.  From then on, the program takes you at your word and employs that chance for every mutated locus subject to detuning mutations in every member of the population in every generation.

I reasoned that mutations with a very large impact would be eliminated rapidly from the population and mutations with a very small impact would not occur in significant numbers because evolution would never have been able to tune the system that finely anyway.  So there had to be a range that was significant.  The chosen number would be some middling value representing the rand for simplicity.

But suppose you decide to build a box, a fairly tight one.  You get together your boards, your screws, your hinges and your latch.  These are your nutrition.  Then you take down your saw, your drill and your screwdriver.  These are your genes (or the enzymes they will produce).  You start to work.  You saw a nice piece off the end of your board.  Your skill is superb, and all your cuts are perfectly square.  You now have a bottom for your box.  You saw off another piece for the front of your box.  With deft hands you make your holds and screw it into the bottom.  It doesn’t quite fit, because it’s a somewhat different length, but a bit sticking out is no problem.  I said tight, not pretty.

When you finish with the front, you do a side and the back.  Then you notice that this particular board is used up, so you take another.  It is a somewhat different width, but no matter, you probably have pretty good luck finishing the fourth side.  Of course the top isn’t even all the way around.  But you recon you can work around that.  Now to cut the top.  Then you realize that you cannot finish the job.  You will need a ruler.  So you down tools and have your friend drive you to the hardware store to buy a ruler.  On the way back, sadly reflecting that the parts of the box need to be fine tuned to each other, you examine the tool that will let you finish. 

The ruler is not going to be part of the structure.  Nor is it going to cut or drill a board.  It is going to let you do your fine tuning.  So you look at this thing that is not a gene but will let you appropriately control your gene.  And lo and behold it is a series of identical repeats.  That is the easiest way to fine tune something.

I already suggested that the fine tuning elements would be together.  And of course they are not coding elements.  And now it appears they are probably all about alike.  In other words, if you were looking for a fine tuning mechanism, you would look for runs of non-coding repeats. 

And when you wanted to build another box, or evolve into another life form, you would already have the fine tuning mechanism on line.  You would not need to blunder around waiting for just the right genetic mutation to show up.  You would only need to wait for a mutation that adjusted the number of repeats. 

Those runs of repeats are just what we would expect.  That does not prove they are the control mechanism.  It is just they look a lot like the control mechanism in the computer model and are a plausible form for that mechanism in the real world.

There have been 1,752 visitors so far.

Home page.