Nils Reinton and I are discussing junk DNA on his blog [
More crap from the junkies]. It might surprise you to learn that this "junkie" still isn't convinced that junk DNA is dead. Nils isn't convinced that junk DNA exists.
This is what a real scientific controversy looks like.
Can you elaborate upon this statement:
ReplyDeleteThe rest may not be junk because it may have a purpose that doesn’t depend on sequence but that’s not a very likely possibility.
Why do you think this is unlikely, and can you put a value on how unlikely you think it is? 10%? 1% 0.01%? Sufficiently unlikely as to not merit experimental effort?
anonymous asks,
ReplyDeleteWhy do you think this is unlikely, and can you put a value on how unlikely you think it is? 10%? 1% 0.01%? Sufficiently unlikely as to not merit experimental effort?
I think it's unlikely that a significant percentage of our genome consists of DNA that has an essential function that doesn't depend on its sequence.
The reason this is unlikely is that we already have good evidence that 50% of our genome consists of defective transposons and pseudogenes. I can't think of a reason why most of this should be functional but if you can come up with a reasonable hypothesis then, by all means, test it.
Meanwhile, in the absence of positive evidence—or even reasonable speculation—I think it's unlikely. I haven't been able to think of anything in thirty years but maybe I'm just not thinking hard enough.
Furthermore, we known that at lest 10% of our genome is intron sequence that doesn't seem to be required by any test that have been done so far. There's no reason to think that this extra intron sequence contains some mysterious function that we don't know about. However, if you can come up with a reasonable hypothesis that says there must be a function in there somewhere then, by all means, apply for a grant and test it. I won't try and stop you.
Sure, but you didn't answer the question. Since we stipulate that the function (if any) is not dependent on sequence, much of the content being defective transposons and pseudogenes isn't really relevant. The question was, approximately how unlikely do you think such a non-sequence dependent function is?
ReplyDeleteMy 2c: ISTM that the likeliest candidates for non-sequence-dependent function would be the effect of C-value on the cell cycle, in light of which development has been tuned, and the effect of homology on recombination, obliging individual genomes to not depart too far from a vague 'species average' - a filtering effect of trisomies and other meiotic failures.
ReplyDeleteWith suitable experimental design (again, ISTM), sequence-independence (ie known non-functional sequences can be pasted instead with no effect) is a testable definition of junk. We might hesitate to put a figure on it, and we may still be invited to prove a negative, but we can apply a confidence level.
To use impact-free excision as a criterion removes both sequence and any contribution to emergent effects, rendering distinction more difficult as to which function has been excised.
Defining junk in the second way allows you to spot non-junk, but not to ascertain junk (2) status - any more than I could indicate which cells you can remove from an organism before you start to affect function.
anonymous asks,
ReplyDeleteThe question was, approximately how unlikely do you think such a non-sequence dependent function is?
I'm not sure why you're so hung up on a precise answer to this question? Is it a trick of some sort?
I think there's only a slim chance (1%) that a *majority* of the DNA in our genome has a function of any sort. I believe we could delete at least half our genome (the junk part) without any effect on the survival of individuals or the species. In fact, I think we could delete as much as 90% without effect but I'm a little less certain of that number—say, 75% likely to be correct?
There, does that make you happy?
If you disagree, then please address The Onion Test before continuing. That will at least define the question so I can understand where you're coming from.
Sure, that's a great answer. It's not a trap question. For me, I'd go with 10% likely to have some non-sequence dependent function. What I wonder about for example is why the human Y-chromosome has stabilized around the 50 Mb size. Intramolecular meiotic recombination can delete repetitive sequences so if the Y chromosome is stable in length evolutionarily either the generation of new junk is coincidentally equal to the frequency of recombination mediated deletions, or alternatively, there is some selective pressure that maintains the Y-chromosome at 50 Mb length, regardless of sequence content of that 50 Mb. Something to think about anyway.
ReplyDelete