Re: Junk DNA: A hypothesis

From: Larry Moran (lamoran_at_bioinfo.med.utoronto.ca)
Date: 01/23/05 

Date: Sun, 23 Jan 2005 15:18:42 -0500 (EST)

On Sat, 22 Jan 2005 13:01:45 -0500 (EST),
Tim Tyler <tim@tt1lock.org> wrote:
> Larry Moran <lamoran@bioinfo.med.utoronto.ca> wrote or quoted:
>> On Thu, 20 Jan 2005 23:53:54 -0500 (EST),
>> Tim Tyler <tim@tt1lock.org> wrote:

[snip]

>> > There are dozens of theories of possible "functions" of "junk"
>> > DNA.
>>
>> Indeed there are. Let's see if any of them make sense ....
>>
>> > Junk DNA affects linkage - and the probability that genes are
>> > split during meiosis - and if that probability is small it
>> > can help some deleterious genes form selfish cartels.
>>
>> This one doesn't make any sense.
>>
>> > Another theory is that introns decrease the probability of
>> > crossovers occurring in the middle of functional genes - and
>> > turn meiosis into more of a proper shuffle - rather than
>> > a shuffle that rips cards apart and glues them together again.
>>
>> This one is dead wrong.
>>
>> > Another theory is that "junk" DNA improves evolvability by
>> > creating "neutral networks".
>>
>> This one is "evo-babble" from people who don't understand evolution.
>>
>> > Selection for small program size in genetic algorithms can hinder
>> > evolvability by eliminating intermediate forms that lie between
>> > adaptive peaks.
>>
>> More "evo-babble" - in this case it's from computer scientists.
>
> I don't agree with these assessments - except for the case of
> the first "linkage" hypothesis - where you might well be correct.
>
> I'd be particularly interested in hearing any evidence against the second
> hypothesis (that junk DNA allows a less destructive shuffle to take place).
> I'm relatively new to that idea.

Here's what you said .....

    "Another theory is that introns decrease the probability of
     crossovers occurring in the middle of functional genes - and
     turn meiosis into more of a proper shuffle - rather than
     a shuffle that rips cards apart and glues them together again."

Introns, especially large ones such as we find in mammalian genomes,
will INCREASE the chances of recombination within a gene. I really
don't see why that would make any significant difference at all to
how combinations of alleles are shuffled at cell division. There
will be more crossovers on large chromosomes but this includes more
double crossovers. I just don't see the point ....

> My impression is that the effect is likely to generate a selection
> pressure favouring spacing out genes - and that it would be worthwhile
> attempting to quantify the effect.

Why would stretching out a gene by inserting introns be an adaptation?

> On what grounds do you describe the idea as "dead wrong"?

Well, for one thing, because the presence of introns would INCREASE,
not DECREASE, the probability of a crossover occurring in the middle
of a functional gene. Even if you incorrectly define a gene as only
the coding region (exons) there's no way to make sense of what you
said. In that case there would be no effect because the target size
didn't change when introns were added.

>> > Also, DNA harbours multiple regions such as LINE-1. These
>> > act as mutagens - and are likely to be directly deleterious.
>> >
>> > However they generate a different pattern of mutations from (say)
>> > cosmic rays. They generate patterns of mutations that affect
>> > different regions of DNA more frequently than others. That
>> > whole set-up could serve some functional purpose.
>>
>> There's no way that a population could select for enhanced deleterious
>> mutations on the remote chance that some of them will be beneficial
>> every million years or so. That sort of thing would require an
>> intelligent designer with foresight.
>
> That was essentially Williams view in 1966 - that organisms could not
> do anything but attempt to minimise their mutation rate.
>
> IMO, that's not correct.

I'm not against any form of group selection and I'm quite well aware
of situations where there are trade-offs between mutation rate and
other things.

> The issue of to what extent organisms do this is one of the relative
> strengths of species and individual-level selection - and I don't
> think anyone can authoratitively answer such questions today.

You are proposing something very specific. You are suggesting that
junk DNA is not junk. It's there for a reason and that reason is
the possibility of future benefit. In other words, junk DNA is an
adaptation. However, you fail to explain your mechanism even though
you seem to be aware of the main criticisms of such schemes.

It's true that species-level selection can't be ruled out but you
can't hide behind that generality. You need to propose and defend
your specific example to see if it makes any sense. It doesn't
make sense to me.

I think that junk DNA is junk and its presence is historical accident.
I don't think individuals or species are under such strong selection
that every imaginable deleterous character will be removed as soon
as possible. That's why I'm not troubled by the presence of something
that looks like junk. I'm also not troubled by wisdom teeth, sore backs,
hernias, and male pattern baldness. I don't feel the need to invoke
weird adaptionist explanations to account for these things. Why do you
feel the need to find an adaptionist explanation for junk DNA?

Maybe at some point in the future there will be an advantageous mutation
arising out of the junk DNA that's accidentally carried in genomes.
That's not a reason for claiming that most of the DNA isn't junk right
now and it doesn't count as a sophisticated way of turning junk DNA
into an adaptation.

>> > We know that some evolution takes the form of duplication and
>> > modification. Rapid elimination of unnecessary non-functional
>> > genes might significantly hamper this process.
>>
>> It might. But once again you are postulating selection for some
>> future possibility.
>
> I wasn't doing that here - or anywhere else.

Yes you are.

>> In this case you suggest that junk DNA has to be kept around because
>> it might become useful in the future. How does an individual know
>> that it needs to store excess DNA in its genome because theres' a
>> small chance that one of its descendants might find a use for it at
>> some time in the far distant future?
>
> Modern organisms tend to be products of past events - and are
> not constructed in anticipation of future ones. They are typically
> constructed as though the future environment is expected to be
> like the ancestral one was.

Right. Junk DNA is just junk.

> I never suggested that the future in which evolvability might
> pay off would have to be very distant. Evolvability is needed
> all the time in populations where pathogens are present.

Now you're changing the argument. Why is junk DNA an adaptation
for protection against pathogens?

>> Note that the genomes of bacteria appear to be under selection for
>> small size so they can replicate quickly. That's why they don't have
>> very much junk DNA.
>
> Indeed.
>
>> If your speculation is correct then there should have been much less
>> gene duplication and divergence in bacterial lineages. Does the data
>> lend support to your speculation?
>
> It's not *my* hypothesis - as my reference should illustrate.

You must agree with it. I can't imagine that you would have posted
a bunch of silly ideas that you know are wrong.

> I am not aware of any investigation of the issue on the basis of
> examination of related organisms. It seems to me that there would be
> many confounding factors if any attempt to compare organisms on very
> different scales to investigate the point was made. I'm rather
> sceptical about whether such a study would reach any useful
> conclusion.

Me too. In fact, there doesn't seem to be any point in doing the
experiment since the hypothesis is silly to begin with.

> It seems to me that a useful approach would be deliberately excising
> the "junk" DNA from small organisms and seeing how they fare over an
> extended period in comparison with wild-types.

That's a difficult experiment. However, we can examine very closely
related species, like frogs and some plants, that differ by several-
fold in the amount of DNA they have. The amount of DNA doesn't seem
matter. How do you explain that?

>> > Some references to theories of the "function" of "junk" DNA:
>> >
>> > "Inserting introns improves genetic algorithm success rate:
>> > Taking a cue from biology"
>> >
>> > - http://citeseer.nj.nec.com/context/17647/246003
>>
>> This seems to be mostly computer scientists discussing biology. I
>> didn't see anything there that's worth remembering. Perhaps you could
>> summarize their very best argument for the selective advantage of
>> excess genomic DNA?
>
> I don't know if I'm competent to judge that.

Then why did you post a summary of this nonsense?

> The hypothesis about avoiding breaking genes seems to be a good one -
> and needs further investigation. That's also true of the "neutral
> network" theory.

Please explain. You can't have your cake and eat it too. Either you
don't know anything about these speculations, in which case you should
have kept quiet and not answered the question, or you think these
speculations are interesting, in which case you should try and defend
them.

Tell me about the "neutral net" theory. It looks like techno-babble
to me but I assume you see something that I don't. Why do you think
it's a "good" hypothesis?

Larry Moran

Relevant Pages

  • Re: DNA varies widely between closely related organisms
    ... > I'm not sure "junk DNA" is an apt or wise term for something we do not ... More likely new genes arise from gene ...
    (sci.bio.evolution)
  • Melting into Our Genes
    ... Our genes are spelled out one by one in long twisting molecules of DNA, ... many of these useless "introns" lie between chunks of gene that melt at ... genetic code determines where the introns appear in our genes. ... precisely how junk DNA has worked its way into the genetic code over ...
    (sci.physics)
  • Re: Shall we get rid of our junk?
    ... Presumably there is a small negative fitness effect on ... those species that can tolerate large amounts of junk DNA. ... of evolution may not be due to natural selection. ...
    (sci.bio.evolution)
  • Re: Junk DNA: A hypothesis
    ... >> mutations on the remote chance that some of them will be beneficial ... On the other hand, negative selection, coupled to a quantum-mechanically ... In this case you suggest that junk DNA has to be ... > aggravated by adding arrogance (not the least AEVASIVE arrogance) ...
    (sci.bio.evolution)
  • Re: Cold water on micro-RNAs
    ... There is a cost for replicating the genome, ... However your prediction that mutation and selection would quickly dispose of junk DNA is not solidly based. ... In eukaryotes the cost of DNA replication is a relatively small part of the cell's energy and materials budget, compared to bacteria where the cost of DNA replication is significant and junk DNA is essentially absent. ...
    (talk.origins)