More Recent Comments

Tuesday, September 23, 2008

Discussing Junk DNA with an Adaptationist

Adaptationists are scientists who like to find adaptive explanations for all features of organism. For them the concept of junk DNA is difficult to swallow in spite of abundant scientific evidence and in spite of the fact that counter-explanations do not account for the data. Nils Reinton is a molecular biologist working in the field of medical diagnostics and he has been challenging the concept of junk DNA in the comment section of a recent posting. The title of that posting, Everything Is There for a Reason?, was direct response to an earlier posting from Nils where he claimed that we shouldn't label DNA as "junk" because it's a science stopper.

During the discussion in the comment to my posting, I challenged Nils to answer a number of questions. He has responded on his blog SciPhu with Hey junk people, I accept your challenge (part I).

Here are the first four questions with my personal summary of his answers.
Q: Why do pseudogenes and most of the transposon-related sequences look so much like broken genes?

A: They may look like broken genes but they probably have some function.


Q: Why is the DNA sequence in most of our DNA not conserved?

A: It's not conserved because it's a reservoir for evolution. In addition, it probably contains genes for small RNAs and, as we all know, those predicted small RNA genes are not conserved.


Q: Why can we delete large segments of mammalian DNA with no observable effect?

A: There is an effect. We just haven't found it yet.


Q: Why is there so much variations in repetitive DNA within a species? Some people have segments that are ten times longer than segments found in other people. Are all of the nucleotides in the longer segments functional?

A: There are some examples of differences in repeats that do make a difference. Therefore, it is wrong to conclude that most of the variation has no effect. Furthermore, the discovery of copy number variation is a new phenomenon and it may turn out the have profound effects.


Nils concludes Part 1 by repeating his earlier complaint ..
This belief that there’s hidden function to be found, treasures to unearth if you will, is the difference between those advocating these parts of DNA as “junk” and me. In my opinion, It’s not the details of what is junk and what isn’t, ..- and how much, that bothers me…..

It’s the attitude. To dismiss something as junk is contrary to my idea of science being driven out of curiosity and the need to explore. Curiosity may kill a cat every now and then, but I’ll take that risk and continue to praise the scientist who recognize possibilities in the junk rather than dismissing it.
This is nonsense but I already covered that complaint in my previous posting.

Nils seems to think that the adaptationist program is the only way to remain curious and excited about science. This is in direct contrast to the original Spandrels paper by Gould and Lewontin. They argued that strict application of the adaptationist program prevents people from seeing other possibilities. It's a science stopper.

Gould and Lewontin argued that a pluralist worldview was far superior because it considers a wider range of possible explanations.

Personally, I'm as excited about the possibility that our genome could be 95% junk as I am about the possibility that there may be strange new features that we don't know about. At least the tentative conclusion that much of it is junk has the advantage of being a superior explanation of the data.

The conclusion that most of the DNA has some unknown function in spite of much evidence to the contrary strikes me as non-scientific. To justify it on the ground that such a belief is required in order to maintain an interest in the subject is almost unbelievable.

Incidentally, despite some initial skepticism1, Richard Dawkins—not usually thought of as the best example of a pluralist—has now resorted to using junk DNA as one of his arguments against Intelligent Design Creationism.
Gene duplications have occurred from time to time throughout the genomes. It is by these, and similar means, that genome size can increase in evolution. But remember the distinction between the total capacity of the whole genome, and the capacity of the proportion that is actually used. Recall that not all the globin genes are actually used. Some of them, like theta in the alpha cluster of globin genes, are pseudogenes, recognizably kin to functional genes in the same genomes, but never actually translated into the action language of protein. Genomes are littered with with nonfunctional pseudogenes, faulty duplicates of functional genes that do nothing, while their functional cousins (the word doesn't even need scare quotes) get on with their business in a different part of the same genome. And there's lots more DNA that doesn't even deserve the name pseudogene. It, too, is derived by duplication, but not duplication of functional genes. It consists of multiple copies of junk, "tandem repeats," and other nonsense that may be useful for forensic detectives but which doesn't seem to be useful in the body itself.
            Richard Dawkins in "The "Information Challenge",
            The Skeptic magazine, 1998

1. The Extended Phenotype p. 157

23 comments :

Anonymous said...

Eehhh, Allrighty then, First: - I repeat and repeat again: I am not an adaptationist. It's perfectly credible to believe that what you call junk in fact isn't junk, and base this belief on chance events, - and you know this. It is sad that you have to put that label on me to get ahead in the discussion. Second: my answers to your questions are reproduced badly in your post, - for those interested in the full answers, go see my original post.

PhDP said...

...I agree with you, but I would also ask; is this radical form of adaptionism even theoretically possible ?

Unless population size is infinite or all mutations have large effect on fitness (and we know this isn't the case), then the theory of evolution predicts that genetic drift is going to be an important force, often much more important than natural selection.

Of course the point is not that natural selection doesn't exist, but that it has to share the spotlight with genetic drift, mutations and recombination (along with other forces).

I don't want to resort to insults, but this radical panselectionism seems quite similar to religion in at least one aspect; randomness and coincidences are rejected in favor of a quest for purposes.

A. Vargas said...

"There is an effect. We just haven't found it yet"

Dude, you're SO adaptationist.

Anonymous said...

Jeez, of course there is junk DNA. One of the main laws of evolution in my book: Nature goes by "good enough", never by perfect design. Here is an analogy I frequently use to explain it to students (usually during beer drinking sessions because, ahem, I never lucked out to publish in the Big Three and thus am not qualified to teach smart kids in good universities):

Mechanic workshop. There is always a mess. Variable amount of junk. Some of it even turns out useful on occasion. And, to be successful, grow and expand, there is always a balance to be struck between keeping things pristinely tidy and letting the mess take its place for as long as it does not interfere with the job that needs to be done NOW. And so on. Pretty simple and obvious illustration, no need to go into details. Always worked so far (for the people who don't confuse the Mechanic with the Creator :-)).

Rosie Redfield said...

Natural selection simply is incapable of creating a 'reservoir' of material because it might be useful in the future!

James Goetz said...

I hope that Nils knows what he's talking about when he's doing medical diagnostics research.:)

Anonymous said...

Rosie Redfield: Natural selection simply is incapable of creating a 'reservoir' of material because it might be useful in the future!

The truth of this statement depends on your definition of 'future', specifically whether the length of time until cyclically applied future selective pressure is long or short relative to the speed of gene loss in populations through mutation and drift.

A trivial example: Mosquitos proliferating in summer don't have any selective pressure maintaining the function of genes that will prove to be necessary for future generations of mosquitos to survive the winter. Nevertheless, the timescale of degradation of winter-survival genes is long relative to the time until the application of cyclically-applied selective pressure (freezing temperatures) and those individuals that have lost these genes are removed from the population the next time the selective pressure (winter) is applied, so these genes are effectively preserved through the summer in the overall mosquito population.

Of course, having these "junk" winter genes around in summer is really the result of selection creating a reservoir of material that was useful in the past (the previous winter), rather than any kind of prediction of future winters. Nevertheless, with cyclical population stressors, the present-day population genetic makeup sure gives the appearance of predicting future necessity.

Anonymous said...

I have a Major issue with labelling DNA one way or another when the evidence is far from irrefutable.

On simple systems you can test what each variable does simply by removing it and observing.

This is not the case with DNA, if you remove a piece and you can't observe any efect whose to say the affect just isnt observable through the means employed by the experiment.

The only way i can see to properly test the 'junk' DNA hypothesis would require a large batch of clones of of several genotypes of one original genome... The normal including all the DNA that is said to be 'junk' and several others with only the remaining 5%~10%. Then many many years of observation, comparative testing, and reproduction by indepent labs.

Such a conclusive study would be required since there is no other way to really see if there is some infintestimally small purpose to all the "junk" DNA.. (its only purpose might be to allow large scale chromosomal compaction to take place without causing destabilization)

So until such a study is done the large regions of DNA should simply be innominate. For they have no discrenable function but there is no complete proof that they are simply junk.

Anonymous said...

I believe that all DNA-changes occur by chance. I also believe that things stay reasonably constant due to selection pressure, - which is the "reason" I am always referring to. Does this make me an adaptationist ? Even if I get this label, this will not diminish my resistance towards the term "junk" and my argument that we_do_not_know_yet_and_should_keep_an_open_mind, still hold true. And, to make it perfectly clear, I am not advocating the genome as "perfectly tidy", on the contrary, - biology is about extracting useful information from chaos, - but that does not make anything "junk" by default. All I have is "a Major issue with labelling DNA one way or another when the evidence is far from irrefutable.". Thank you Matt for stating it so clearly.

Larry Moran said...

matt says,

I have a Major issue with labelling DNA one way or another when the evidence is far from irrefutable.

I'm sorry, but that boat don't float.

There's a huge amount of scientific evidence is support of the idea that most of our genome is junk. There's absolutely nothing wrong with provisionally accepting the evidence and stating that most of our genome is junk.

As with anything in science, this conclusion is tentative. Nothing is ever proven to the point at which it is irrefutable.

We do this all the time and it never causes a problem. For example, most people are prepared to say that birds are dinosaurs, that mitochondria descend from bacteria, and that Neanderthals were a separate species.

The "problem" only seems to come up when adaptationists get upset over a non-adaptationist explanation. Then, all of a sudden, there's a problem with provisionally accepting the evidence.

Ironically, the adaptationists don't seem to have a problem with declaring that white skin color is an adaptation or that menopause can be explained by the grandmother effect. Neither of these positions are irrefutable either.

From what I can see, the adaptationist position is to "keep an open mind" whenever the evidence points in the direction of accident (non-adaptation) but not when it looks like an adaptation.

The controversy is more fundamental than most people realize. It boils down to the default position when it comes to mechanism of evolution. Should one assume that everything is an adaptation unless proven otherwise? Or, is the default mechanism random genetic drift and the burden of poof is on the adaptationist to demonstrate selective advantage?

Most people who have thought about the problem think that drift is the default position and selection has to be demonstrated. This is why we're seeing so many papers that are trying to detect selection.

Anonymous said...

You clearly use the quote function right but miss the important semantics... the proof of junk DNA is as i said far from irrefutable... (sadly this is far from an ideal world but so much more exciting for being so)

Current evidence is somewhat akin to looking to a group dynamics study that only focuses on the solution... Say we take a group with a leader and give them a problem, they solve it. Then we give another similar problem to the group but this time take away the leader and they solve the problem just as well... Now insofar as we are only looking at the final result we see little, if any difference, but we just call that <1% experiement error... From this we could assume that the leader we thought we removed was actually a nobody since we can't fully see and comprehend what happens behind the closed doors of the decision making process. Alternatively we could postulate that the group is good enough that any member can act as a stand in and cover for the lapse of leadership. There are many possibilites and many variables to look at and unless we just happen to hit on the right one in our study we can't really know what effect our removed leader had on the group. And so it is with this "junk" DNA.
Now i have not done alot of reading on this in the last few months but this is what most of them sounded like.

Lets take take the case of knock out mice. In some knock out mice studies, where the genes knocked out are known to have serious implications, the mice do not demonstrate their deficiencies. That is until they are put under certain stresses or tested in very specific ways.

If i were to run a conventional "junk" DNA test on these same knock out mice but not look into the very specific nature of their knock-out, just do the general tests and comparison with the wild type, I would come to the conclusion that the DNA i just deleted was in fact junk DNA.

This is an over simplification i know as the segements called junk do not seem to possess any expresion factors, but the point remains. The tests being performed are simply not rigorous enough to give a truly strong and complete proof that all (or even just most) of said "junk" DNA really is just garbage.

PhDP said...

matt; I think you forget at least two things;

1) The theoretical argument in favor of junk DNA. Evolution has to produce junk DNA, or the theory is simply wrong.

2) Most of the evidences about Junk DNA come from comparative genomics and the patterns we observe in molecular evolution.

Anonymous said...

It seems to me that junk DNA is a null hypothesis. It is incumbent upon those proposing a function for a particular sequence to empirically demonstrate said function. Otherwise, we assume it has none.

I think this is the argument for why junk DNA is junk, but please correct me if I'm wrong:

Large duplication and deletion events are extremely rare. Since to live organisms need a minimum toolkit of genes, a large deletion early in evolution would be lethal. So, individuals experiencing large duplication events will leave survivors while those experiencing large deletions will not. Once a large duplication event occurs in a lineage, it is exceedingly unlikely that a large deletion event will occur in that lineage subsequently. Moreover, the many small deletions that occur after that duplication will be insufficient to wipe out a significant proportion of the now useless material. So, if by chance the last common ancestor of a lineage experienced a large duplication event (perhaps whole genome duplication) and such events are so astoundingly rare that the chances of having both a huge duplication event followed by a similarly huge deletion event are vanishingly small, then what is left is a particularly junk-rich taxon - no selection required.

Additionally, small, random duplication events are unlikely to harm an organism, regardless of whether the duplicated region is useful. But, since deletions are just as random, on the occasions when they remove or disable an important gene the individual leaves no (or fewer) descendants. Thus, more (perhaps only marginally so) individuals with random additions will survive than those with random deletions. Again, this accumulates junk without selection.

Lastly, viruses leave behind scraps in the genome that are occasionally passed down to future generations, and won't be deleted as often as added for the same reason as for duplications, thus allowing for more junk to accumulate.

Over time, little bits of junk add up and make for lots of useless crap in the genome.

John Pieret said...

All right, since you are kicking this around already and since any dispute like this creates grist for creationists' mills, anyone want to take a moment or two to swat the permanently wrong Casey Luskin?

SteveF said...

John,

What Casey has kindly done is mentioned some research that is getting a lot of people rather excited. The usually reserved John Hawks says:

I think this is an extraordinarily important result. You don't see me write those words very often. This is a paper that every biological anthropologist should read.

http://johnhawks.net/weblog/reviews/genomics/evo-devo/prabhakar-2008-hacns1-selection.html

The paper concerns a region of non-coding DNA, an "enhancer" known as HACNS1. These enhancers act as controls, switching other genes on and off. In this particular case it's involved with thumbs, wrists, ankles and feet and, particlarly excitingly, in ways that are distinctively human. I'm not a biologist so I'll point you in the direction of an excellent write up here:

http://scienceblogs.com/notrocketscience/2008/09/did_a_gene_enhancer_humanise_our_thumbs.php

and the paper itself is:

Prabhakar, S. et al. (2008) Human-Specific Gain of Function in a Developmental Enhancer. Science, 321, 1346 - 1350.

Changes in gene regulation are thought to have contributed to the evolution of human development. However, in vivo evidence for uniquely human developmental regulatory function has remained elusive. In transgenic mice, a conserved noncoding sequence (HACNS1) that evolved extremely rapidly in humans acted as an enhancer of gene expression that has gained a strong limb expression domain relative to the orthologous elements from chimpanzee and rhesus macaque. This gain of function was consistent across two developmental stages in the mouse and included the presumptive anterior wrist and proximal thumb. In vivo analyses with synthetic enhancers, in which human-specific substitutions were introduced into the chimpanzee enhancer sequence or reverted in the human enhancer to the ancestral state, indicated that 13 substitutions clustered in an 81–base pair module otherwise highly constrained among terrestrial vertebrates were sufficient to confer the human-specific limb expression domain.


BTW, in another essay criticising Sean B Carroll's most recent book, Casey argued (as does Jerry Coyne interestingly enough) that most cases in the shiny new science of "evo-devo" involve the loss of a trait (an old creationist argument ultimately). As far as I can tell (and I'm not a biologist) this research shows otherwise. But Casey neglects to mention this. Which is odd.

Anonymous said...

The paper concerns a region of non-coding DNA, an "enhancer" known as HACNS1.

Since when enhancer region is a junk DNA? You seem to be equating junk and non-coding, which is obviously wrong. HACNS1 minimal sequence is extremely conserved, which alone makes it a non-junk DNA.

SteveF said...

DK,

Regarding HACNS1, all I wrote was:

The paper concerns a region of non-coding DNA, an "enhancer" known as HACNS1.

As far as I am aware, this is correct (e.g. see the abstract - "a conserved noncoding sequence (HACNS1)"). What do you disagree with here? As I say, I'm not a biologist so I'm happy to be corrected. However, I think in this instance you may have misread me.

Anonymous said...

SteveF wrote:
Regarding HACNS1, all I wrote was ... What do you disagree with here?

I disagree that the example of HACNS1 has any place in the discussion of junk DNA.

Anonymous said...

@ Phdp
while your points are generally correct, and would be great if the issue was simply over the existence of junk DNA.

The point i'm tryign to make is that there is simply just not sufficient evidence to claim that 95% of it is junk.

Afterall just being a spacer between genes or aiding in the overall stability of the chromosome means the segment of DNA has a purpose.

SteveF said...

DK,

I disagree that the example of HACNS1 has any place in the discussion of junk DNA.

OK. But I didn't mention junk DNA. In response to John's question, I described it as non-coding DNA, which is accurate.

Larry Moran said...

John Pieret asks,

All right, since you are kicking this around already and since any dispute like this creates grist for creationists' mills, anyone want to take a moment or two to swat the permanently wrong Casey Luskin?

As usual, Casey Luskin illustrates the wisdom of Alexander Pope who said 300 years ago, "A little learning is a dangerous thing..."

Nobody who is up on this subject claims that all "non-coding" DNA is junk. We are well aware of the fact that genes have regulatory regions controlling their expression. We've known that for almost 50 years. Regulatory sequences are not junk DNA as I explain in [Junk in Your Genome: Protein-Encoding Genes]. A generous estimate is that 0.6% of our genome could be regulatory sequences controlling transcription.

Casey Luskins doesn't understand any of this. He thinks that the discovery of a tiny regulatory sequence is an Earth-shattering event. It may be to him, but then any little increase in his learning is a big improvement no matter how much he screws it up.

Anonymous said...

Junk advocates are probably not going to like this paper either...

Evolution of the mammalian transcription factor binding repertoire via transposable elements
Guillaume Bourque, Bernard Leong, Vinsensius B. Vega, Xi Chen, Yen Ling Lee, Kandhadayar G. Srinivasan, Joon-Lin Chew, Yijun Ruan, Chia-Lin Wei, Huck Hui Ng, and Edison T. Liu
Genome Research published online before print

Identification of lineage-specific innovations in genomic control elements is critical for understanding transcriptional regulatory networks and phenotypic heterogeneity. We analyzed, from an evolutionary perspective, the binding regions of seven mammalian transcription factors (ESR1, TP53, MYC, RELA, POU5F1, SOX2, and CTCF) identified on a genome-wide scale by different chromatin immunoprecipitation approaches and found that only a minority of sites appear to be conserved at the sequence level. Instead, we uncovered a pervasive association with genomic repeats by showing that a large fraction of the bona fide binding sites for five of the seven transcription factors (ESR1, TP53, POU5F1, SOX2, and CTCF) are embedded in distinctive families of transposable elements. Using the age of the repeats, we established that these repeat-associated binding sites (RABS) have been associated with significant regulatory expansions throughout the mammalian phylogeny. We validated the functional significance of these RABS by showing that they are over-represented in proximity of regulated genes and that the binding motifs within these repeats have undergone evolutionary selection. Our results demonstrate that transcriptional regulatory networks are highly dynamic in eukaryotic genomes and that transposable elements play an important role in expanding the repertoire of binding sites.

Larry Moran said...

Anonymous says,

Junk advocates are probably not going to like this paper either...

I know this is going to sound trite but it's a good idea to be skeptical of everything you read in scientific journals.

You need to ask yourself whether what you're reading makes sense in light of everything else you know. Keep in mind that real "breakthroughs" don't happen very often.

As a general rule, when you see something that doesn't make sense then that's probably because it's nonsense.

I can't see the paper right now but I'll look at it tomorrow. What I'm looking for are the controls where they show that binding is significant. One good control would be to use chromatin from another species, such as zebrafish or Arabidopsis, to see if they still detect a lot of "specific" binding "near" genes. I'll also look for the references to well-studied examples of DNA binding proteins to make sure they understand that most transcription factors will be non-specifically bound to DNA most of the time [DNA Binding Proteins].