I am loath to resurrct yet another old thread, but I’ll link to the posts for context.
From the NY Times: RNA Trades Bit Part for Starring Role in the Cell
So the question is: does this cast doubt on the claim that junk or non-coding DNA is actually not being used? It would seem to me that it does, but I’m no molecular biologist.
Also, in general, how did the idea ever come about that the pseudogenes were non-coding? Was it through actual experiments, or simply theory?
A few points (and I might garble this, since I have a cold.)
1.) Your article is not about pseudogenes. It’s about a different form of junk DNA.
2.) Pseudogenes can’t code for protein, because they contain stop codons. Nevertheless, they look exactly like protein genes otherwise.
3.) I’ve never claimed that pseudogenes are proof of evolution because they serve no function. Let me repeat that, because I didn’t make it clear last time: I am not using the vestigial organs argument. All I’m saying is that here we have some genetic features which evolution can explain, and creationism cannot. Just waving your hands and saying, “that’s the way it is because God made it that way to serve some function, but we don’t know what the function is or how those features serve it” isn’t science. If you want to claim that pseudogenes code for RNA instead of protein, you need to explain how the properties of pseudogenes relate to that function.
Hey, i met that guy when i worked there!
a lot of junk DNA we don’t know what it does or if it does anything. A lot of the coding regions we don’t know what they do. We are still working on the genome, and will be for many years figuring out what codes from where and what this mutation does to that. And then there is extra junk from retroviruses and whatnot mucking up the works. plus looking at ACGTTTGTGCGCGCTGAACGTGACGTACGTACGTA all day gets pretty boring and i spend time reading this site instead.
Gimme a year or two and i’ll figure everything out;)
Thanks for the correction here. I was unaware of the existence of different types of junk DNA.
I did not say that you made this claim. As I understand it, your claim is that because pseudogenes serve no purpose, there would be no reason to expect them to be similar in similar animals, unless they shared common ancestry. But this is dependent on the notion that they indeed serve no purpose.
Well it might not be “science”, but might be true. I would think the burden of proof is on the one making the claim. In this case the claim is that the fact that these genes serve no purpose proves (indirectly, through similarity across species) that there is common ancestry. And my point was that to the degree that we cannot say with certainty that these genes are indeed junk, any use of this assumption to prove anything else will fail. So the fact that I (or anyone else) cannot say with certainty what exactly they do is not the issue - the issue is if indeed we do know with certainty that they do nothing.
So: do you agree that if indeed it turns out that much of what had been thought to be non-coding is actually serving a purpose, it would cast some doubt on the amount of certainty with which we can say that non-coding DNA is “junk”? And again, just how has it been established that the stuff was junk to begin with?
mostly because people didn’t see it coding for anything, and there was a lot of it not coding for anything (~90 to ~97%, depending on source), and a lot of it repeated, so they just thought it was useless garbage.
link from Wired, of all places.
another linkinvolving inheriting junk DNA, i haven’t read the whole website, but that part seemed fine
The problem with your analysis, IzzyR, is that “pseudogenes serve no purpose” isn’t an assumption. It’s a theory.
Let’s contrast:
Formulated in that fashion, it’s an example of the fallacy of argument from ignorance. It basically says, “We can’t find a use for this, so it must have no use.” But you can’t really say for sure whether it’s the case that you just haven’t found a use yet, and that a use might be found in the future.
BTW, I’m not sure that that formulation of the vestigial organs argument isn’t a creationist strawman. A much better version of the argument goes like this:
Here’s an organ which looks like something else, but has been jerry-rigged to serve a different purpose. Since God could make any organ from scratch to perfectly serve its purpose, this jerry-rigging proves evolution.
It’s what I call the “Isn’t it funny?” argument. Evolutionists point out that whales have vestigial legs. Creationists declare that those legs aren’t “vestigial”- they serve a purpose! They are support structures for the whales’ genitals. Isn’t that funny? The perfect genital-supporting structure just happens to look like a shirvelled-up leg, not an arm, not a wing, not a skull, but exactly what evolution would have jerry-rigged into service.
By contrast with the classical vestigial organs argument (i.e., it serves no purpose, therefore creationism is wrong) we have the pseudogenes argument:
Yes, the evolutionist explanation happens to involve pseudogenes serving no function, but there’s a world of difference between argument #1 and argument #2.
In argument #2, we just don’t say, “See? look- we can’t find any function. Creationism is wrong.” We say this:
here we have some things that look like extra copies of regular genes.
They have all the earmarks of normal gene-copying processes, such as (in the case of processed pseudogenes) the marks left by splicing RNA back into the genome.
Ok, my hypothesis at this point is that these genes were produced by normal copying processes, instead if by creation ex nihilo. How can we test this? For one thing, there’s the question of the arrangement of genes in the hemoglobin cluster, which I discuss in the FAQ. Also, so far as I know, retroposition of RNA has been seen in vitro, and it’s definitely been seen in a clinical setting. So clearly we have good reason to think that pseudogenes arose by natural processes.
We also find:
From which we expand our hypothesis: these must be extra gene copies which got switched off by those mutations. Therefore, they don’t serve a coding function.
How can we test this? If it’s right, then the mutation rate should be higher for pseudogenes that for regular genes, and should be comparable to the mutation rate for introns and other noncoding regions. And, as a matter of fact, it is. (For that matter, the mere existence of consistent mutation rates proves common descent, because the rates are measured across different ‘kinds’ with reference to the fossil record.)
And, since we don’t know of any function pseudogenes could possibly serve, we can conclude that they serve no function.
Now, your news story reports that someone has found that one class of junk DNA serves a purpose. So? What does that have to with pseudogenes? If that class of junk DNA has a mutation rate that is comparable to that of pseudogenes, then it says something about the constraints on noncoding regions, and we’d have to posit that pseudogenes might serve a noncoding function which was subject to very loose evolutionary constraints. But that’s just a minor modification to what we’ve said thus far: pseudogenes are copies of normal genes, produced through naturalistic physical processes, which have had their coding function destroyed by mutations, and which therefore accumulate mutations more quickly than real genes, in a pattern consistent with the fossil record. And, BTW, creationists still can’t provide an explanation for pseudogenes.
I think the vestigial organs argument is a different issue. But your response to the pseudogenes issue seems to be this:
I gather that you are saying that you have an explanation and a logic for how it would come to be there with no purpose, as you later elaborate. OK. Still, call it an assumption or call it a theory, it just might be wrong. There’s theories and there’s theories, as you know, some stronger than others. As I’ve noted, the field is relatively young, and much is not understood at this point. And if it turns out that the theory about one set of non-coding genes was wrong, that kind of makes you wonder about the parallel theory about pseudogenes.
I see that in the link provided by Tars Tarkas, the claim is made that:
That article is about 2 years old, so I guess my article is not a revolutionary as I had thought.
The following should be noted.
“Pseudogenes (b. 110 BC-d.71 BC). Native of Polycarpum. Illegitimate and disowned son of the Lesser Demosthenes, half-brother of Mutagenes, father of Endogenes. Minor lyrist of the late clasical period. Only extant work ‘Ode to a Grecian Urinal’.”
–from Hektor Bobalides, MEDIOCRITIES OF HELLENIC ANTIQUITY
I thought the name rung a bell.
IzzyR, why don’t you believe in evolution?
I think Izzy is questioning the logic of your argument, not the validity of your conclusions. Just because he accepts evolution on the basis of the entirety of evidence available does not mean that each bit of information presented as evidence is necessarily true.
In fact, since you don the cloak of Venerable Scientist when you speak out against creationism, it is much more important that you make no claim at all that is not firmly based on good evidence, and valid logic. If you are willing to accept less, you are just describing how your religion differs from someone else’s.
So, you have to find out if mutation rates for RNA encoding genetic regions are the same as introns, and pseudogenes. And, if you really want support, you will have to show that they are, in fact more like the rates for protein encoding portions of the genome. Given that, you have your argument back, and even stronger. Failing that, you have theology.
Tris
**
I think Izzy can describe his own beliefs, if he’s really interested in continuing this conversation.
Excuse me? Why is it my job to waste time combing through the literature in order to rebut the arguments of someone who understands the subject matter so poorly that he didn’t even realize that the study he cited wasn’t about pseudogenes? If IzzyR wants to rebut my arguments about pseudogenes, it’s up to him to provide the relevant information about mutation rates. As is, he’s basically waved his hands vaguely and declared that maybe one day we’ll find out that the current understanding of pseudogenes is wrong, even though at the moment there’s no direct evidence that that’s the case- and never mind that he hasn’t even addressed the fact that the question of whether or not pseudogenes serve a function is a fairly minor aspect of how they serve as evidence of evolution.
I’ve discussed evolution in previous debates. But here I was hoping to address one specific argument. The difficulty in doing that is part of the reason these boards are less then ideal for discussing evolution IMHO - it is too much of a passionate crusade for many people.
It is not your job to do anything on these boards, AFAIK. As you wish. I do appreciate the information you have occasionally provided, however (e.g. your FAQ about DNA etc.)
I think I’ve addressed all points you have raised here about pseudegenes.
IzzyR, let me make sure we understand each other.
Do you agree that even if pseudogenes were found to serve a function, they would still constitute evidence for evolution and against special creation?
I don’t see why this would be so. Though I imagine it might depend on what that function might be.
Pseudogenes are a class of genetic sequence. Different classes give consistent numbers when you calculate their mutation rates, based on the fossil record.
If special creation were true, then the fossil record would be bunk, and the numbers you get should be random. Thus the mutation rate calculated for a pseudogene when you compare, say, horses and frogs should be wildly different than the value for the same pseudogene between, say, zebras and salamanders.
In any event, your original article answers the question about mutation rates as an indication of function:
“Areas of the genome that are similar are thought to have important functions, explaining why they have not mutated as species evolved. At least part of this overlap appears to be genes that produce RNA as their end product.”
In other words, in the case of these RNA-coding regions, their mutation rate was markedly less than that measured for pseudogenes. Thus the reason they were suspected of having a function is precisely that they didn’t act at all like pseudogenes do.
Assuming that the differences between Species A and Species B represent a number of mutations assumes that they share common ancestry. So one cannot talk about “mutation rates” unless one already accepts that they began from a common starting point. Otherwise its just differences.
If pseudogenes are completely useless, then there might be no reason for a horse and a lion (or any other random species) to differ any more than a horse and salamander. But if they serve some purpose, then the argument that these reflect external differences falls back into place. And since horses and frogs differ (from each other) about as much as do zebras and salamanders, one might expect the number of molecular differences to be about the same.
Well the article does not actually compare them to pseudogenes. But if this is indeed what was meant, then you have a point that you cannot easily extrapolate from one to the other.
But the problem is that you’re just rehashing the same old creationist misunderstandings of the evidence. In fact, you’re doing so without even trying to address the real substance of my point.
Firstly, mutation rates for particular classes of genetic information are actually more stable if they don’t serve a coding function. Protein-coding genes are under tight evolutionary constraints, and react strongly to changes in the environment. You have to do a lot more careful work to calibrate the “molecular clock” in such cases, although my original point still holds for protein genes. This is not at all what one would expect if, as you say, differences in genetic material just reflect overall anatomical or functional differences. (For that matter, you find that in times of greater selection pressure, the ratio of silent to non-silent mutations changes. And, of course, the mutation rate for silent mutations within a protein gene is roughly the same as the mutation rate for pseudogenes.)
Secondly, as I’ve said time and again, overall genetic difference doesn’t correlate to anatomical differences. Birds, for example, are genetically a subgroup within the reptiles, precisely as one would expect from the fossil record. Creationist classification schemes have always made birds and reptiles separate groups. What the genetic studies do correlate to is the fossil record. In the example above, you could, instead of using zebras as a point of comparison, use another close relative of horses: rhinos. You’ll still get a consistent mutation rate, even though rhinos are a lot more different from horses than, say, deer are. In fact, the big thing that horses and rhinos have in common is that they have an odd number of toes. Funny how the number of toes makes all the difference for their pseudogenes, eh?
And, as I said, you haven’t even addressed my real point. When you calculate mutation rates you aren’t using numbers based on some sort of overall similarity. You take the number of differences between the pseudogenes in the two different species, and you divide that by twice the time that has elapsed since those two branches of the evolutionary tree diverged, according to the fossil record. If common descent is bunk, then the fossil record should have no correlation to the number of differences between pseudogenes. But, as it turns out, using the fossil record makes all the numbers come out consistently.
This is, of course, entirely aside from the question of processed pseudogenes. They prove evolution for the same reason that retrogenes do: they bear all the earmarks of being mRNA that was spliced back into the genome.
Could you explain a little more this bit about calibrating the molecular clock?
A bit more here as well. How is it possible for genetic differences to not correlate with anatomical differences? Aren’t the genetic differences solely responsible for the anatomical differences? I can understand if it should turn out that a giant panda is closer genetically to a lesser panda than to a polar bear despite seeming superficially closer to the latter. But that would only be an indication that superficial observations are not everything. I imagine that if you examine them closely enough you would find that the giant panda function internally more like a lesser panda than a polar bear - the very result of genetic similarity. So all that remains is that the human brain wraps itself around superficialities at first glance. This is not significant.
This makes no sense at all, unless I’m misunderstanding you here. If you were to say that horses differ from rhinos as much as much as horses differ from zebra, and say this is consistent with the fossil record but not external similarities, you’d be making a valid point. As it is you are saying that horses differ from salamanders as much as rhinos differ from salamanders. This says nothing at all about the similarity between horses and rhinos - they could be radically different - only that they both differ from salamanders by about the same amount.
Here’s where you might be making a point. Assuming that I understand your point correctly. I assume you are saying the following:
The fossil record shows that birds and reptiles are more closely related than are either one of them and mammals. This is because animals that are thought to be intermediate bird/reptiles have been found in more recently forms geological material than animals that are thought to be intermediate mammal/reptiles. But yet, today, birds and reptiles are as different from each other as each is from mammals. This is because they have had more meaningful mutations in the interim that have separated them. This, because the rate of meaningful mutations, subject as they are to survival pressures, can vary between species and circumstances.
However, the rate of meaningless mutations, i.e. mutations to non-functioning DNA (pseudogenes) is not subject to survival pressure, would follow a constant pace for all species. Thus, while the later-separated birds and reptiles might have the same number of meaningful DNA differences as the earlier-separated reptiles and mammals, they would be expected to have fewer pseudogene differences, in keeping with the shorter timespan in which these arose. This is indeed the case.
If this is indeed your argument - and if it is all true - you have a valid point.
One question that I would ask you is about your claim that the pseudogene mutation rate is in line with evolution. What would be your reaction if it should happen to turn out that pseudogenes have a function after all?
IzzyR, I’ll write a more detailed reply later, but for now I want to say that you consistently seem to be confusing “mutation rate” with “number of mutations.”
If you compare, say, horses and rhinos (closely related) with horses and deer (more distantly related), you find that pseudogenes contain more differences between horses and rhinos than they do between horses and deer. But the mutation rate for their pseudogenes is the same. Why?
Let’s say that for a particular pseudogene horses and rhinos have 10 differences, while horses and deer have 40 differences. Horses and Rhinos diverged 5 million years ago. The mutation rate is thus 1 difference per million years (because we divide the number of differences by twice the time elapsed.) Horses and deer diverged 20 million years ago. The mutation rate for them is thus 1 per million years, too. (Obviously I’m pulling numbers out of thin air here, just to illustrate the principle.)
My point here doesn’t really have anything to do with number of mutations versus overall similarity. Nowhere in this do we use some sort of measure of “overall similarity” to calculate the mutation rate. We just use number of differences, and elapsed time in the fossil record.
If the dates calculated from the fossil record don’t represent a real divergence (for example, if YEC’s are correct, and it’s all hydrodynamic sorting) then the values of “5 My” and “20 My” in my above example aren’t correct. But if they aren’t correct, then why do they give consistent results? The problem is really the same (although less obviously so) for OEC’s. If God decided to make new species pop into being from time to time, and the fossil record accurately records the sequence of events, why did he make it happen in such a pattern that the pseudogene calculations would give consistent results, when the pattern of appearances really has nothing to do with accumulation of changes over time?
As for your question of how physical similarity can be separate from genetic similarity, let me say first off that the “why” is a separate issue from the fact that it’s clearly true. (In fact, creationists sometimes whip it out as an anti-evolution argument, saying that frogs are more genetically diverse than mammals as a whole or some such, but it’s not clear to me what their point is.)
But the “why” is actually pretty simple. Only non-silent mutations are reflected in anatomy. Like I mentioned before, in times of high selection pressure, the ratio of voiced to silent mutations changes. Birds accumulated a lot of changes where it counts, while they accumulated silent mutations at the ordinary rate. Thus they’re genetically more similar to alligators than turtles and snakes are, but they’re so different that creationists thought for centuries (ever since Linnaeus) they were a group outside the reptiles.
You seem to be approaching this with creationist logic, by which if the pseudogenes serve a function, then every single basepair is important (because it was made to have a specific sequence by God) and thus all sequences should be strictly conserved.
Lots of “junk DNA” serves a function. If I understand the article you cited, it’s about microsatellite DNA, and that has long been known to serve a function in some circumstances. Introns serve a function. But none of these things serve coding functions, so there are not strict constraints on their sequence. If pseudogenes were found to serve a function, I’d be surprised, but that wouldn’t affect them as evidence of evolution unless their function implied a degree of conservation incommensurate with their mutation rate (which, as is, is similar to that of introns.)
In fact, some pseudogenes do serve a function, if you want to get technical. Creationists love to bring up alu elements as “proof” that pseudogenes aren’t evidence for evolution.
Why do they choose alu elements in particular? It’s one of the few- perhaps the only- pseudogenes that is also a transposon. Everybody knows that transposons serve a function. (The function, ironically enough, is to enhance the ability of the organism to evolve.) It’s like “proving” that all Germans are Nobel Laureates by pointing to Einstein. He was German, right?