No, not at all. Summers et al have demonstrated experimentally that the minimum requirement for conferring low chloroquine transport activity to PfCRT is two specific mutations, which confirms Behe’s inference in the Edge.
For those interested, here is a quick reprise by Behe himself of the argument developed in the Edge.
Except that they don’t conclude that there are two specific mutations required.
these results indicate that PfCRT acquires the ability to transport CQ via one of two main mutational routes, both of which entail the introduction of K76T plus the replacement of an asparagine (N75 or N326) with an acidic residue.
3 Likes
Okay, but does that mean you agree with me that convergence is best understood as the opposite of divergence?
1 Like
Interesting. Name those two specific mutations, then. From the demonstration in the paper. Specifique is the French cognate, so there’s no linguistic issue.
1 Like
You just avoided mentioning that P. vivax developed resistance to chloroquine without those “multiple, specific changes”, suggesting that they might not be needed.
Have fun trying to show that the chloroquine resistance route taken by P. vivax isn’t possible for P. falciparum.
1 Like
Ok. Let me reformulate more accurately then:
Summers et al have demonstrated experimentally that the minimum requirement for conferring low chloroquine transport activity to PfCRT is two mutations entailing the introduction of K76T plus the replacement of an asparagine (N75 or N326) with an acidic residue. This result confirms a key inference made by Behe in the Edge, ie the need for multiple, specific changes in PfCRT for the development of resistance to chloroquine.
And here is Moran’s detailed summary of the several ways Behe’s argument fails.
A quote that should help explain why Behe’s article misses the mark:
I do not deny that the observed routes to chloroquine resistance were highly improbable (10-20) but I account for this low probability by trusting the results of Summers et al. (2014) who showed that four separate mutations were required for effective chloroquine resistance and the mutations had to occur in a particular order. In addition, there are several other factors that contribute to the low overall probability; the most important is the demonstration that the particular combination of mutations are probably the only possible routes to resistance.
Michael Behe tries really hard to counter my objections but as so often happens his best laid schemes go agley (awry). He never really grasps the objections to the chloroquine frequency data: it’s not that we are disagreeing with the frequency of chloroquine resistance (about 10-20), it’s that we are disagreeing with his explanation of that frequency. That’s why I was surprised to see him admit defeat but then claim that it didn’t matter
2 Likes
It doesn’t even matter if, entirely hypothetically, the PfCRT mutations identified are the only ones possible through which CQR can evolve, because it’s just a single isolated example of an adaptation with a constrained pathway from ancestor to descendant.
That result can’t just be assumed to apply to anything else, particularly when there are so many other examples of adaptations having uncountable numbers of pathways through which they can evolve.
And when it comes to convergence, we have numerous observations of adaptations that are constrained in a way that makes convergence probable, while others are much more unconstrained and can evolve in many different ways. Either way, we do not end up with a problem for convergence, nor do we need to accept appeals to CQR as applying to just any and all retrospective calculations for other adaptations observed in life.
No, Gunther Bechly can’t just extrapolate the case of CQR to imply two presumed coordinated mutations in the evolution of, say, the land to ocean transition of ancestral whales, are too constrained to have evolved in the timeframes implied by the fossil record.
3 Likes
It is more accurate to state that Behe was wrong and that chloroquine resistance did not require two specific mutations.
3 Likes
Precisely.
I think there is a lot of confusion around Behe’s “Edge of Evolution” argument, and its fundamental weaknesses often get obscured in arguments over the minutiae of peripheral issues such as the precise mechanisms by which chloroquine resistance evolved. Here is how I understand his attempted argument:
Suppose there is a beneficial trait that requires two mutations. Call them A and B. The beneficial trait does not exist until both mutations exist in the same individual organism. For the sake of simplicity, let’s assume the frequency with which the trait will arise is solely determined by the frequency with which these specific mutations will arise, and the degree of selection to which each is subject.
If both A and B are neutral with regards to selection, then once one of these mutations occur in the population its prevalence at any given time will be determined by random drift. If A occurs in one individual, then it may persist in the population and, given enough time, one of this organism’s offspring will also have the B mutation (or the two mutations will occur in the same individual thru recombination), at which point the AB genotype will now exist. Of course, this could happen in reverse order with B preceding A. Someone who is expert in pop gen can probably give reasonably accurate calculations of these odds.
However, if either A or B on their own, or both A and B, are beneficial, then the odds of the AB trait arising are greater (that is to say, the time and population size required for the trait to arise would be reduced). The reason being that the individual A or B alleles are more likely to be maintained in the population than if they were just subject to drift. The degree to which the odds are increased (i.e. the waiting time/population needed reduced) will depend on the degree of positive selection.
Similarly, if A and/or B are deleterious, then the waiting time/population needed will be increased, because the individual alleles are more likely to be lost from the population before the AB genotype occurs than if the alleles were neutral or beneficial.
Let’s consider an extreme example, in which both A and B are highly deleterious to the point that they are immediately lethal to an individual carrying either mutation. In this case, the AB genotype will only arise if both mutations arise in the same individual organism simultaneously. The odds of this happening will be determined by the square of the rates for each individual mutation. So if the odds of each individual mutation occurring are 10-10 (which according to Moran is the correct value), then the odds of AB arising are 10-20.
Which is exactly the frequency with which Behe claims chloroquine resistance has arisen in Plasmodium falciparum. OK, but so what? Let’s just ignore all the kerfuffle about malaria specifically, and presume that even if Behe was wrong about malaria there is a trait out there somewhere that only arises 1 in 1020 generations because it requires two mutations that, on their own, are highly deleterious. What does this tell us about the odds of any trait requiring two mutations arising?
Nothing. Absolutely nothing. That there may be some traits that only arise very rarely, if at all, because these traits require two mutations that, individually, are highly deleterious does not mean that every single trait requiring two mutations is equally unlikely to arise.
It is a breathtakingly inane argument that Behe makes. That it continues to be defended by some only speaks of Behe’s skill as fast-talking con man, and of his intended audience’s willingness to be deceived.
That’s how it appears to me, anyway. I admit I have not read Behe’s book, but base this on Behe’s own descriptions of its core argument and discussion of this argument by others. If I am misrepresenting or misunderstanding anything, I know I’m in the right place to be corrected.
EDIT: In re-reading the above, I realize I have committed the same error many of Behe’s critics have also made, which is to give Behe undue credit for making a more coherent argument than he actually does. To my knowledge, he has never explicitly argued that CQR is so uncommon because it requires two simultaneous (or nearly simultaneous) mutations. For that matter, he never commits to any explanation beyond the utterly trivial, non-controversial and unhelpful point that it requires at least two mutations.
All his argument boils down to is “CQR evolves once in 10 20 generations. Therefore, any trait requiring two mutations to protein coding genes evolve once in 10 20 generations.” When it was pointed out that CQR likely does not require two simultaneous mutations (and when it was confirmed experimentally that it does not) Behe response was that this does not matter, because CQR evolves once in 10 20 generations. Which, OK, sure. But what argument does he provide for his assertion that this frequency applies to all beneficial traits requiring two mutations? None. He doesn’t even try. He just spews out a bunch of irrelevant bafflegab, for instance in the article @Giltil links above, confident that the people he is trying to convince will not notice or care.
Basically, it’s the Chewbacca Defense.
3 Likes
That’s not accurate at all. Behe inferred that two specific changes were required. The paper falsifies that.
Behe inferred that those two specific changes were required to occur SIMULTANEOUSLY. The paper falsifies that.
You should compare Behe’s original words to the data, not Behe’s words to Behe’s words sneakily written after moving the goalposts. Remember:
You’re obviously avoiding the data.
2 Likes
No, I would only agree that a better term would be helpful, as evolutionary convergence means converging on the same phenotype from different places using at least some different changes.
Flight in bats vs birds qualifies, even though front limbs are involved in each, because wings are constructed differently. That’s a very different phenomenon from two Plasmodium species evolving altered transport selectivity.
K, well then let me just state I find your view highly idiosyncratic, I strongly doubt it has much support within evolutionary biology, and I will just continue to consider convergence the opposite of divergence. Among other good reasons, because it also makes the sense I have in mind consistent with how the word is used in everyday speech.
Could you correct your sentence that, as it is, is nonsensical (a single mutation requiring two mutations is indeed unlikely to arise!)
Convergence is often defined as the evolution of similar states beginning from different initial states. But the changes themselves don’t have to be different. The evolution of similar states beginning from the same initial state is often differentiated as parallel evolution. But as far as I know neither is ever used to refer to single mutations, especially single nucleotide changes, and is never used to refer to separate changes in a single population or species.
3 Likes
That there may be some traits that only arise very rarely, if at all, in organisms with very large populations and short generation times because these traits require two mutations that, individually, are highly deleterious does not mean that every single trait requiring two mutations is as unlikely to arise.
That being said, it’s a gross oversimplification, because in this case, selection for chloroquine resistance is only present during part of the Plasmodium life cycle, in humans. No one is treating rats with chloroquine.
Of course, if things are already identical, they can’t converge any further. For convergence to happen, things already have to be different in some way. But things can be different within the same population that doesn’t consist entirely of clones.
Sure.
It may be true that that is the typical usage, but I don’t see why not, if whatever the “initial” state is, is not identical. That would then qualify as convergence (the genotypes, sequences, molecules, or whatever, have become more similar, instead of more dissimilar, so converged instead of diverged). If we can use divergence to describe the becoming more dissimilar of genotypes, then we can use convergence to describe the becoming more similar.
Incidentally, I’m not the only one to say this idea that it’s only convergence if it happens from “distantly related” points, is vague and ill-defined too. They go even further and argue that even parallelism should just be called convergence.
It has over 600 citations, and sampling a few of them indicated it’s being hotly debated how to use the terms with little consensus on what is or should be the right usage. Call me weird but it seems so simple to me. Again, things can diverge(become more different), stay the same, or converge(become more similar), and that’s regardless of whether we are talking phenotypes or genotypes.
They are needed in P falciparum. They are not needed in P vivax because this bug is a different species that has a different biology. Not very hard to understand.
Given that in science it’s well known that it’s almost impossible to prove a negative, I’m not going to take up your silly challenge. However, if it happens that the chloroquine resistance route taken by P. vivax is possible for P. falciparum, then what can still be said is that it’s a very improbable event no higher than 1 in 10^20. So your P vivax distraction does nothing to meet the challenge posed by Behe in the Edge.
Yes. To put it another way, Behe pretends that the only relevant factor is the frequency with which the mutations occur. He ignores that the mutated gene must also go to fixation, and all the factors affecting this. Given that the malaria parasite has an unusually complicated life cycle, any generalizations that can be drawn from it will be limited.
But, again, let’s not get distracted by Behe’s attempts at diversion. If CQR in P. falciparum does not require two simultaneous mutations, maybe something else in some other species does. If so, so what? How does this demonstrate that all beneficial traits requiring more than two mutations lie beyond the “edge of evolution”?
Behe has no answer beyond “But CQR in P. falciparum occurs with a frequency of 10-20.” Which is the equivalent of “Chewbacca is a Wookiee from the planet Kashyyyk. But Chewbacca lives on the planet Endor.” Or, to mix our pop culture references, “These go to 11.”
I just remembered that there is another problem with Behe’s argument: He commits that old favourite of the ID Creationists, the Texas Sharpshooter Fallacy. All he has calculated (or, more accurately, lifted from someone else’s paper) are the odds of one specific set of mutations occurring in one specific organism. The very definition of the Texas Sharpshooter.
3 Likes
The challenge is meaningless, because there’s just no reason to suppose some other adaptation we could have in mind is similarly constrained to how CQR might be.
It’s a totally vacuous statement in the end. If an adaptation that can only evolve through a series of mutations this complex, takes this long to evolve in a population of this size and mutation rate, then if the population size and mutation rate was much smaller, it would take much longer for that same adaptation(or one of similar constraint).
Yeah… and?
4 Likes