The intelligent design of evolution
by bipodAccording to at least one of the commentors (sp?) on our blog, the causes of these experiments cannot be discovered and will not be discoverable experimentally (that's the fallacy of intelligent design)…
The intelligent design of evolution (link takes you to Nature article)
Mark P Styczynski, Curt R Fischer and Gregory N Stephanopoulos
The debate between intelligent design and evolution in education may still rage in school boards and classrooms, but intelligent design is making headway in the laboratory. In this case, though, the designer turned out to be just some clever scientist. A recent paper in Nature (Yoshikuni et al, 2006) presented the iterative evolution of highly specific catalysts from a promiscuous wild-type enzyme via what the authors refer to as designed divergent evolution.
The paper investigated whether catalytic functionality could be rationally engineered into a protein, without recourse to the high-throughput screening techniques necessary for directed evolution. Yoshikuni et al (2006) started with a terpene synthase enzyme,?-humulene synthase, that is promiscuous not in its substrate specificity but in its product selectivity?it catalyzes the formation of 52 different sesquiterpene products from one single substrate, farnesyl diphosphate. (Sesquiterpenes naturally occur in a variety of plants, and their derivatives are used in applications ranging from chemical feedstocks to antifungal compounds.) The predominant product for the wild-type enzyme is?-humulene, but Yoshikuni et al designed seven mutant variants with improved selectivities for eight of the products.
May 4th, 2006 at 7:58 pm
bipod, interesting article, and reminiscent of the intelligent guidance that is used by pharmaceutical companies to search large spaces for therapeutically-useful compounds.
The comments in the article about irreducible complexity seemed completely gratituous and uninformed. Presumably a weak attempt at keeping ID folks from looking favorably on the paper?
Comment by Eric Anderson — May 4, 2006 @ 7:58 pm
May 4th, 2006 at 8:33 pm
From Nature:
Wow. Thanks, bipod! This appears to validate the notion that thinking in design terms can indeed provide useful information and avenues of investigation. The cited "limitations" of this approach do relate to the applicable substrate and its nature (how reducible it is for this application).
So I can't say I'm surprised that "irreducible complexity" of functional constructs got replaced with "reducible simplicity" in the classes of individual proteins this design approach is expected to work with.
Comment by Joy — May 4, 2006 @ 8:33 pm
May 5th, 2006 at 10:48 am
[...] File this under, "I fail to see the point". Some scientists used a bit of design ingenuity in the lab? Big deal. You can't argue from "they did it" to "some unsubstantiated, super-being did it". Besides, what these guys have done is not, in principle but only in complexity, different from what humans have done for thousands of years with domestic plants and animals. Of course we can design stuff. We do it all the time. That's hardly the point. [...]
Pingback by hell’s handmaiden » Telic Thoughts in the Lab — May 5, 2006 @ 10:48 am
May 5th, 2006 at 1:12 pm
Bipod,
1) The link doesn't take you to the Nature article; nor is there a link to the Nature article in the commentary you linked to. Did you even read the Nature article? Does commentary trump data? Didn't the authors of the Nature article use saturation mutagenesis and artificial selection? Doesn't this show how easily nature can do the same thing by RM+NS?
2) As someone who intelligently modifies a protein family that has been intensively studied for 100 years (we mutate its active site to allow the hydrolysis of a synthetic substrate without losing its ability to hydrolyze its natural substrate), I don't see how you get the induction that the protein must have been intelligently designed because we can intelligently modify it. I don't view our strategy as particularly sophisticated; I can describe it sterically to a ten-year-old.
Would you (or Joy) please explain this induction to me in detail?
3) Do you realize that the data from these approaches demolish the underlying assumption of all probabilistic attacks on MET–that conservation of a residue among species allows one to conclude functional constraint?
Comment by Smokey — May 5, 2006 @ 1:12 pm
May 5th, 2006 at 1:56 pm
Didn't the authors of the Nature article use saturation mutagenesis and artificial selection? Doesn't this show how easily nature can do the same thing by RM+NS?
I love the easy conflation of blind, unintelligent natural selection with intelligence-based artificial selection here.
There will be no scientific understanding of intelligence in evolution until there is an understanding of what intelligence is.
Comment by MatthewCromer — May 5, 2006 @ 1:56 pm
May 5th, 2006 at 2:02 pm
Matthew,
There was no conflation. Moreover, the "blind" aspect of MET is random mutation, not natural selection. Do you know what "saturation mutagenesis" means?
Comment by Smokey — May 5, 2006 @ 2:02 pm
May 5th, 2006 at 3:49 pm
Addressing the comment from left field:
"I don't see how you get the induction that the protein must have been intelligently designed because we can intelligently modify it"
This is OK, because that is not the line of reasoning. No-one is arguing that because system x can be modified to become system y, that we therefore conclude that system x must have arisen from intelligent input.
What we do observe, however, is that system y arose from intelligent input, not an unguided "evolutionary" process. Thus, the formation of system y (if we are talking about a biological system) can be viewed as a live example of an intelligent agent creating a biological system (which thus refutes one of the silly and over-used complaints about intelligent design: "we have never seen anyone design a biological system, so you cannot analogize to non-biological systems"). Furthermore, it demonstrates that system y, a known biological system after all, indeed *can* come about through intelligent guidance, as presently demonstrated, which serves to underscore the complete lack of evidence that system x came about by unguided natural processes.
It doesn't "prove" anything about system x. But it does clearly indicate that intelligent guidance is a live possibility for biological systems, which may also raise our level of skepticism about non-intelligent origins for system x.
Comment by Eric Anderson — May 5, 2006 @ 3:49 pm
May 5th, 2006 at 3:49 pm
So now natural selection is not blind?
Time to get Dawkins to reissue his classic. . .
Comment by MatthewCromer — May 5, 2006 @ 3:49 pm
May 5th, 2006 at 6:09 pm
"Thus, the formation of system y (if we are talking about a biological system) can be viewed as a live example of an intelligent agent creating a biological system…"
No "system" is being created. A single enzyme is being modified by saturation mutagenesis. There's miles of difference between the two.
"… (which thus refutes one of the silly and over-used complaints about intelligent design: "we have never seen anyone design a biological system, so you cannot analogize to non-biological systems")."
No, you can't, as we're merely modifying a single enzyme, not designing a system. Your conflations of "design" with "modify" and "enzyme" with "system" are comical.
"Furthermore, it demonstrates that system y, a known biological system after all,…"
Single enzymes aren't systems.
"… indeed *can* come about through intelligent guidance, as presently demonstrated,…"
The intelligent guidance involved was picking the residues to randomly mutate. Again, it's just a single enzyme. Using intelligence and prior knowledge to pick the residues merely reduces the work and time involved; nature doesn't have that limitation. We picked only a single residue. Would you call that "creating a system"
"… which serves to underscore the complete lack of evidence that system x came about by unguided natural processes."
There is abundant evidence that biological systems arose through natural processes, starting with the gigabytes of sequence evidence ignored by evolutionary critics. Every new sequence provides an independent test of MET.
Comment by Smokey — May 5, 2006 @ 6:09 pm
May 5th, 2006 at 7:01 pm
Wow, look at all those new emoticon options today!!
Let me comment, for the completely non-random reason that one of the authors of the article is named 'Gregory,' as am I, that this paper shows how un-sexy the topic of both evolution and 'i+d' really are when it comes to nuts and bolts. 'Designed by a scientist' is quite different than 'designed by an unembodied intelligence.' I can already hear Rock's LOL's coming in waves.
Where's the intelligence (it might help to sell copies, increase citations)? Isn't it just engineers and graphic artists who 'design' things? Nobody's really even speaking about 'design,' are they? Wasn't it a teleological accident, a meaningful scientific discovery, something that happened randomly, with a purpose?
Arago
Comment by g arago — May 5, 2006 @ 7:01 pm
May 5th, 2006 at 8:24 pm
Smokey, I did not make this as explicit as I perhaps should have, so allow me to clarify: First, I use the word "system" in a broad sense. Second, I am not trying to comment on the purpose of bipod's thread nor whether or not the particular item reported on in the article should be viewed as such. Rather, I was responding to a comment you made that appeared to be unrelated to bipod's and Joy's points. Further, I was addressing a common misconception/allegation: that analogies to human designs are off limits because we have never seen any human design life. While there is still a long way to go, that allegation is growing more and more tenuous.
As for the gigabytes of sequencing information, I presume you are referring to genome sequencing — containing that information rich code that arose from, what do you imagine, fortuitous chemical accidents?
Comment by Eric Anderson — May 5, 2006 @ 8:24 pm
May 5th, 2006 at 9:47 pm
It's pretty amazing how IDists choose to totally ignore the interesting and obvious.
Terpene synthases, as a group of enzymes, are positively amazing. At the center of the enzymatic mechanism is the very transient production of a positively-charged carbon center (a carbon cation). This cation is stabilized every so briefly by interaction with the pi electron cloud of a nearby aromatic amino acid side chain in the enzyme's active site. Without this interaction, no cation forms, and nothing happens.
The cation is highly reactive, and rather non-specific (even random LOL). The end product(s) are determined by the geometry of the substrate as it is held in the active site - those parts of the molecule in proximity will react, and the initial reaction products will resolve accordingly (just how depends on the initial event, and the subsequent steps can be several and complicated). All of this happens very quickly, probably before the enzyme can change conformations. So the product profile is determined by how the substrate is held initially.
Terpene synthases are not all promiscuous; indeed, most give rise to one or a few products. But as a group they are responsible for chemistry that puts to shame to full sum of all synthetic organic chemists in history. What is important about the paper bipod cites, from an evolutionary as well as ID perspective, is that it shows that the evolution of new chemistries is a simple process "“ it doesn't require massive re-structuring of the enzyme, tens or hundreds of simultaneous changes, anything of the sort. As one should expect when one considers the chemistry, chemical macroevolution of this sort needs as few as single amino acid changes. This paper shows that the entire universe of terpenes is accessible to RM+NS, and that this evolution is decidedly a zero-CSI proposition.
There's nothing ID friendly in this story.
Comment by Art — May 5, 2006 @ 9:47 pm
May 5th, 2006 at 9:51 pm
Comment by Rock — May 5, 2006 @ 9:51 pm
May 5th, 2006 at 9:54 pm
Yeah, I forgot
:twisted::evil::mad::razz:
Comment by Art — May 5, 2006 @ 9:54 pm
May 5th, 2006 at 10:18 pm
The Intelligent Design/Evolution debate, in smileys:
Comment by Douglas — May 5, 2006 @ 10:18 pm
May 5th, 2006 at 10:20 pm
(Alternating participants.)
Comment by Douglas — May 5, 2006 @ 10:20 pm
May 5th, 2006 at 10:24 pm
Eric, my comments are highly related to bipod's post and Joy's comment, because bipod was quote-mining from a commentary and apparently hasn't bothered to read the actual Nature paper. Despite what was in the commentary, Yoshikuni et al didn't really design the mutants. They intelligently identified the residues that should mediate specificity, and subjected them to saturation mutagenesis (i.e., blind). Then, they did artificial selection to pick the winners. We picked out a single residue and made the most radical substitution we could in terms of size, but not charge. We then screened a library of synthetic substrates to select the one that was hydrolyzed by our mutant, but not our wild-type control (>100-fold selectivity).
Both cases are intelligent modification, not design. If I put those fashionable blue bulbs in my headlights, have I really done any designing?
It doesn't matter how broad your use of "system" was, it was simply wrong to conflate a system with a single enzyme. The analogies are ludicrous; I don't view any of my science as playing God. If you want to make an analogy between me and the designer of life, how can you portray yourself as understanding this alleged design better than I, an alleged practicioner, do? What are your published accomplishments in the field of intelligent protein modification?
As for the sequences, I'm referring to the relationships between the gigabytes of sequences. Every time we sequence something new, MET predicts these relationships and has the potential to be falsified. ID avoids discussing these relationships because if one assumes design, they clearly say a great deal about the nature and tendencies of the designer(s), and politically, we can't have that. So once again, ID proponents actively avoid doing real science because they are afraid of learning. So, I'd like you to summarize the underlying basis for your claim that there is a "complete lack of evidence," since to ethically make such a claim, you'd have to be very familiar with every bit of evidence ever offered.
"Fortuitous chemical accidents" is laughable, particularly because the relationships can be gleaned by limiting our consideration to totally silent (i.e., nonfortuitous) sequence differences. Moreover, we get the same relationships between proteins or organisms if we consider only the silent or only the nonsilent changes. What does that say about a designer? Are you denying the existence of genetic drift?
Finally, MET predicts that we will still find clear structural relationships between protein families, even as the sequence homologies vanish into statistical insignificance. Precisely what does the notion of ID predict about these relationships, Eric?
Comment by Smokey — May 5, 2006 @ 10:24 pm
May 5th, 2006 at 10:29 pm
Art, I was trying to lead them to the same conclusion more slowly.
These (and our) experiments show that it is incredibly easy to derive new functions without (ours) or with (the pseudocited paper) losing existing functions.
The assumptions underlying the probabilistic attacks on MET are bunk.
Comment by Smokey — May 5, 2006 @ 10:29 pm
May 5th, 2006 at 11:04 pm
Douglas, I think I resemble that remark.
Smokey, sorry for being impatient.
Comment by Art — May 5, 2006 @ 11:04 pm
May 7th, 2006 at 11:44 am
Smokey:
Actually, you left out an important piece, they then used the data collected from the single mutations to recombine them and constructed enzymes with high product selectivity. The design methodology described in the commentary and used by the researchers is called "adaptive walking", and bipod's point is that this methodology can likely be gleaned from the experiment.
Comment by Guts — May 7, 2006 @ 11:44 am
May 7th, 2006 at 11:51 am
Art:
It's probably far more complicated since similar protein folds can lead to different products in different enzymes, and the enzyme must trigger the generation of the cation and then control when and where it is quenched.
Comment by Guts — May 7, 2006 @ 11:51 am
May 7th, 2006 at 12:31 pm
Actually, you left out an important piece,…
I left it out because I was correcting bipod's & Joy's initial misunderstanding of the method.
…they then used the data collected from the single mutations to recombine them and constructed enzymes with high product selectivity.
"Intelligently combining them" still doesn't get you to "intelligent design," Guts.
The design methodology described in the commentary…
Again, bipod clearly only read the commentary and didn't bother reading the paper, falsely claiming that he had provided a link to it. The commentary was inaccurate.
The only "design methodology" here is experimental design. They were simply testing the hypothesis that the changes were additive. They were, which was surprising, which is why the paper that bipod didn't bother to read made it into Nature. As Art says, there's no good news here for ID. Our data make it even worse: we changed an almost completely conserved residue in the active site that contacts the substrate to something not found in nature, without losing the original function. This demolishes a major assumption underlying all the probabilistic attacks on MET by Dembski, Behe, et al.
…and used by the researchers is called "adaptive walking", and bipod's point is that this methodology can likely be gleaned from the experiment.
As bipod and Joy didn't bother to read the paper, they didn't really have a point. It was just wishful thinking.
I'm interested in how Eric can claim "the *complete lack* of evidence" when it is clear that he hasn't bothered to examine the evidence in good faith. Doesn't a negative claim like that require reams of research?
Comment by Smokey — May 7, 2006 @ 12:31 pm
May 7th, 2006 at 12:39 pm
Smokey:
I don't see any misunderstanding in either of their posts.
I have no idea what a better example of intelligent design might be.
Now you're making things up? Where is this claim? The full citation is found within that commentary. The commentary was not inaccurate.
Which includes this: "the selected mutations were added to -humulene synthase"
You keep leaving out the relevant piece, they also concluded:"divergent evolution by rational design may be feasible on a significantly larger scale than currently possible"
Are you claiming to have done something here? Who are you? What paper are you referring to?
Comment by Guts — May 7, 2006 @ 12:39 pm
May 7th, 2006 at 12:41 pm
One more thing: the main reason that we have to go to such lengths to determine the function of a single one of this large family of enzymes is a phenomenon that makes other methods less reliable.
This phenomenon is consistent with MET, but not with intelligent design.
Comment by Smokey — May 7, 2006 @ 12:41 pm
May 7th, 2006 at 12:44 pm
Smokey, I don't think you really understood this paper yourself, ironically.
Comment by Guts — May 7, 2006 @ 12:44 pm
May 7th, 2006 at 1:02 pm
Duh - one thing the cited study shows is how very modest variations, subtle differences in how the substrate is bound, can lead to dramatically different chemistries.
I think yer making things up here, Guts. Do you know how these enzymes work? Specifically, do you know what active roles the enzyme plays in determining the steps subsequent to cation formation? How many of these come into play after initial substrate binding? Feel free to expound and provide pointers.
The bottom line remains the same - the entire universe of terpene chemistry in nature is eminently accessible to RM+NS.
Comment by Art — May 7, 2006 @ 1:02 pm
May 7th, 2006 at 1:11 pm
Art:
Duh, any protein with a function probably has plasticity residues. Enzymes especially, they need to form specific chemicals, catalyze the right chemical reaction, recognize substrates. This requires obtaining residues that exclude cross reactivity,accelerate any of those things, etc. If you have a group of proteins that do not have these residues, they might be just random peptides without any functions.
Art:
Naa, I'm not an anti-IDist.
Art:
The reason I gave contradicts your assertion of sloppy simplicity, I'm happy to play the reference game, you first.
Comment by Guts — May 7, 2006 @ 1:11 pm
May 7th, 2006 at 9:11 pm
Guts wrote:
I have no idea what a better example of intelligent design might be.
How about starting from scratch and actually designing a functional protein? Have I designed my car if I change the headlight bulbs from stock to blue?
Where is this claim? The full citation is found within that commentary.
Crikey. Bipod claimed that there was a LINK to the Nature paper. His/her claim was patently false. Citations are not links, just as intelligent modification is not intelligent design. Therefore, it is reasonable to hypothesize that bipod didn't bother to read the Nature paper.
…[the experimental design] includes this: "the selected mutations were added to -humulene synthase"
Which followed this: "The combinations of mutations were SELECTED so as to decrease the m value and to maintain Pm." This means that they were chosen by artificial selection, not designed. The design went into the selection of residues and the construction of a screen. The quote you mined just makes the authors look illiterate, as one doesn't add mutations to an enzyme, one makes mutations in a gene encoding an enzyme. The resulting enzyme is a mutant enzyme, not a mutation.
You keep leaving out the relevant piece, they also concluded:"divergent evolution by rational design may be feasible on a significantly larger scale than currently possible"
The authors' speculations aren't the relevant piece, the data are the relevant pieces. Moreover, the authors are pointing out that the only thing being rationally designed here is the evolution. Do you know what "saturation mutagenesis" means, Guts?
Are you claiming to have done something here?
Obviously, yes. What tipped you off?
Who are you?
I'm someone who works in this field of intelligent mutating of proteins to change their substrates and/or products; therefore, according to the analogy favored by the IDists here, I am analogous to God. Who are you?
What paper are you referring to?
I didn't refer to *a* paper. I referred to our work, which is published in multiple papers in Cell, JBC, Neuron, and PNAS.
Comment by Smokey — May 7, 2006 @ 9:11 pm
May 7th, 2006 at 9:57 pm
Smokey:
Smokey, you have a serious reading comprehension problem. Here is what Bipod said:
The link does indeed take you to the Nature article with the title "The intelligent design of evolution". (note the journal is a "joint endeavor" by EMBO and Nature Publishing Group).
I already mentioned this.
Once again, you have a serious reading comprehension problem. The entire point of the paper was this:"The systematic recombination approach described herein enabled us to design enzyme specificity rapidly and efficiently without a screen for the desired activity." They did use a screen on the previous step. But, they used site-directed saturation mutagenesis and site-directed mutagenesis to actually construct the enzyme specificity.
That isn't even coherent. I never claimed that this was not an example of intelligent design of evolution, thats the whole point this particular blog.
The data provide the basis for their conclusion that rational design of evolution is possible and can be done on a greater scale.
Do you know what site directed saturation mutagenesis and site directed mutagenesis is Smokey?
You mentioned a single experiment, just cite the one you did. I find it strange that this is the umpteenth post that you've mentioned this in and you still havn't cited it.
Comment by Guts — May 7, 2006 @ 9:57 pm
May 8th, 2006 at 9:44 am
As that all this result is analogous to, changing the color of a bulb?
This is rich. Intelligent modification is not intelligent design. So "the designer" can only *poof* things into existence. No modifications allowed.
While over at ARN they assert that RM+NS is a straw-man.
Comment by Mung — May 8, 2006 @ 9:44 am
May 8th, 2006 at 12:56 pm
Smokey is now hilariously claiming that the authors of the paper themselves are confused, and that this ":"The systematic recombination approach described herein enabled us to design enzyme specificity rapidly and efficiently without a screen for the desired activity." is some kind of vague abstract conclusion. He also continues to refuse to cite the experiment he supposedly did. This isn't a website for that kind of nonsense.
Bipod, I'm closing comments for now.
Comment by Guts — May 8, 2006 @ 12:56 pm