|
Author
|
Topic: Convergence or Divergence?
|
Jay
Member
Member # 268
|
posted 06. July 2002 12:42
Hi Frances,
I like the new angle this thread is taking in that it is going over specific examples, many of which you have kindly provided! However, I feel like we're not quite connecting on one big point.
Frances: "The best hypothesis seems to be at present homology based on the sequence motifs and the structural simularity. Plausible evolutionary pathways have been formulated and will be tested. If homology fails to be supported then we still have convergence or common design."
Really, all of these arguments for homology that I've seen so far just boil down to 'it looks kinda similar, so it must have been related'. This is rather ad hoc, and it makes the fatal mistake of already throwing out its competitors( convergence and common design) *simply based on the very shallow fact that the systems look 'too similar' (whatever that means) to not be due to homology*.
From the very start of this thread, the point was that almost all homology inferences operate in this very way. They compare two systems, and seem to kinda sit back and say, 'Gee whiz, those things really have some neat similarities, they must have been related!'
You say that plausible evolutionary pathways have been formulated and will be tested. What are those? What pathways for FtsZ/tubulin have been proposed, and how will they be tested? What has been proposed for any of these widely 'divergent' systems? I haven't seen anything! Are there any even remote estimates of the probability of divervgence of these systems? Is such a topic even considered?
If not, then how can we say that we have ruled out convergence/common design? Look at the way that they compare convergence and homology for FtsZ. Notice how superficial and non-rigorous it is! They are just looking for 'enough similarity' in their minds to justify calling it homologous. Talk about subjective!! They make comments like: ":It was speculated that an ancestral protein containing a Rossman fold-like GTP-binding domain diverged in two directions. In one direction it evolved into the typical GTPase family of proteins and in the other, into the atypical GTPase proteins FtsZ and tubulins (3). Alternatively, the similarity in structure between FtsZ and tubulin could be coincidental and there may not exist true homology between these proteins (21)."
Erickson, H.P., (1998). Trends in Cell Biology. Atomic structures of tubulin and FtsZ. 8: 133-137.
Can you imagine if I said that FtsZ appeared to be related by common design as it shared some key design ideas with tubulin, while being different in actual construction - but that maybe this is just coincidental and the two systems really are convergent by natural processes. Would such an interpretation ever be considered? Suppose that I had no rigorous back up to make this inference, but that it was just a guess because the systems just had a certain 'feel' for common design, in my opinion. Would you quote me in a debate if I infered things this way?
Frances: "So far the argument seems to be that the sequence alignment showed low similarity yet the two proteins showed strong similarity and shared motifs. Seems to be the best working hypothesis until new data becomes available."
In other words, there was very little sequence similarity, but there was still some kind of similarity, and so now we are justified in calling the two systems homologous? So is it the rule that *any* kind of similarity that we can dig up is automatically evidence for homology, and in many cases, enough to 'confirm' that inference as the correct one?
These FtsZ/tubulin examples nicely highlight this rather grave problem in the science today. Homology is a subjective inference made with very little data either for the model or against it. The only thing we seem to go by is whether there is 'enough' similarity of any sort to assume homology. Hopefully, we will see this jump in logic for what it is.
Thanks,
jay
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 06. July 2002 12:48
I wrote: "Yet many now proceed with the conclusion that ftsZ and tubulin are related." Frances replied by noting this is "Hypothesis Mike, hypothesis." Well, that's what it should be. But consider the sources Frances has supplied:
"Thus, prokaryotes are now known to possess homologues both of tubulin, namely FtsZ, and of actin." A hypothesis would read, "Prokaryotes may possess homologs of tubulin and actin."
And
"The bacterial cell division protein FtsZ is a homolog of tubulin." A hypothesis would read, "The bacterial cell division protein FtsZ might be a homolog of tubulin."
In other words, Frances provides two papers that conclude homology as something that is now known. Yet as Jay and I have been asking, exactly how was this established?
Again, I appreciate the links and abstracts provided by Frances, but none of them address this question in the context of our discussion. All of my arguments from my previous post still stand.
When I asked, "This is an interpretation of the structural similarity. Yet for our discussions, why should we conclude that structural similarity establishes a homologous relationship?"
Frances replied:
quote:
I would say because at the moment, combined with all the data it provides for the best hypothesis.
And it's not just structural similarity, which of course is an important factor but it also includes the finding that the tubulin signature GTP-binding sequence motif, GGGTG(T/S)G is found in FTSZ.
But why is this the best hypothesis? If it's a fuzzy judgment call that serves as a working hypothesis, fine. But the "best hypothesis?" If it is someone's subjective opinion that this is the "best hypothesis," fine. But the "best hypothesis" for us all?
Take the tubulin signature GTP-binding sequence motif. The odds of this sequence appearing by random chance are only 10^-9. This supporting evidence doesn't even tease apart common descent from coincidence.
When I noted that even when science excludes, a priori, intelligent design as a hypothesis, things remain very fuzzy. Frances replied, "controversy in science is quite common so to suggest that this is a surprise hardly is convincing evidence." I did not raise this as a surprise or evidence. The point is that this is a fuzzy, fuzzy topic. Thus, it looks like special pleading is involved when folks claim that ID can only sit down at the table when it brings its crystal clear demonstrations and evidence of the designers. It would be different if ID was brining a fuzzy hypothesis to an otherwise clear discussion.
So let me again stress something. When scientists argue, "We conclude that kinesin and myosin, and possibly G proteins, are probably directly related via divergent evolution from a common core nucleotide-binding motif, and describe the likely topology of this ancestor," we must remember that the conclusion was not reached by weighing it against teleological explanations. Nor was it reached through some robust calculations, excluding convergence. It is simply an argument from analogy. And this is highly ironic given that analogy is given a bad name by many critics of ID.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 07. July 2002 21:31
More on the FTSz Tubulin relationship
from An archaebacterial homologue of the essential eubacterial cell division protein FtsZ.]An archaebacterial homologue of the essential eubacterial cell division protein FtsZ.
It is argued in this paper that phylogenetic analysis reinforces the evolutionary linkage between FTSz and Tubulin.
Similarly in FtsZ Dynamics during the Division Cycle of Live Escherichia coli Cells
It is argued that it is likely that FtsZ evolved into the three types of tubulin proteins.
That Mike seems to forget that most anything in science is tentative. Also Mike's comment about 'analogy' or in this case 'homology' and ID ignores that homology is neither strict analogy nor is homology the only evidence.
But it's not analogy perse that makes ID problematic but the reliance on elimination combined with claims of 'no false positives'.
So I am still asking Mike how we would approach the evidence of 'design'? Does it include natural design? Do designers typically use analogy or homology in their design? If so, is the evidence of convergence found elsewhere evidence of intelligent design as well? How do we limit the explanatory power of design without understanding the designer or the causal history or the pathway?
As far as the GTP-binding sequence motif is concerned, it is found in quite a few proteins that bind ATP to GTP. It's not just the binding motif but also the structural simlarity and function.
In his Thesis Dirk-Jan Scheffers argues that
"The simultaneous elucidation of the
structures of FtsZ from the thermophilic
archaeon Methanococcus jannaschii (77) and
the αβ tubulin heterodimer from bovine brain
(108) provided conclusive evidence for FtsZ
and tubulin homology (figure 3)."
We can ofcourse argue about the relevance of these comments and how we can infer alternative pathways.
[ 07 July 2002, 22:02: Message edited by: Frances ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 08. July 2002 01:03
Back to the clamps, an interesting hypothesis was raised which places our discussion into context. It's not just tubulin/FTSZ or the clamp proteins which show homology but many others. Thus rather than looking at a single protein we should look at the whole picture.
Thje authors propose that the differences can be explained by DNA viruses. In fact this would make sense given the recent findings that horizontal or lateral gene transfer played a much larger role in early evolution.
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 08. July 2002 03:45
Hi Frances,
I read this guys paper that you cited awhile back. He raises some good points about the viruses that might be applicable to systems where it looks like there was more divergence than time would allow. However, in the case of these DNA clamps, we're playing a whole new ballgame that really isn't won or even helped by the viruses.
If you look at alignments of the PCNA (the archaeal/eukaryotic sliding ring type) it becomes rather clear that this is a very recognizable, conserved type. And when we compare it to the Beta Clamp (the bacterial analogue, which is also well conserved) it is evident that the two types are very different from each other, even though each is well conserved within its type. (For those interested, e-mail me and I'll send you a file with the alignments for each of these ring types, all ready to go!)
Within the PCNA ring type, we have variation from distant archaea all the way to humans, mycoplasmas, alveolates, fungi, plants, etc... In other words, we have probably well over a billion years of evolution here to check just how solid this PCNA conservation is. Likewise, for the Beta Clamps, we have nearly all phyla from eubacteria represented (we may have all of them in this one).
But again, when we compare the two types, we see that they are clearly not even close to each other. They are robustly very different in nearly all details.
All of this to say that when we look at the historical data, we have every reason to suspect that these have always been very different types, viruses or not. It should be noted again that there is almost no sequence similarity between the two types, and different subunit numbers. But on the other hand, their structures are identical, as well as their functions.
When I sit back and look at the alignments, and then look at the comparative crystal structures (which are found in that thread I linked to), I can't help but get the clear feeling that these are two independent design events, each separately created to accomplish a very similar task in two different sets of organisms.
There is no reason to think that a faster evolution (viruses) or even more divergence time would solve this problem. And to constrain the rings even further, we rule out the possibility of duplication and divergence (to speed up mutation) since their structures are so conserved. You can't speed up mutation like that and not wreck the structure.
Thanks,
jay
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 08. July 2002 11:16
Frances cites three more studies where the investigators claim homology between ftsZ and tubulin. One takes the structural similarity between the two to mean "provided conclusive evidence for FtsZ and tubulin homology," underscoring how many scientists have moved beyond the tentative, hypothesis stage with regard to these similarities.
Yet contrary to Frances helping his case by citing these studies, he is actually serving Jay's point. Again, how is this homology being established? As Jay says, "They compare two systems, and seem to kinda sit back and say, 'Gee whiz, those things really have some neat similarities, they must have been related!'"
The argument seems to be this: ftsZ and tubulin are very similar in a structural sense, have some functional similarities, and share a small sequence motif. Yet these observations do not help us tease apart the three possibilities raised by Jay and Frances.
It's understandable why scientists don't attempt to rule out a teleological explanation, as this is not done in science. But how was convergence ruled out? It wasn't and isn't.
In this case, homology is inferred through an analogical argument. This is why all the evidence Frances cites is in the form of similarities between the two proteins. Since the homology inference is an analogical argument, perhaps we should focus on some of the differences?
Frances asks me a set of questions:
So I am still asking Mike how we would approach the evidence of 'design'?
I laid out four angles of approach on my web page. But keep in mind all the things Jay and I have already written. Scientists conclude homology without rigorously ruling out the two other possible explanations. There is no reason my approach must therefore be different. What we have are fuzzy data that can be interpreted differently. So one chooses a particular interpretation, for whatever fuzzy reasons, and runs with it.
Does it include natural design?
How would this differ some convergence?
Do designers typically use analogy or homology in their design?
Good question. This is the type of thing ID researchers want to start thinking about.
If so, is the evidence of convergence found elsewhere evidence of intelligent design as well?
Another good question. This could be. For example, it is possible that the many examples of convergent evolution reflect front-loaded evolution. That is, a designer working with the knowledge of physical law and emergent properties may carefully choose initial conditions to increase the likelihood of intended outcomes.
How do we limit the explanatory power of design without understanding the designer or the causal history or the pathway?
For starters, engineers usually design things for a reason. Thus, similarities between ftsZ and tubulin, for example, that cannot be explained in terms of function are unlikely to have arisen from design.
[ 08 July 2002, 11:18: Message edited by: Mike Gene ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 08. July 2002 13:09
I asked how do limit the explanatory power of intelligent design to which Mike suggested that "engineers usually design things for a reason". Does this mean that we use only human engineers as a template for design? And how do we understand their reasons? Is there always a reason? Usually? Do we claim to understand these reasons?
Sometimes engineers may design something reusing earlier designs but sometimes engineers like to be creative and do it their own way. Some even like to place their 'signature' in their designs.
Mike also suggest front loading but in world in which chaos is far more common, front loading seems to be utterly impossible. Of course if we assume front loading then for all practical purposes ID is not different from natural law. That is evolution ran its natural course.
Mike also states that quote:
Scientists conclude homology without rigorously ruling out the two other possible explanations.
You seem to believe that science is eliminative in fact the approach that is used is to propose a hypothesis and test it as thoroughly as possible. Of course IDers are free to propose their own hypothesis and show why it is a better one.
Jay suggests that the evidence points to different origins and that the structures look as if they were designed. My question is how do we separate natural origins from intelligent design here? Why do we need intelligent design to explain? What does it explain?
[ 08 July 2002, 13:12: Message edited by: Frances ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 08. July 2002 13:40
Frances asks several questions concerning the use of human engineers as a template for inquiry. No one claimed that this approach would be easy or straightforward. Nevertheless, it can be used to provide constraints on one's inquiry. In fact, most ID critics seem to realize this. How? A common argument against ID is the "bad design" argument. I'm sure we have all heard, "No engineer would design the backward wired retina of the vertebrate eye." I happen to think there is something to this argument. But this is again taking us off the main topic.
Frances claims that front-loading seems impossible, given all the chaos. It probably depends on the type of front-loading and I have addressed these claims in the posting I linked to above. Suffice it to say that I don't think it impossible to front-load unicellular life forms to make them more likely to evolve into multicellular life forms. Is this just "natural law" at work? FLE causes us to focus on the initial conditions and not simply treat them as brute givens. Is it true that any ol' unicellular life form is just as likely to evolve into a mulitcellular life form as any other possible unicellular life form? I doubt it. But again, this is taking us off the main topic.
Frances says science is not eliminative, but instead proposes a hypothesis and tests it. Well, that's the very approach I took on my web page article.
As for Frances last set of questions, again I refer to my essay. This time I'll quote.
quote:
Thirdly, we might expect these differences to be very important, explaining why a designer would employ the different variations on the GPD theme. And one of the facts not mention thus far in this thread is that although both ftsZ and tubulin have very different amino acid sequences when compared to each other, the sequences of both ftsZ and tubulin are highly conserved in bacteria and eukarya, respectively. In other words, when we compare ftsZ sequence within bacteria and tubulin sequence with eukarya, we find strong sequence conservation. FtsZ, for example, shows 40-50% identity when very different forms of bacteria are compared and I believe the tubulin conservation is even higher. In fact, one paper on my desk states "tubulins are among the most conserved proteins known."
This pattern is consistent with independent origins by design. That is, the first bacteria were endowed with a GPD variant known as ftsZ that has been conserved for billions of years due to its important design objective. Similarly, the first eukaryotes were endowed with a GPD variant known as tubulin that has been conserved for billions of years due to its important design objectives.
On the other hand, if we try to force common descent on the two distinct, highly conserved proteins, we face a strange situation. For prior to the evolution of ftsZ and tubulin from this hypothetical ftsZ/tubulin-like precursor, there was no apparent functional constraint. If there was, it is difficult to explain how the two sequences so radically drifted from each other only to be locked into place (of all places) in the last common ancestors of eukaryotes and bacteria. But wait a minute. The 3-D structure was being conserved. That's the basis for inferring the common descent. Yet what was it doing prior to the two sequences getting locked into place? Nothing bacterial. Nothing eukaryotic.
To put it simply, we have billions of years of sequence conservation within groups, and sequence differences between groups so dramatic that they could be related by chance. To force them into a common ancestral state has all the appearance of a ad hoc move that ignores this pattern. And why do we force them into the homology category. Because they are too similar, in other regards, for convergence (although this is a intuitive argument). In other words, they are too different for homology, and too similar for contingency. Yet an extensively experienced process of rational protein design (first specifying global folding and then decorating for fine-tuning) might very well produce this pattern.
The ID approach is simply hindered by our own inexperience in designing proteins.
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 08. July 2002 15:48
Hi Frances,
Frances: "Jay suggests that the evidence points to different origins and that the structures look as if they were designed. My question is how do we separate natural origins from intelligent design here? Why do we need intelligent design to explain? What does it explain?"
I like Mike's asessment that some things look too similar to be contingent, yet too different to be homologous. That is precisely what these DNA sliding clamps suggest to me. They are two very old, well defined groups that strongly argue against having ever been related, if you look at the alignments.
However, as I pointed out, they are still so similar that some in the field are sort of surrendering to calling them homologous because they don't seem to know what else to call them. In other words, when scientists come across this paradox of too similar yet too different, they tend to fall towards homology. And from the start of this thread, I've pointed out that this reason for falling on this side of the fence has almost no empirical support in most of these cases.
What I think of as potential hotspots for common design are exactly these sorts of systems like Mike's and mine where scientists seem to warily cave into "homology peer pressure" if you will. They know that they are too similar to not be related *in any sort of way*. And, they won't have to answer too many hard questions if they just go ahead and call them homologous... and as we all know, any thought of common design (sharing of common ideas) is already thrown out, so there really is no choice but to give in and go with the flow. And so homology it is! This in spite of the fact that they do not know any of the corresponding probabilties needed to weigh this asessment. IDsts would be shot for such tactics! ![[Smile]](smile.gif)
Frances: "Do designers typically use analogy or homology in their design? "
I'd say it's likely that both are used. However, again, the question at hand is whether there are instances that we can detect, rather than asking what all possible instances there might be. That is why we focus on analogous systems. These appear to be much clearer examples of possible common design than would, say, two proteins that share 99% sequence identity.
So we may have millions of examples of common design for all the we know, but for now, we're only able to detect some of the clearer ones. But the value in this is in extablishing that there are, in fact, real examples of common design. This then adds that factor into the equation when homology asessments are made.
And once again, this is not without value. This 3rd party to the possibilities may turn out to better explain certain similar systems and may clean up our alignments used in homology modeling, for instance. I, for one, would love to see something like this as it would be of practical benefit to me.
Thanks,
jay
[ 08 July 2002, 15:54: Message edited by: Jay ]
IP: Logged
|
|
Leonid Andreev
Member
Member # 282
|
posted 08. July 2002 18:45
Reading the above exchange, I can’t help but draw a parallel with a hypothetical discussion of whether one can speak of the evolution of automobile bolts and nuts while knowing nothing about automobiles, automobile construction history and evolution. I believe the answer is: yes, one could, but better should not. And by the way, the automobile design is much less complex than that of the living matter. There is certainly ID in automobile bolts and nuts, which is determined not only by a design engineer, but also by a market, fashion, technology level, and many other factors.
Advances in exact sciences have led to a diehard belief in their omnipotence, as well as in the possibility of scientification of any knowledge. Researchers in 16sRNA homology since long ago do not consider biological properties of organisms to be a primary factor, obviously thinking that if something does not look right to a biologist, that’s biology’s problem. For decades, numerical methods in chemotaxonomy, which, too, deal with homology, are being developed based on a naïve, childish idea: the more characters of biological objects you use, the more accurate biological (biochemical) description you produce. Mixing the unmixable by comparing the ability for arabinose fermentation with flagellum shapes, peptide hydrolysis with respiration inhibition, and thousands of other properties displayed by organisms in an alive state, numerical taxonomists discuss thus obtained senseless dendrograms with as much seriousness as a discovery of an unknown plant genus would deserve. For those who think that I must be unjustly exaggerating the funny side of chemotaxonomic approaches, I would recommend to see the materials on some of the commercially available software packages for clustering analysis, for example, Clustan (www.clustan.com). One of the sample case studies used in Clustan deals with an evolutionary tree of 25 animals based on five variables of respective mother milk: ash, lactose, fat, protein, and water contents. Although the tree shows an orangutan to be at the same node as a mule, and a whale together with a deer, that does not seem to bother anyone. A good evolutionist can always find an explanation.
Coming back to proteins. Among the basic primary substances of alive biomass (lipids, carbohydrates, nucleic acids), proteins are perhaps the only ones that are capable for demonstration of their functional activities outside a cell. There is an important ecological meaning to that, because lytic enzymes of various kinds that are released upon cells death are further involved in processes that make the residues, including various macromolecules, ready for consumption by other organisms. Therefore, an exchange of “innovations” and their improvements (as well as convergence and divergence) is immanent in the living matter of the Nature.
In addition to the extracellular factor, an important thing about proteins is that even though they are functionally autonomous, they are capable for changing their activities and even for re-directing certain processes under the influence of membrane environment and numerous co-factors and co-enzymes. In physiologically similar organisms, effector profiles may be quite similar (development towards convergence). In other organisms, e.g. those developing new physiological-ecological niches, same proteins are striving for divergence in pursuit of better adjustment to new conditions.
Therefore, an abstract discussion of convergence/divergence issue isolated from a carrier, i.e. a biological object, may be only as much scientific as a kabuki theater performance. May a Rolls-Royce screw look exactly the same as a 1950 Mercedes-Benz screw? Yes, it may…, no, it may not…
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 09. July 2002 01:10
Mike suggests that since a common argument against ID is imperfection of design that this might help us explain the limitations of ID. But that assumes that we are able to understand the limitations of the designer? Why should we assume that the designer was more or perhaps less able than we are? In fact the common argument is more typically used in context of an Intelligent Designer (in caps).
Why should the front loading be limited to the origin of life? Perhaps the front loading took place at the Big Bang? That would also hide the designer behind the Planck time. Surely that would solve a lot of problems. Of course we may never know if this front loading was real or merely an anthropic principle.
Mike argues that the data could be due to chance, after all two very different sequences have 'survived' in two different lines. Too different for homology and too similar for contingency. But can we thus infer design ala Dembski? Is the homology to similar or too dissimilar for design? How will we know if we know nothing about the designer? And is it too similar for contingency? Why? You yourself showed that the probability was well outside the limit set by Dembski. When asked if a designer were to use homology or analogy, Mike circumvents the problems by suggesting that this makes for interesting thought. But how are we going to resolve this without knowing more about the designer?
Mike also suggests that similarities which cannot be explained in terms of function are unlikely to have arisen by design? Why is this the case? Are we limiting the designer to be mechanical? No embelishments? No creativity?
Perhaps it's time to look at the full picture and not just at one protein? As Villarreal et al argue several 100 genes are homologous between bacteria and eukaryotes. They provide with a plausible hypothesis. Jay rejects their hypothesis based on the conservation of the bacterial analogue but what does that have to do with the argument that it was a (retro) virus which caused significant similarity between replicative DNA polymerase of eukaryotes and certain DNA viruses. They perform interesting phylogenetic analysis of these data. In fact similar findings had been found earlier.
Baldauf and Doolittle in "The root of the universal tree and the origin of eukaryotes based
on elongation factor phylogeny, 1998" show how their phylogeny show a weak link of the crenarchaeotes as a sister group of the eukaryotes.
Jay is interested in the outcome of applying design in the hope that it may lead to something. Of course I would be the last one to prevent Jay from going down this path. After all Newton found much solace in the thought that it was an intelligent designer who was keeping the planets in line and yet he came up with some very insightful theories and laws.
Perhaps applying Dembski's design inference to the proteins might be helpful? Any takers?
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 09. July 2002 01:48
Hi Leonid,
Interesting thoughts. I have to agree with you on the 'more data points is better' way of asessing homology. This is especially true when we move back to phyla and domains. There really is no universal phylogeny at all. You have, on the one hand, those that analyze some of the informational genes assuming that archaea are more similar to eukaryotes than bacteria, but then when you move over to metabolism, you get eubacteria and archaea as sister groups. In fact, with some genes such as HSP70, you get even weirder results like the robust paraphyly of archaea clustering in between some of the Gram + eubacteria!
So, trying to take an overall average phylogeny doesn't work here at all. There is no overall theme of phylogeny in the domains that we are aware of, but rather, many robustly conflicting phylogenies. Is it not possible that perhaps we are asking the wrong question when we ask how the three domains were related by common descent? I see this as a nice hotspot for possible common design of organisms, rather than common design of proteins as we've been discussing (how cool if the two line up!!). But I digress....
Leonid: "Therefore, an abstract discussion of convergence/divergence issue isolated from a carrier, i.e. a biological object, may be only as much scientific as a kabuki theater performance."
But... the ring clamps! The cytoskeleton stuff!! Is that not what you're looking for? We're trying, really!
Thanks,
jay
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 09. July 2002 01:58
Hi Francis,
Francis: "Perhaps it's time to look at the full picture and not just at one protein? As Villarreal et al argue several 100 genes are homologous between bacteria and eukaryotes. They provide with a plausible hypothesis."
Ah, but here we go again! So what if they share 100 genes? Does that really mean that they are related by descent? Is this not just an extension to the organismal level of the 'gee, they look really similar and so must be homologous' argument?
In fact, there are other reasons to think that these domains may not really be homolgous, but rather due to common design. Try this one if you got the patience (it's long!):
Brown, JR, Doolittle, WF. Archaea and the prokaryote-to-eukaryote transition. Microbiol Mol Biol Rev. 1997 Dec;61(4):456-502. Review.
They basically show that the story between the domains is not simple. In fact, it really flies in the face of much of what homology would expect. The three domains are full-out chimeras of each other in the genes that they do share, and have no consistent phylogeny at all, but robustly different ones. In addition, they share many instances of analogous systems like those we've shared.
Finally, they have novel IC systems introduced into some of them that are not shared in the other domains. Taken together, I believe that these data points support the idea that the domains are related by common design and construction. The reasons for this I went over more in detail a few replies back. But the bottom line is that the relationship between the domains from an evolutionary perspective is strained at best, and leaves wide open a common design inference, just as many of these individual protein systems do.
Thanks,
jay
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 09. July 2002 02:18
Jay,
YOu are right, horizontal gene transfer is surely going to complicate the picture but from a contingency perspective the overlap in 100's of genes is surely going to eliminate chance and strengthen the homology argument. Of course we can't say much about design since we have not much to determine how a designer might go about it.
Of course common design will always remain an 'option' but until we can really formulate intelligent design/common design, it will remain just that.
Also ICness perse cannot be used to infer intelligent design either.
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 09. July 2002 11:38
Frances asks no less than 18 new questions (although he has not answered many of the fundamental questions Jay and I have asked). Most of his questions stem from the designer-centric position, which assumes that independent knowledge of the designer is necessary to carry out a design inference. Not so. We can simply follow the example of science, which constructs what are known as "working hypotheses." For example, in science, a common ancestor is inferred through analogical reasoning. Although we have no independent knowledge of this "common ancestor," we assume things about it, using analogy, and then use these assumptions to generate working hypotheses that are then tested.
In my case, I generate a working hypothesis that the designer possessed a human-like intelligence (otherwise, we have no hope of detecting its design, since it is experience with our design that serves as the template for inquiry). I further constrain the hypothesis to life being "bioengineered," for as Janitor constantly points out, we already possess a "science of design," thus it makes sense to use this as a guide (although I am hindered by not being an engineer*). Further yet, I hypothesize that the designers has access to a superior base of knowledge and advanced technology. This assumption is based on the fact that our current abilities cannot come close to designing novel life forms, let alone those sufficient to the task of seeding a distant planet.
These are working hypotheses that do provide methodological constraints. Ironically, this should please Frances. For the constraints I employ favor the false-negative. If life was designed by an intelligence completely unlike human intelligence, or by an artist, or by someone who plays practical jokes, etc., I will miss this design.
Of course, one can always question the assumption, just as is done in science. The point is not in coming up with a deductive argument that cannot be questioned. The point is not to deliver certain knowledge. The point is in coming up with a teleological approach to explore the biotic world, in the quest of generating insight and fruit. After all, that I could ask a dozen unanswerable questions about some common ancestor doesn't cause scientists to abandon the use of common ancestry.
As for front-loading, Frances is asking philosophical questions. Again, his questions are all answered in the other thread I linked to. Suffice it to say that I infer design behind the OOL for independent reasons. It is then the realization that the design of life entails some element of front-loading. The research question is then to explore the nature and extent of the front-loading.
Frances then asks a set of questions dealing with homology vs. contingency. How do we know the two proteins are too similar for contingency? We don't. We can only look to scientific consensus, which has rejected it and sided with homology. However, since ID proponents have a third option, they can the likewise question homology (since scientists reject contingency, they have no other option, thus do not seriously question homology, as Jay has been explaining). In the end, I claim to have established nothing. Only another hunch that gets fitted into the working hypotheses, which are weaving themselves into a grander story. In science, things are supposed to be tentative. ID need do no better.
When Frances asked if the designer uses analogy or homology, I did not "circumvent this problem." I expressed an honest "I don't know." These are questions to be explored, not answered with a priori reasoning. Perhaps Frances can clarify by providing an example-in-question. I was about to speculate, but then I realized I'm not quite sure what Frances means here. Is he using 'homology' in the traditional sense?
As for looking at the full picture, and not just one protein, I agree. After all, I do not envision tubulin and ftsZ to have been designed and deposited in an otherwise non-designed organisms. My working hypothesis is that the first life forms were designed (a heterogeneous pool). Unless one can explain, from an engineering perspective, why every individual life form would have a different genetic code, different translation mechanisms, and completely different ways of extracting energy, the housekeeping similarities don't clearly point in one direction or another. A hundred shared genes is just a variant of the tubulin/ftsZ problems amplified one-hundred-fold.
[ 09 July 2002, 11:42: Message edited by: Mike Gene ]
IP: Logged
|
|
|
Printer-friendly view of this topic