|
Author
|
Topic: The Other Flagellum II
|
Mike Gene
Member
Member # 149
|
posted 20. September 2003 09:02
Nic,
I’m not playing any word games. I raised the issue of a cooption frequency as it relates to DE vs. FLE. You dismissed such an effort as ‘ridiculous’ and suggested we should just stick with the notion that cooption is “ubiquitous.” But as we can see, the word can carry different meanings. I will once again simply draw attention to the fact that the concept of FLE, and not DE, has apparently spawned the desire to measure something about biotic reality. Look closely enough and we see that FLE and DE are not indistinguishable.
Before going there, however, let’s pull in another mechanism to unpack front-loaded designs – gene duplication. When it comes to the cooption of gene products, would you say that most of these events followed from an earlier gene duplication?
As for cooption being “non-Darwinian,” I never claimed this. I pointed out that there are many evolutionary events that do not involve cooption, but no one would think of them as being “non-Darwinian.” Therefore, cooption is not an essential feature of Darwinian evolution. Of course Darwinian evolution does not mean only improvement of function. It only means variation and selection. And that can occur without cooption. In fact, it would seem it occurs more often without cooption than it does with cooption. Thus, the insufficiency of ‘ubiquitous’ and the need for a measurement.
[ 20. September 2003, 09:03: Message edited by: Mike Gene ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. September 2003 09:25
I happen to agree with Mike that cooption could, at least in principle, be measured from a purely frequentist perspective.
By that I mean at the coarsest level of description of processes, and with some carefully defined (and possibly fuzzy) objective measure of when "cooption" occurs, we could determine its frequency. But then only in principle. There are real difficulties determining the "actual" cooption rate, because the knowledge of events of biological descent are sketchy in many instances--over the larger picture. This is due to the biological record in fossil and other evidence being quite incomplete.
The evidence being incomplete does not rule out the theory of common descent. Nor should it rule out bounding the rate of cooption.
Now once this was found, it would need to be used very carefully. For example the background rate of cooption would say very little about any individual case. The reason is that cooption probability will certainly be very different depending on circumstance.
With that consideration, and suitable definitions of "ubiquitous", surely cooption is ubiquitous!
Mike on page 2:
quote:
For example, how often does cooption occur? Pointing to examples and proclaiming it “common” doesn’t help.
Now pointing to examples does indeed help in one regard. When we use the normal so-called "theory" of "design inference", that of the Explanatory Filter, we find that we cannot "eliminate chance" when we discover that a known "chance" mechanism readily provides a pathway for events. Likewise for a Bayesean comparative method. So any claim needs to be backed up by some other evidence, if it is to be credible, even by ID enthusiasts standards.
I asked for a definition of how “IC” was being used (back on page 1). And I notice that IC is no longer being used as a particular claim of evidence. That is of course because the definition used was not one that necessarily precluded a natural non-intelligent action explanation, e.g. of “cooption”.
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 20. September 2003 10:33
I originally wrote, “Recall that you predicted that “all major structurual components of IFT complexes, that are well-conserved between cilia IFT complexes, will have identifiable homologs in noncilial tubulin-based transport systems.” Well, as I discuss in the OP, as it appears this cooption prediction has failed.”
Nic replied: quote:
This is based on your ruling toprim domains out of the picture from the start, based on some very dubious logic that independent convergence on a toprim domain is more likely than common descent. Your non-homology explanation still doesn't explain sequence similarity between said domains, and sequence similarity is what the BLAST searches hit on.
First, let’s return to the context of your prediction (found on page 6 of the original Flagellum thread).
I asked, “If, for example, it turns out that many of the components of the assembly machinery don’t have homologs in intracellular transport, will you concede that the flagellum didn’t evolve?
You replied:
quote:
If none of them did, it sure would cause me to doubt things. 'Course, we already know that some homologies are readily identifiable in the cilium, especially for the most important components, but it's still a prediction of the evolutionary model.
But hey, I'll stick my neck out: all major structurual components of IFT complexes, that are well-conserved between cilia IFT complexes, will have identifiable homologs in noncilial tubulin-based transport systems.
You claimed that all major components would have identifiable homologs. Thus far, they don’t. This would take us to the meaning of predictions and the difference between a soft prediction and a hard prediction. But that’s another topic.
As for ruling out TPR domains (with one of the IFT proteins), I do not consider this “dubious logic.” We now know that instead of there being a gazillion protein domains out there, there are only a few thousand. And TPR domains do not appear to be high information domains. Thus, I fail to see why they are out of the reach of convergence, especially when the TPR motif carries out the generic function of protein-protein interaction.
What’s more, you can’t come to the table where teleological and non-teleological causes are in dispute and demand that only non-teleological mechanisms be considered. From the teleological perspective, there is no reason to think a designer would constrain the use of TPR domains only to IFT88. Why would any designer do this, especially given the generic utility of such a domain. In a way, it’s like arguing that a designer would only use screws to piece together a VCR and nothing else.
As for picking up similarities with BLAST, what’s being picked up is the TPR motif. For example, there is a hypothetical protein from Nostoc (NP_484932) that is picked up with a BLAST search using IFT88 sequence with a E value of 10^-6. The Nostoc protein has 600 amino acids and IFT88 has over 800. Yet what is being matched is a short span of ca. 114 amino acids, where 36 are identical, including positions 448-561 from IFT88, a region that contains TPR motifs.
Finally, you wrote: Furthermore, I think that the "noncilial tubulin-based transport systems" have barely begun to be characterized. (See bottom)
But that is not that meaningful in the age of genomics. For example, if you search the entire genome of Arabidopsis or baker’s yeast, you don’t find ‘homologs’ anywhere. Such organisms have lost their flagella. But why would they also both just happen to lose the supposed flagellar precursors involved in “noncilial tubulin-based transport systems”?
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 20. September 2003 10:46
Gedanken,
I should clarify that I am not thinking of a cooption frequency as it applies to all of evolution. That is, I am not suggesting The Cooption Frequency of Evolution. Instead, I am thinking of a cooption frequency respective to any particular gene. This is one way that we can effectively rule out FLE. But first we need to address the question of the relationship between cooption and gene duplication. There seems to be plenty of anecdotal evidence that indicates gene duplication must occur prior to a cooption event. Or, even if the relationship is not necessary, evolution typically unfolds in a “duplication, then cooption manner.” Does anyone agree?
You write, “When we use the normal so-called "theory" of "design inference", that of the Explanatory Filter, we find that we cannot "eliminate chance" when we discover that a known "chance" mechanism readily provides a pathway for events.”
But that would depend on the genes involved and their respective cooption frequencies. What would be scientifically sloppy is to invoke cooption for a flagellar gene product by pointing to ear bones.
[ 20. September 2003, 10:47: Message edited by: Mike Gene ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 20. September 2003 11:42
I'm developing some interest in this thread, since cooption is one of the topics we are looking at studying in the context of Avida. Already in that context the question of level of analysis has become a can of worms. For example, there are (at least!) three levels at which cooption could occur in Avida: single instructions, substrings of instructions within (sub)programs performing one or another function, and the (sub)programs themselves. Recall that Avida employs a population composed of assembly language programs that evolve to perform input-output mappings corresponding to various logic functions, some relatively simple, some more complicated. Within those larger programs there are (sub)programs that perform each function, say one that performs AND. Within those (sub)programs several subfunctions must be performed - I/O, register contents comparisons, and so on. Those (sub)programs apparently often evolve to share code - some instructions (and perhaps short strings of instructions) are used by several (sub)programs in service of performing several logic functions. (I am reminded of frameshifts that enable multiple encodings in a string of base pairs.)
Cooption is evidently necessary for the Avida digital organisms to evolve to perform the more complicated functions. The Lenski, et al., Nature paper control conditions showed that for a population to evolve to perform the most complicated mapping - EQU - simpler functions must be available to be performed, though no particular simpler function (nor pair of simpler functions) is necessary. It is therefore not clear how performing those simpler functions allows more complicated functions to evolve. Lenski, et al., did not do the detailed analysis of lineages that would speak to that question. I have not done so either, but some less-than-complete examination of a couple of lineages suggests that performing the more complex functions does not involve coopting the code that performs a simpler function whole. Rather, it appears that substrings of previously evolved (sub)programs are coopted - a part of the code that evolved to perform a simpler function may be recruited by evolution to contribute to the performance of a more complicated function. Sometimes the part is a single instruction, sometimes a string of several instructions. The fact that no particular simpler function nor pair of simpler functions is necessary to evolve lineages that perform EQU lends some credence to the notion that cooption involves substrings of (sub)programs rather than whole (sub)programs.
Mike Gene's remark that he is thinking of "cooption frequency respective to any particular gene" implies one level of analysis - the gene - but I suspect that cooption occurs at multiple levels in biology as it appears to do in Avida. That obviously has implications for finding potential cooptable precursors of biological structures and processes.
RBH
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. September 2003 17:10
[Anyone reading this, go back and read RBH's just previous post for context in a different field.]
Given Mike's restriction of subject of "cooption frequency", then I am not not sure that one can reasonably determine a specific "cooption frequencey" or even determine a meaning for that statement.
Does it mean take the actual frequency of actual biological history (not the known history but actual sequence of all descent occurrences) and find the ratios? That would clearly be impossible due to missing evidence.
Does it mean take some simulation and produce a theoretical model for that frequency? That might be more meaningful.
But then Mike's comments don't quite fit in. The issue I raise with respect to EF is the issue of whether there is a high probability that one has missed a "probability distribution" in calculating the EF result. My thread on "Does intelligence imply 'motive'" is relavant. (Mike has made notes there, and may wish to follow that when I get more information now in development posted.) But the main issue is that ubiquitous cooption hardly points to "sweeping the field clean" of possibilities.
Mike must actually defend that. To do so is not only show an incredibly low "cooption frequency", but to defend that he has a low probability of having missed some "distribution."
AH--RBH's point is exactly to the issue of the "missed distribution" in the EF! (That apparently cannot be reduced to the "cooption frequency" on the specific gene. This is precisely the type of occurrence not previously thought about, that was an example used in my thread mentioned above.)
[ 20. September 2003, 20:33: Message edited by: gedanken ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 20. September 2003 18:08
Ged writes:
quote:
Given Mike's restriction of subject of "cooption frequency", then I am not not sure that one can reasonably determine a specific "cooption frequencey" or even determine a meaning for that statement.
As I noted, co-option can be identified at a specific evolutionary point if it can be shown that the ancestral state did not use the gene the way the descendant does. This is most evident with deuterostomes. I would not say that co-option is ubiquitious, more so because we don't have the required cladistic information to say such a thing. Although work with co-option frequencies can determine the extent to which co-option occurs.
[ 20. September 2003, 18:24: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. September 2003 19:05
Alonso points out information that helps show why cooption cannot be so easily characterized.
His answer to measuring cooption frequency is in terms of noting the particular cooption on the specific gene of a specific case. But then the cooption frequency is either 1 or 0. Either that very specific cooption occurred, or it did not.
However if we broaden the cooption frequency to the issue of all possibilities to create a particular structure, then we have very uncertain knowledge possible of possible ways to create a structure in that class.
Aother view is not to limit to a particular "target", just as a particular dice throw pattern of a long sequence can have low frequency as an assembly--yet high probability of the next throw being something. Once again you have to look at likelihoods of something coopting IN to the given one as some sort of Bayesean and/or Markov view, the probability of the target by itself is useless. (In fact such a "probability" does not exist, only relative probability of transitions suitably defined and constrained, if they can be described at all. This will get us to the possibility/probability that the analysis missed an important "distributiion", as mentioned before, thus invalidating any significance factor of an eliminative "design inference".)
I don't see how Alonso's point helps us to determine a cooption probability at a relevant level.
Also on "ubiquitous" cooption, I am taking Yersina's point and definition--meaning that cooption is very common in the overall observation of biological evidence. I won't get into word games on that.
[ 20. September 2003, 19:14: Message edited by: gedanken ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 20. September 2003 19:41
Ged writes:
quote:
Alonso points out information that helps show why cooption cannot be so easily characterized.
His answer to measuring cooption frequency is in terms of noting the particular cooption on the specific gene of a specific case. But then the cooption frequency is either 1 or 0. Either that very specific cooption occurred, or it did not.
This is of course, characterizing the co-option event, which can be easily done with examples such as deuterstomes. The brachyury gene is involved in very different tasks in these clades. Although they have different body plans, they seem monophyletic. This gene is expressed in posterior mesoderm and hindgut. However, each specific clade seems to have co-opted for different functions, which can be characterized. For example, sea urchins express this gene at the bastoporal ends of the gut. Of course, this is probably not the area that Mike is looking at (devo) however, it is an example that illustrates how a co-option frequency, for, say the brachyury gene, can be measured.
As to the supposed ubiquitous nature of co-option, notice that the example with deuterstomes is not seen in other organisms (such as CRY in mice). Or we simply do not have the cladistic information. So I would not say that co-option is ubiquitous.
This has absolutely nothing to do with "all possible ways to build a structure".
[ 20. September 2003, 19:43: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 20. September 2003 20:14
I still think that Alonso wants to have it both ways.
He wants to specify the possibilities in a general way.
Then he wants to talk about the frequencies in terms of a very specific type of existing expressions that are known of a particular present day structure in comparable organisms.
By doing so, you have a very high probability of a "missed distribution" with respect to any intended inputs to the Explanatory Filter. To acutally find such "gene frequencies" in the specific case of interest (in stead of actual frequencies in some other present day case) you would have to do either simulated or real genetic engineering trials on all the genes that could relate. (And not miss any--that would be another "missed distribution".)
The difference between observed cooption frequency in one case, and possibilities in another case is quite significant. The possibilities in one case cannot be measured in terms of the actuals in another, because the possibilities include more than has actually been tested even in the present case--if that were in any way a valid analody which I suspect is not.
Then once again one has to put in the Bayesean or Markovian framework--beyond all the difficulties above--to get a "fan in" probability of any relevance. This aspect has not even been addressed at all.
If you look at "frequencies" of expression in terms of strictly those multiple expressions in today's organisms, then you are not (as Alonso admits) looking at the case in question. You are in fact just doing a spot check as part of generalized overall cooption frequencies. This is because it is not directly comparable to the case in question, it is only a spot check of some type of similar case, thus applicable only to general cases in some manner. This is of course NOT what Mike is talking about, as he has said, as Mike wants to focus on a specific case. So I still don't see relevance.
[Sorry about changes after your post Alonso, but if you post in first minute after my post you might mis an edit. I don't want to continue a back-and-forth, rules and all... OK, I am through editing, Alonso, you can change next post as appropriate.]
[ 20. September 2003, 20:27: Message edited by: gedanken ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 20. September 2003 20:20
Ged writes:
quote:
He wants to specify the possibilities in a general way.
I am not sure where you got this from in my post. I am not talking about possibilities. Simply whether a co-option event occured and the frequency of it's occurance for a particular gene.
Ged writes:
quote:
If you look at "frequencies" of expression in terms of strictly those multiple expressions in today's organisms, then you are not (as Alonso admits) looking at the case in question.
No, I did not admit any such thing. I think that the analysis I did is looking at the case in question.
Ged writes:
quote:
You are in fact just doing a spot check as part of generalized overall cooption frequencies. This is of course NOT what Mike is talking about, as he has said. So I still don't see relevance.
Actually, you misunderstood Mike here and Mike posted his clarification:
quote:
I should clarify that I am not thinking of a cooption frequency as it applies to all of evolution. That is, I am not suggesting The Cooption Frequency of Evolution. Instead, I am thinking of a cooption frequency respective to any particular gene.
Thus my post was extremely relevant. I was thinking of a co-option frequency wrt the brachyury gene.
[Added in Edit in response to Ged's edit:
Ged writes:
quote:
The difference between observed cooption frequency in one case, and possibilities in another case is quite significant. The possibilities in one case cannot be measured in terms of the actuals in another, because the possibilities include more than has actually been tested even in the present case--if that were in any way a valid analody which I suspect is not.
There is no such thing as an observed co-option unless someone siting there is actually watching the event. What we can do is look at the evidence. What I did for deuterstomes was not actually observe the co-option event, but infer it due to the fact that the brachyury gene performs a different function from that of the ancestor organism. Each clade uses it for a different task (pointing to the toolbox concept of genes). In this way, one can measure and characterize co-option frequencies for particular genes.
done editing]
[ 20. September 2003, 20:46: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 20. September 2003 20:47
Gedanken's excellent posting raises once again the issue surrounding establishing such metrics as frequencies, frequencies are established not by just looking at known expressions. I wonder how one can justify a statement like "So I would not say that co-option is ubiquitous." based on the limited dataset so far. I have to agree so far with Gedanken statement "I don't see how Alonso's point helps us to determine a cooption probability at a relevant level."
Whether or not co-option is ubiquitous cannot be established from looking at single occurrences thus the idea that it does not happen everywhere is not a helpful argument against.
There are some interesting papers that may help us address these issues in a scientific manner
Gene co-option in physiological and morphological evolution. True JR, Carroll SB.
Annu Rev Cell Dev Biol. 2002;18:53-80. Epub 2002 Apr 02.
quote:
Co-option occurs when natural selection finds new uses for existing traits, including genes, organs, and other body structures. Genes can be co-opted to generate developmental and physiological novelties by changing their patterns of regulation, by changing the functions of the proteins they encode, or both. This often involves gene duplication followed by specialization of the resulting paralogous genes into particular functions. A major role for gene co-option in the evolution of development has long been assumed, and many recent comparative developmental and genomic studies have lent support to this idea. Although there is relatively less known about the molecular basis of co-option events involving developmental pathways, much can be drawn from well-studied examples of the co-option of structural proteins. Here, we summarize several case studies of both structural gene and developmental genetic circuit co-option and discuss how co-option may underlie major episodes of adaptive change in multicellular organisms. We also examine the phenomenon of intraspecific variability in gene expression patterns, which we propose to be one form of material for the co-option process. We integrate this information with recent models of gene family evolution to provide a framework for understanding the origin of co-optive evolution and the mechanisms by which natural selection promotes evolutionary novelty by inventing new uses for the genetic toolkit.
Yersinia has collected an impressive list of co-option examples in the literature
It may also be helpful to compare co-option with Dembski's portrayal of such
quote:
To sum up, the Darwinian mechanism requires a selectable function if that mechanism is going to work at all. Moreover, functional pieces pulled together from various systems via coevolution and co-option are selectable by the Darwinian mechanism. But what is selectable here is the individual functions of the individual pieces and not the function of the yet-to-be-produced system. The Darwinian mechanism selects for preexisting function. It does not select for future function. Once that function is realized, the Darwinian mechanism can select for it as well. But making the transition from existing function to novel function is the hard part. How does one get from functional pieces that are selectable in terms of their individual functions to a system that consists of those pieces and exhibits a novel function? The Darwinian mechanism is no help here. Darwin himself conceded this point. Writing in the Origin, he noted: "Unless profitable variations do occur, natural selection can do nothing." To say that those profitable variations are random errors is to beg precisely the point in question.
Darwin on co-option"
quote:
In considering transitions of organs, it is so important to bear in mind the probability of conversion from one function to another, that I will give one more instance. Pedunculated cirripedes have two minute folds of skin, called by me the ovigerous frena, which serve, through the means of a sticky secretion, to retain the eggs until they are hatched within the sack. These cirripedes have no branchiae, the whole surface of the body and sack, including the small frena, serving for respiration. The Balanidae or sessile cirripedes, on the other hand, have no ovigerous frena, the eggs lying loose at the bottom of the sack, in the well-enclosed shell; but they have large folded branchiae. Now I think no one will dispute that the ovigerous frena in the one family are strictly homologous with the branchiae of the other family; indeed, they graduate into each other. Therefore I do not doubt that little folds of skin, which originally served as ovigerous frena, but which, likewise, very slightly aided the act of respiration, have been gradually converted by natural selection into branchiae, simply through an increase in their size and the obliteration of their adhesive glands. If all pedunculated cirripedes had become extinct, and they have already suffered far more extinction than have sessile cirripedes, who would ever have imagined that the branchiae in this latter family had originally existed as organs for preventing the ova from being washed out of the sack?
Dembksi's suggestion that NS does not select for future function but that seems irrelevant for the issue of co-option, novel functions arising through co-option have been described in the literature. If the argument is that the examples are not of sufficient detail (such as Dembski's quote of Harold might suggest) may I point out that this is of no help to the design inference approach.
One may be tempted to claim that absence of detailed scenarios are helpful for the design inference but such a position ignores the eliminative nature of the DI. If the DI cannot establish the likelihood of such scenarios then it cannot eliminate them and thus has to enter into a 'holding pattern'. Without much more detail it is not possible to determine if these 'just so stories' are without any merrit or if they are viable ideas leading to actual hypotheses. From my reading of the literature quite a few scenarios of co-option have risen above 'just so stories'.
Co-opting RM&NS and co-option into front loading is also not going to be helpful for the DI since again regularity and chance cannot be eliminated, and in fact are argued to be the relevant forces.
If as Mike Gene seems to argue that IC systems may arise fully naturally through some vague story of 'front loading' then we have to accept that IC systems can arise naturally. In addition the confusion of co-option with 'pure chance' seems to ignore the issues at hand. In fact one may very well argue that a (final cause) teleological perspective in such instances may lead one into the 'wrong direction'.
Some people may argue that ID in fact may be helpful in exploring these issues in a scientific manner, Mike himself has written many ideas related to ID/IC. But we should be careful to not confuse the prolific writings of Mike with contributions of ID to scientific understanding. While motivations of design and assumptions of teleology may have been helpful for Mike to explore the scientific data, little of his findings seem to have direct relevance to ID.
In fact I would say that we see an excellent example of co-option of a viable theory (RM&NS, co-option) without much of an explanation as to their relevance to ID that go beyond a level of 'just so stories'. Nevertheless I do encourage Mike to pursue these issue, many scientists have found teleological thinking useful in generating hypotheses and ideas.
[ 20. September 2003, 22:05: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 20. September 2003 20:50
Pim writes:
quote:
frequencies are established not by just looking at known expressions.
Which is irrelevant. What Ged was confusing was two concepts, a general co-option frequency for all evolution, and a specific co-option frequency for a particular gene.
Pim writes:
quote:
I wonder how one can justify a statement like "So I would not say that co-option is ubiquitous." based on the limited dataset so far.
One of the reasons I said that was precisely because of the limited dataset (another is that it simply doesn't happen everywhere).
[EDIT: Pim edited out the attribution of a Mike Gene quote to Gedankin, so I deleted the correction.]
Pim writes:
quote:
If the argument is that the examples are not of sufficient detail (such as Dembski's quote of Harold) may I point out that this is of no help to the design inference approach.
Of course it's a help to an extrinsic design inference. If the only argument for the co-option in the evolution of a specific system is a vague just-so story (or a wrong story) then that helps the design inference. However, with the FLE in hand, one can start putting meat on some co-option stories. As Mike has noted:
quote:
But the problems with co-option are deeper. Once we leave the random tweaking of a protein along a linear axis guided by selection and instead appeal the multiple coincidences entailed by different, independent proteins being shaped for various other functions that just happen to coalesce into a brand new system, the role of coincidence itself is brought into question. As I have explained elsewhere, design might also be reflected as a front-loaded state that is likely to find various anticipated solutions. In such cases, IC may serve as the springboard by which one detects design through a front-loaded original state. In other words, IC may have evolved through co-option and RM&NS, but that is not the whole story. The whole story may entail whether the original state was stacked such that a random search was likely to stumble upon a new IC state through shuffling and co-opting what was originally designed. In other words, we may have a situation where evolution was designed to spawn certain IC systems when the conditions were right
I'm wondering if you guys are going to keep editing your post after I reply. I feel like I'm dealing with the mighty morphin power rangers.
[ 20. September 2003, 21:47: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 20. September 2003 21:57
Gedanken:Given Mike's restriction of subject of "cooption frequency", then I am not not sure that one can reasonably determine a specific "cooption frequencey" or even determine a meaning for that statement.
Why not? Nic himself wrote, “You can take almost anything and sooner or later, if you trace it back far enough, you will end up at a change-in-function "event".” If you can do this, and there is a “sooner or later,” there’s something to measure, right?
Does it mean take the actual frequency of actual biological history (not the known history but actual sequence of all descent occurrences) and find the ratios? That would clearly be impossible due to missing evidence.
Obviously, any cooption frequency would not be able to detect undetectable cooption events (but then one has to wonder what is the difference between an undetectable cooption event and one that did not happen). All this means is that our cooption frequency would be an estimate, which is far better than anything science currently has. In genetics, recombination frequencies between two loci don’t detect double crossovers between the two loci. That doesn’t render the recombination frequency useless.
What would be so impossible about taking any gene and attempting to estimate how often it has been coopted across evolutionary time? In the age of genomics, this would not seem impossible to do.
Mike must actually defend that. To do so is not only show an incredibly low "cooption frequency", but to defend that he has a low probability of having missed some "distribution."
Mike doesn’t have to defend this because Mike is not out to prove design. In an investigation where I am trying to determine if FLE or standard DE was involved, I don’t invoke cooption regardless of the evidence. If there is no evidence that cooption was at work, I don’t invoke it. Attempts to further estimate the cooption frequency may then add to the picture. Of course, I would also have to make the caveat that I have not determined that we have a low probability of missing some “distribution.” Investigations rarely proceed with crystal clear insight.
Again, it is noteworthy that even though Darwin introduced the concept of cooption over a century ago, Darwinian theory has ever since been uninterested in something like a “cooption frequency.” As I noted before, the laundry list of cooption events in the literature amount to nothing more than anecdotes. Arguing that “cooption is ubiquitous” is little more than philosophical hand-waving.
Anyway, I keep coming back to my question. For most instances of cooption, is gene duplication a prerequisite? The abstract Pim cites throws support my way:
“This often involves gene duplication followed by specialization of the resulting paralogous genes into particular functions.”
[BTW, I have to mention (as an aside) that I have found the design perspective is indeed a gold mine for ideas. Over a year ago, I wrote five essays on the bacterial flagellum. I have at least four more to go, but then I also got interested in cytosine deamination. That provided three essays and I have another two to go there. Then, it’s the eukaryotic flagellum with two in the bank and at least two more to go. Now, the whole idea of cooption frequencies is giving me all sorts of ideas (not to mention a flood of front-loading ideas). And some people actually believe ID is a science killer. ![[Wink]](wink.gif) ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 20. September 2003 22:05
RBH: Cooption is evidently necessary for the Avida digital organisms to evolve to perform the more complicated functions.
Very interesting. If such a finding extrapolates to biotic reality, this would strengthen the teleological interpretation of FLE, as a designer attempting to front-load multicellular life through unicellular life would need to rely on cooption. That is, cooption, in of itself, ceases to become evidence for a non-teleological origin.
IP: Logged
|
|
|
Printer-friendly view of this topic