Register | Sign In


Understanding through Discussion


EvC Forum active members: 65 (9162 total)
4 online now:
Newest Member: popoi
Post Volume: Total: 915,806 Year: 3,063/9,624 Month: 908/1,588 Week: 91/223 Day: 2/17 Hour: 0/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   An ID hypothesis: Front-loaded Evolution
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 44 of 216 (653289)
02-19-2012 6:04 PM
Reply to: Message 12 by PaulK
02-19-2012 2:16 PM


Re: The obvious problem with front-loading
quote:
Unless preserved by selection the genome can be changed by mutation at a relatively rapid rate. Any "front-loaded" information that isn't actually in use is unlikely to be preserved unchanged for long enough to become useful.
And this is why, under the FLE hypothesis, mRNA transcripts are the sequences that channel evolution, not unexpressed sequences.

This message is a reply to:
 Message 12 by PaulK, posted 02-19-2012 2:16 PM PaulK has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 47 of 216 (653293)
02-19-2012 6:53 PM
Reply to: Message 37 by Dr Jack
02-19-2012 5:12 PM


quote:
Well, I can flat out tell you the answer to that one: we do. The problem for you is that this isn't, in any way, incompatible with conventional theory. Exaptation is common, and well understood. These genes perform important roles in single celled organisms, and are later exapted to perform different roles in multi-cellular organisms. You need to come up with something that will distinguish your front-loaded genes from simple exaptation.
Further, I'd suggest that your hypothesis should lead to us seeing many, probably most, such multi-cellular genes in prokaryotes - we don't.
I'm not sure if it's just me, but those two paragraphs seem slightly contradictory. On the one hand, you state that "Well, I can flat out tell you the answer to that one: we do [find such homologs in prokaryotes]" and on the other hand you say that "I'd suggest that your hypothesis should lead to us seeing many, probably most, such multi-cellular genes in prokaryotes - we don't." Just saying.
A couple of points to be made here. Morphologically, hearts and brains have no homologous counterparts in unicellular organisms or even some "lower" multicellular organisms. Why then do you think that conventional theory would predict that the genes for the development and function of these organs would have homologs in unicellular organisms?
Secondly, you state that "I'd suggest that your hypothesis should lead to us seeing many, probably most, such multi-cellular genes in prokaryotes - we don't. There are some, yes, but most are not found in bacteria; a few more in Archaea and a decent portion in single-celled Eukarya or in multi-cellular but simply differentiated Eukarya. It seems to me that if front-loading was true we would expect a majority of such genes to be present in all species, would you agree?"
For clarification, I would agree that, under the FLE model, we would expect homologs of these genes in many prokaryote taxa - either in the form of structural or sequence similarity. We have sequenced only a fraction of the biosphere. Thus, this prediction can be potentially confirmed. Based on FLE logic, I predict that we will find many, many prokaryotic homologs of key genes involved in the development and function of metazoa. Does conventional theory predict this? You don't seem to think so.
Consider, for example, Pax-6 and Hoxa-3. Pax-6 plays a key role in the development of the eye, and Hoxa-3 plays a major role in heart development. Based purely on the logic of the conventional theory, would you expect these genes to have homologs in prokaryotes?
Of course it does. Highly conserved proteins have higher sequence identities than average proteins whilst, conversely, proteins not under selective pressure rapidly diverge. Proteins shared between the three domains must be highly conserved and thus will tend to have higher sequence identities.
I don't see how this would be the case. If a protein has homologs in all domains of life, this does not mean that it is necessary for life per se; it means that it is beneficial. And most proteins are, of course, beneficial. This doesn't constrain their sequence identity above average.

This message is a reply to:
 Message 37 by Dr Jack, posted 02-19-2012 5:12 PM Dr Jack has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 53 of 216 (653320)
02-20-2012 12:08 AM
Reply to: Message 43 by bluegenes
02-19-2012 6:03 PM


Re: The best of error minimizing codes?
I hope you'll agree that I was right to suggest a thread specifically for your front loading hypothesis. You may not like all the reactions, but you're certainly getting some!
Yes, I agree that it was a good idea. I don't mind objective criticism at all, but for the record, for all of you: now that the weekend is over, I won't have that much time to respond to all of you - only one or two responses per day, unless I can muster some more time. This means that I will be only responding to the highest-quality criticisms (e.g., MrJack and bluegenes and bluejay).
That said:
There's an interesting thing about this. While our standard code has certainly undergone selection for error minimization, and seems to be close to its local peak on the fitness landscape, there are plenty of higher fitness peaks elsewhere which nature could have hit if it had happened across a different random code to start with.
Looking at this from the I.D. perspective, it seems like bad news. Our standard code seems a very unlikely choice for your rationally designing frontloaders if they were aiming at error minimization.
Here's a 2007 paper that might interest you.
Paper on the rugged fitness landscape of genetic codes...
I'll take a look at that paper, but there are some things that need emphasizing. That the standard genetic code is not at a global optimum for error minimization really isn't bad news from an ID perspective. This is because there are other functions aside from error minimization that would be optimized, and this is indeed the case in the genetic code, as highlighted by Bollenbach et al., 2007:
As we learn more about the functions of the genetic code, it becomes ever clearer that the degeneracy in the genetic code is not exploited in such a way as to optimize one function, but rather to optimize a combination of several different functions simultaneously. Looking deeper into the structure of the code, we wonder what other remarkable properties it may bear. While our understanding of the genetic code has increased substantially over the last decades, it seems that exciting discoveries are waiting to be made.
Full optimization of one function may significantly reduce the optimization of another. Thus, a balance would have to be made between various functions.
The genetic code is, nevertheless, at a local optimum for error minimization. And the absence of a phylogenetic tree like I describe in my essay, and the fact that this highly optimized genetic code is nearly universal, points to front-loading.
As for your main point about it, I see no reason why any very early more error prone versions should have survived alongside a prokaryote LUCA, and I don't really see the point of your analogy with non-flagellar functional homologies, as these are nothing to do with sub-optimal flagella.
The prokaryote LUCA could easily have had a thoroughly sub-optimal genetic code. This would, in turn, evolve and be fine-tuned, but many detours and by-ways would be explored, with some less optimal genetic codes branching off, producing a phylogenetic tree of genetic codes. There is no reason why this should not have occurred, under the non-telic model.
I'll respond to your other points later.
Edited by Genomicus, : No reason given.

This message is a reply to:
 Message 43 by bluegenes, posted 02-19-2012 6:03 PM bluegenes has replied

Replies to this message:
 Message 54 by Blue Jay, posted 02-20-2012 1:04 AM Genomicus has seen this message but not replied
 Message 57 by bluegenes, posted 02-20-2012 9:13 AM Genomicus has replied
 Message 62 by Dr Adequate, posted 02-20-2012 4:16 PM Genomicus has seen this message but not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 55 of 216 (653325)
02-20-2012 3:35 AM
Reply to: Message 43 by bluegenes
02-19-2012 6:03 PM


Re: The best of error minimizing codes?
...and I don't really see the point of your analogy with non-flagellar functional homologies, as these are nothing to do with sub-optimal flagella.
The point about the flagellum is that it is predicted by Darwinian evolution that we should find functional pre-cursors, if it did indeed evolve. The same logic holds for the genetic code: we should find sub-optimal pre-cursors. Perhaps a better analogy would be the eye: the human eye is more optimal than, say, the rhinoceros eye, for example. One could build a tree with less optimal eyes in basal lineages, with more optimal eyes in late-branching taxa. That you cannot do the same with the genetic code is interesting, to say the least.
Is the suggestion that the front loaders might be able to predict something like the chance endosymbiotic event that seems to have enabled the evolution of eukaryotes?
Well, for starters, that's assuming that the endosymbiotic event that gave rise to the eukaryotes wasn't planned. The question "how could they front-load that?" is a valid one, but keep in mind that the human race has very little experience in the field of front-loading biological states. I think the answer to this question could be solved if we really thought about it. My personal opinion, of course.
And wouldn't they have to have a very clear idea of the future orbit and physical evolution of the planet itself, not to mention the behaviour of the local star which could radically effect things?
Well, the way I see it is that these front-loaders would have seeded many planets with these life forms. On some planets, these life forms may have gone extinct. Also, convergent evolution at the molecular level seems to indicate that it wouldn't be that terribly difficult to front-load future biological states - the behavior of our local star notwithstanding.

This message is a reply to:
 Message 43 by bluegenes, posted 02-19-2012 6:03 PM bluegenes has replied

Replies to this message:
 Message 59 by bluegenes, posted 02-20-2012 10:42 AM Genomicus has seen this message but not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 56 of 216 (653329)
02-20-2012 5:49 AM
Reply to: Message 49 by Blue Jay
02-19-2012 9:30 PM


Hi, Genomicus.
Welcome to EvC!
Thanks!
Since front-loading predicts biases in the trajectory of evolution, shouldn't we be able to detect these biases by examining patterns in evolutionary trajectories across multiple lineages? For example, if the front-loaders intended to promote the emergence of a certain characteristic, wouldn't we predict that multiple lineages of descendant organisms would evolve that characteristic? Isn't this what "biased evolutionary trajectories" means?
In the first place, I'm not envisioning anything like extreme front-loading, where something as specific as the human species is front-loaded. So if I'm understanding you correctly, you're saying that if multicellularity was front-loaded, we should expect multicellular life forms to evolve independently. But this is what is believed to have occurred; that is, that multicellular life forms arose multiple times but many of these lineages went extinct - see here: Puzzling Out The Tree Of Life Of Green Plants.
"Multicellular life was another great idea that almost certainly arose more than once. A one-celled organism has to do everything -- it's a jack of all trades and master of none -- but in a group of cells, individuals can specialize and be very good at one specific thing," Mishler said. "But again, only a very few major lineages made it through to the present, and those all from one basic stock."
And there is good evidence of biased trajectories of important biological features: to name just one example, eyes have evolved independently in different lineages - indicative of a biased trajectory.
Essentially, it would not be reasonable to argue that "characteristics that only emerged once in Earth's evolutionary history probably were not front-loaded into the original," because lineages can be lost through deep time.

This message is a reply to:
 Message 49 by Blue Jay, posted 02-19-2012 9:30 PM Blue Jay has replied

Replies to this message:
 Message 60 by Blue Jay, posted 02-20-2012 3:08 PM Genomicus has replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


(1)
Message 64 of 216 (653417)
02-20-2012 10:24 PM
Reply to: Message 60 by Blue Jay
02-20-2012 3:08 PM


Well, I wasn't trying to put forward an example yet: I was trying to establish a conceptual principle whereby we could test the predictions of FLE.
Well, I think that you are basically correct that we should lineages evolving multiple times if front-loading has occurred: this does not, however, mean that we should each lineage exactly as the other. We might find analogs of those lineages. I'd expect such multiple evolution to be more noticeable at the molecular/gene level than at the morphological level - since the driving force behind front-loading is at the genomic level.
Still, with some biological features, like blood, it would evolve multiple times only under specific circumstances.
Do note, too, that the more specific the front-loading objective, the less likely it is that that objective will evolve multiple times.
So, based only the single criterion I proposed, I would regard multicellularity as potentially consistent with the front-loaded evolution hypothesis.
IMHO, it's not "potentially consistent" with front-loaded evolution, but rather it's a clue in favor of the idea that multicellularity was front-loaded.
My concern with the specific example here (eyes) is that the bias in trajectory only emerges in one "later branch" of the Tree of Life (Metazoa).
Well, you wouldn't expect eyes in prokaryotes, now would you?
So, it seems that the capacity to develop eyes isn't rooted at the base of the Tree of Life, but at the base of the animal branch of the Tree of Life.
Actually, the capacity for eye development is encoded at the root of the phylogenetic tree of life. Pax6 is a gene involved in eye development, for example. When you BLAST (blastp; default search parameters) the protein encoded by Pax6 (accession number: P63015) against the domain Prokaryota, you get significant hits (E-values < 1e-05). A PSI-BLAST search would almost certainly uncover hits with even greater significance. This suggests that eyes (and other major organs in Metazoa) were anticipated by the first genomes.
Unless FLE allows for front-loading within individual branches of the tree, and not exclusively at the base of the tree, I think the proper conclusion is that eyes probably were not front-loaded.
The base of the tree contains the necessary information to shape future evolution, and it would contain the genes "anticipating" the evolution of Metazoa; for example, genes (or homologs) needed for eye function would be encoded in the first genomes such that when Metazoa do appear on the scene, eyes can develop.
I will accept your "deep-time" argument if you withdraw your objections to the similar argument in relation to sub-optimal genetic codes.
My "deep time argument" was more of a musing. Lineages can be lost through deep time, but this would be an ad hoc rationalization. Still, even though, for example, animals have not evolved multiple times independently AFAIK, you have to consider this in a broader context. While this might be a clue counting against front-loading, you would also have to consider the evidence favoring the idea that animals were front-loaded.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.

This message is a reply to:
 Message 60 by Blue Jay, posted 02-20-2012 3:08 PM Blue Jay has replied

Replies to this message:
 Message 65 by New Cat's Eye, posted 02-21-2012 11:09 AM Genomicus has not replied
 Message 70 by Trixie, posted 02-21-2012 3:25 PM Genomicus has replied
 Message 72 by Taq, posted 02-21-2012 3:51 PM Genomicus has replied
 Message 81 by Blue Jay, posted 02-22-2012 12:09 AM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 73 of 216 (653486)
02-21-2012 5:32 PM
Reply to: Message 72 by Taq
02-21-2012 3:51 PM


Mr Jack said:
I'd suggest that your hypothesis should lead to us seeing many, probably most, such multi-cellular genes in prokaryotes - we don't. There are some, yes, but most are not found in bacteria; a few more in Archaea and a decent portion in single-celled Eukarya or in multi-cellular but simply differentiated Eukarya. It seems to me that if front-loading was true we would expect a majority of such genes to be present in all species, would you agree?
Taq said:
Yes, that is exactly what we would expect to see if evolution is true. There is nothing front loaded about it. Evolution is descent with modification. Of course the earliest ancestors are going to have the genes that were later modified through random mutation and natural selection. Where else would you expect to find them?
Am I the only one that is puzzled by this discrepancy of statements?

This message is a reply to:
 Message 72 by Taq, posted 02-21-2012 3:51 PM Taq has replied

Replies to this message:
 Message 76 by Taq, posted 02-21-2012 5:47 PM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 74 of 216 (653487)
02-21-2012 5:33 PM
Reply to: Message 70 by Trixie
02-21-2012 3:25 PM


You might want to do your PSI-BLAST before making claims about what it will find. The results are pretty much what you get with a BLASTP.
Possibly, but PSI-BLAST uses a different scoring system than blastp. Just a thought.

This message is a reply to:
 Message 70 by Trixie, posted 02-21-2012 3:25 PM Trixie has replied

Replies to this message:
 Message 80 by Trixie, posted 02-21-2012 6:36 PM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 75 of 216 (653489)
02-21-2012 5:38 PM


Into the swamps...
As Bluejay predicted, I am effectively getting swamped with responses. So is there anyone who would especially like me to respond to their points? I do owe bluegenes a response, though, so responding to his post is priority.

Replies to this message:
 Message 77 by Dr Adequate, posted 02-21-2012 5:50 PM Genomicus has not replied
 Message 78 by Taq, posted 02-21-2012 5:52 PM Genomicus has not replied
 Message 79 by bluegenes, posted 02-21-2012 6:09 PM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 82 of 216 (653522)
02-22-2012 5:03 AM
Reply to: Message 57 by bluegenes
02-20-2012 9:13 AM


Re: The best of error minimizing codes?
Yes, that's the answer I was expecting, but it was optimization for error minimization that you mention in the O.P. In fact, the paper suggests that the standard code is frozen some way off even its local fitness peak so far as error correction is concerned. However, the other improved error minimization peaks are so numerous that it would seem unlikely that there aren't those that would achieve better all round balanced function (or more rational design) than the one we've got. The designers had a lot of choice.
I have read the paper you cited earlier. Their conclusions are in direct contradiction with the conclusions of Freeland et al.
Novozhilov et al. conclude that:
"The standard code appears to be the result of partial optimization of a random code for robustness to errors of translation. The reason the code is not fully optimized could be the trade-off between the beneficial effect of increasing robustness to translation errors and the deleterious effect of codon series reassignment that becomes increasingly severe with growing complexity of the evolving system."
While Freeland et al. state that:
"Here, we show that if theoretically possible code structures are limited to reflect plausible biological constraints, and amino acid similarity is quantified using empirical data of substitution frequencies, the canonical code is at or very close to a global optimum for error minimization across plausible parameter space."
Why is there this discrepancy? IMHO, the answer is that these two different studies used different methodologies. Novozhilov et al. primarily use the Gilis scoring matrix to arrive at their results, while Freeland et al. use the PAM74-100 substitution matrix. The Gilis scoring matrix is based on protein stability, while the PAM matrix is based on observed amino acid substitution frequencies. Given that the PAM matrix more accurately portrays amino acid similarity from an evolutionary perspective (that is, the PAM matrix shows amino acid similarities based on what amino acid substitutions have been accepted by natural selection over evolutionary time), while the Gilis scoring matrix portrays amino acid similarities on the basis of protein stability, IMHO Freeland et al.'s conclusions are more biologically realistic from the perspective of evolution. Also, on the side, I don't think Novozhilov et al. consider biosynthetic restrictions on the genetic code, while Freeland et al. do - again, making their results more realistic.
Furthermore, the results of a 2009 paper fly right in the face of the conclusions of Novozhilov et al. A study by Butler et al., 2009 (Extreme genetic code optimality from a molecular dynamics calculation of amino acid polar requirement), used Monte Carlo simulation to assess the optimality of the genetic code, concluding with:
"The extreme optimization of the genetic code therefore strongly supports the idea that the genetic code evolved from a communal state of life prior to the last universal common ancestor." (Emphasis added)
Thus, I must disagree with the argument that the genetic code is only "partially optimized." It is extremely optimized for error minimization, and as Freeland et al., 2000, point out, it is very close to the global optimum across plausible parameter space.
It fits the scenario of an initial random functional code evolving to become optimized, and hitting a local peak on the fitness landscape fairly quickly then getting frozen. The paper suggests that random codes hit their local optimums easily and quickly.
A couple of points to be made here:
1) An extremely optimized genetic code, like the standard genetic code, wouldn't seem to be terribly advantageous to unicellular organisms, in contrast to less optimized genetic codes. Radical substitutions would be far more likely to be non-deleterious in unicellular organisms than in complex, multi-cellular organisms. In fact, a less optimized code might be more advantageous for unicellular organisms, in one sense: it would accelerate the rate of protein evolution.
As I explained earlier:
Also note that there might actually be an advantage for prokaryotes to have a sub-optimal genetic code: radical mutations would be more frequent, and this could possibly accelerate evolution. You wouldn't have to go through multiple amino acid substitutions to get to a radically different amino acid. I.e., if you take a look at the PAM1 substitution matrix, there is a 0% probability that an alanine --> tryptophan substitution will occur. So, in order to get alanine to change into tryptophan, it'd have to be like this, for example: alanine --> arginine --> tryptophan. The canonical genetic code, then, is a good system for complex organisms, where radical substitutions will be very likely deleterious. But for bacteria, a sub-optimal genetic code, where radical mutations are far more frequent, this might just be a selective advantage.
2) Secondly, our different perspectives on the level of optimization come into play here. I view the genetic code as extremely optimal, in agreement with Freeland et al. and Butler et al. The evolution of genetic code optimality, AFAIK, is thought to be Gaussian, instead of linear. This means that there really would be plenty of room for alternative, less optimal codes to branch off, creating a phylogenetic tree of genetic codes, with sub-optimal codes in basal lineages. It doesn't seem all that likely that a highly optimized genetic code will quickly evolve from a random code. You'd also have to gradually incorporate all 20 amino acids. Yet there is also no phylogenetic tree of genetic codes, with gradually increasing numbers of amino acids.
There's quite a gulf between your view and that of the side-loaders.
True.
Edited by Genomicus, : No reason given.

This message is a reply to:
 Message 57 by bluegenes, posted 02-20-2012 9:13 AM bluegenes has replied

Replies to this message:
 Message 83 by Wounded King, posted 02-22-2012 11:00 AM Genomicus has replied
 Message 85 by Dr Adequate, posted 02-22-2012 11:21 AM Genomicus has replied
 Message 87 by Taq, posted 02-22-2012 11:48 AM Genomicus has not replied
 Message 178 by bluegenes, posted 02-27-2012 5:46 AM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 107 of 216 (653723)
02-23-2012 10:27 PM


General Response to Objections
General Response to Objections
Instead of getting bogged down with replying to every one of you, here I will address the objections to the FLE hypothesis that I think are surfacing the most frequently. The objections I will address here are (a) front-loading specific objectives (e.g.,eyes or blood) is not very feasible, (b) the branching problem raised by Dr Adequate, (c) the issue of predictions made exclusively by the front-loading hypothesis, that is, predictions made by the FLE model but not made by conventional evolutionary theory.
Keeping perspective in mind
My objective here is not in the slightest to win the debate or anything like that. I freely admit that there are problems with the FLE hypothesis; the human species has little experience in anything like directed panspermia or front-loading, so I imagine that these problems will start disappearing once we really start getting serious about panspermia and directing evolution with unicellular organisms. My real objective here is to refine the front-loading hypothesis so that when the proper time comes, it can be submitted to the scientific community for review.
Front-loading Specific Objectives
The gist of this objection can best be summarized by this statement by Bluejay:
The more specific the objective, the less likely the objective is to be met at all using a crude technique like front-loading. I find it highly dubious to suggest that something as specific as an eye could feasibly be front-loaded at all for this exact reason.
I’d like to point out that the argument I can’t imagine how the eye could have been front-loaded, therefore it could not have been isn’t really all that valid.
It’s also important to understand something here: front-loading does not necessarily entail directly front-loading something like the eye. Front-loading is stacking the deck (to borrow from Mike Gene) such that the evolution of certain features becomes much more plausible. For example, if we design rhodopsin in a bacterial genome, we could predict that when eyes evolve, rhodopsin will be used. This would make the evolution of the eye much more plausible, since rhodopsin would not have to evolve from a non-rhodopsin-like precursor. Similarly, if we loaded the first genomes with globins, then we could predict that when animals do arrive on the scene, their blood will be built around hemoglobin. Designing the initial genomes with hemoglobins (or hemoglobin homologs) encourages the evolution of blood.
Another example will be cited here. If we designed cilia into simple eukaryotes, this could help channel the evolution of more complex eukaryotes like animals. This is because the cilia would encourage the evolution of organisms with tissues like that seen in complex animals. This is an example of course, and this needs to be emphasized before anyone starts saying that I think the first eukaryotes were directly designed instead of front-loaded.
Thus, front-loading is largely about stacking the deck, encouraging the evolution of certain biological features through initial designed states and making the evolution of those features more plausible.
Furthermore, convergent evolution has effectively demonstrated that the front-loading of specific biochemical systems is indeed feasible. For instance, hemoglobin has evolved independently in lampreys and all other vertebrates. It thus seems plausible that specific proteins can be front-loaded from initial states. A more dramatic example comes from the evolution of C4 photosynthesis. This biochemical pathway is thought to have evolved independently more than 45 times (see this review by Sage, 2003: The evolution of C4 photosynthesis). And C4 photosynthesis is mediated by four different enzymes (see figure; taken from MetaCyc): pyruvate orthophosphate dikinase, NADP malic enzyme, phosphoenolpyruvate carboxylase (PEPC), and malate dehydrogenase.
Figure. A diagram of the C4 photosynthesis pathway, taken from MetaCyc.
If I were to propose that a specific biochemical pathway like C4 photosynthesis was front-loaded, one might be tempted to argue that this is extremely infeasible. In short, while front-loading a biochemical pathway isn’t like front-loading an organ such as the eye, one could still argue that I find it highly dubious to suggest that something as specific as a C4 photosynthesis pathway could feasibly be front-loaded at all. Yet C4 photosynthesis has evolved independently many, many times, proving the feasibility of front-loading a specific biochemical pathway. Why does this prove the feasibility of front-loading? Precisely because it proves that fairly specific targets can be reached by the blind watchmaker independently and multiple times— this means that FLE is feasible — given certain initial states.
I like the example of C4 photosynthesis because I think it’s a good illustration of how initial states can shape subsequent evolution, such that the blind watchmaker stumbles upon a certain target. For the record, I do not think that C4 photosynthesis was front-loaded. But it’s a good example of how feasible FLE could be.
Sage, 2003, describes the initial states that must be present before C4 photosynthesis can evolve (see figure; from Sage, 2003).

Figure. In this figure, Sage describes the initial states that are required before C4 photosynthesis can evolve. Thus, if we designed a plant genome such that it is prone to many gene duplication events (and we would have to design the right genes into this genome), and designed the plant such that it has closed veins and its bundle sheath organelles are enhanced, and we placed this plant on another planet (supposing this planet was not hostile to plant life, of course!), then we could predict that when we visit this planet many centuries later, C4 photosynthesis will have evolved in this plant species. Botanists et al., please don’t quibble about my lack of a discussion on how we would need many plants with this designed genome, etc.

Thus, it seems to me that specific proteins can be front-loaded — as evidenced by the independent evolution of hemoglobin; specific biochemical pathways can be plausibly front-loaded, as demonstrated by the multiple, independent evolution of C4 photosynthesis; and furthermore, specific molecular machines could be front-loaded: this is evidenced by the independent evolution of two protein export systems from the bacterial flagellum — namely, the type III secretion system and a protein export system in Buchnera very similar to the TTSS, and one that is also derived from the bacterial flagellum — but evolved independently.
Given the above considerations, I really do think that to argue that an organ like the eye — or an analogous organ — couldn’t be feasibly front-loaded is a gaps argument. If specific proteins can be front-loaded, and if specific molecular machines and specific biochemical pathways can be front-loaded, you’re just pushing the goal-posts back and demanding that I explain how something like the eye could be front-loaded. Admittedly, at the present moment, this is a difficult question to answer — and this would be a good line of research, but as the evidence shows, specific biological objectives can be feasibly front-loaded.
The Branching Problem
This objection to FLE was raised by Dr Adequate, who stated:
According to your hypothesis, LUCA must have had genes to turn it into a whale, and a spider, and an oak tree, and a camel. So what decides which it will actually do?
This question is a bit flawed to begin with. I am not proposing that specific species or genera or families or orders were front-loaded. The initial genomes would not, therefore, contain the genomic information necessary to give rise to oaks and camels and whales and spiders.
However, I do propose that the first genomes were capable of front-loading the appearance of both animals and plants. How would this be possible? Recall that front-loading is about stacking the deck such that the appearance of certain biological features becomes much more plausible.
Consider an analogy from a game of cards (say, poker). If we loaded a deck of cards with aces, such that 25% of the card deck contains aces, it would be very likely that we’d get a hand with four aces — far more likely than if we didn’t tamper with the deck. Similarly, if we loaded a deck of cards with queens, so that another 25% of the deck contains queens, we would be very likely to get a hand with four queens. Importantly, it would be very probable that one person will get a hand with four queens, while another gets a hand with four aces. We’ve front-loaded the appearance of four queens and four aces so that these combinations are far more likely than with a standard deck. Note that, with this analogy, it is perfectly feasible for one person to get four aces (I shall call this front-loading objective x) and for another to get four queens (front-loading objective y) - even though this would be a branching pattern, where both front-loading objectives x and y have been realized.
So, the first genomes could be loaded with genes necessary for plant development and function and with genes necessary for animal development and function. And thus the origin of both plants and animals would be effectively anticipated by the first genomes. It’s not like an evolving animal lineage must use plant genes; nor does an evolving plant lineage have to use animal genes.
I’m going to use another example here, albeit a hypothetical one. Suppose we designed plant that (a) was prone to many gene duplication events, had closed veins, its bundle sheath organelles were enhanced, and had a photorespiratory CO2 pump, and (b) had globin proteins that could, in a number of mutational steps, be converted to hemoglobin. In effect, a population of these designed plants would be poised to evolve the C4 photosynthesis pathway and the hemoglobin molecule — front-loading both biological features. If Dr Adequate’s objection is indeed valid, then we must believe that this hypothetical plant population could not front-load the appearance of both C4 photosynthesis and hemoglobin (I am, of course, supposing that there’d be a selective advantage to both of these biological features). If I understand his objection correctly, he’s suggesting that only one of these features could be front-loaded. But this obviously not logical; as we have seen, it wouldn’t be very hard to front-load the C4 photosynthesis pathway, nor would it be difficult to front-load the appearance of hemoglobin. Nothing is stopping the front-loading of both of these features. Furthermore, it would be indeed possible to front-load hemoglobin and C4 photosynthesis such that they do not occur in the same population. All that would be needed is for this plant population to split into two populations, say A and B. A simple deletion event of the globin gene in population A would lead to the appearance of only C4 photosynthesis; deletion of the four enzymes involved in C4 photosynthesis in population B would lead to the appearance of only hemoglobin. This simple example shows how Dr Adequate’s objection is answered: there is no conceptual problem with different lineages being front-loaded.
Predictions of the FLE Hypothesis
It is important for a hypothesis to make testable predictions. Here, I will try to briefly describe how the FLE hypothesis makes predictions that are not made by conventional theory. Before beginning, however, I would like to point out that, in this thread, I do not intend to discuss in depth the issue of whether some of these predictions have, in fact, been confirmed. In this thread, I am primarily interested in discussing if these predictions differ from those generated by conventional theory.
Let me begin with a prediction concerning the origin of molecular machines like cilia. Intra-flagellar transport (IFT) particles are involved in ciliary function in most eukaryotes. These proteins contribute to ciliary function, and any eukaryotes that lack these IFT proteins — such as Plasmodium -- are probably degenerate cilia and do not represent the structure of the last cilia common ancestor.
The point is this: under the non-teleological framework, co-option events are primarily responsible for the origin of this motility organelle and its IFT proteins. Under this model, random co-option events of proteins in the cell resulted in the functional association of different proteins, which would have been preserved by natural selection — and over time, through repeating this step, finally a cilium arose. This is, in essence, the non-teleological hypothesis for the origin of the eukaryotic flagellum.
Given that the existence of Metazoa seems to require the existence of cilia, under the FLE model, cilia were front-loaded. How would cilia be front-loaded? The FLE hypothesis is only at its beginning stages, so one should not expect, at the present moment, a rigorous FLE model for the origin of the cilium. However, I can offer a cursory model for the FLE origin of the cilium. In this model, the first genomes would be designed with components that would later be used by the cilium. In other words, homologs of the core, necessary IFT proteins would be designed into the first genomes. They’d be given a function, such that their basic 3D shape is conserved over deep-time. If they were given a function where their 3D shape would be substantially changed over deep-time, then the front-loading designer couldn’t possibly hope that when these proteins associated, their shapes would complement each other correctly such that a cilium could arise.
From here, we can develop our FLE prediction. The non-telic hypothesis for the origin of the cilium does not require or predict that the prokaryotic homologs of IFT proteins be well-conserved in sequence identity. In fact, it’s certainly possible that the non-telic hypothesis predicts that most of the prokaryotic homologs of the core IFT proteins will be loosely conserved in sequence identity: a protein that is not under stringent functional constraints will be more likely to be co-opted into a novel role by chance without being deleterious. For example, H4 histone is one of the most highly conserved proteins in eukaryotes. To me, at least, it seems that it would be much more likely that if H4 histone was duplicated and then co-opted into an entirely novel function a non-adaptive effect would occur than if a fibrinopeptide, for example (which are not at all highly conserved), were co-opted into this novel role. This would be an interesting line of research, but I don’t intend to explore this argument further, because the fact remains: the non-telic hypothesis for the origin of the cilium does not require or predict that the prokaryotic homologs of core IFT proteins be well-conserved in sequence identity, while the FLE hypothesis for the origin of the cilium predicts that the prokaryotic homologs of core IFT proteins would be well-conserved in sequence identity, more so than the average prokaryotic protein. This is a testable prediction: we would need to find a prokaryotic homolog of a core IFT protein, and then conduct pairwise comparisons of that IFT homolog with its prokaryotic orthologs, and check its degree of sequence conservation. There is nothing in the non-telic hypothesis that predicts this hypothetical prokaryotic homolog will be highly conserved in sequence identity, more so than the average prokaryotic protein. You will not find anything like this prediction in the scientific literature. IMHO, this is an exclusively teleological/FLE prediction.
Deep Homology and Front-loading
I argue that the FLH predicts that proteins of major importance in eukaryotes and advanced multi-cellular life forms (e.g., animals, plants) will share deep homology with proteins in prokaryotes. I have discussed this prediction with various critics of the FLH, and the most common objection seems to be that non-teleological evolution also makes this prediction. I disagree, so let me explain.
Life seems to require a minimum of about 250 genes (Koonin, Eugene V. How Many Genes Can Make a Cell: The Minimal-Gene-Set Concept, 2002. Annual Reviews Collection, NCBI) — a proto-cell would not require that many genes. Thus, it would be perfectly acceptable, under the non-teleological model, that the last common ancestor of all life forms had approximately 250 genes, add or take a few. From this small genome, gene duplication events would have occurred, subsequently followed with mutations in the new genes, would lead to a novel protein. Over time, then, and through gene and genome duplication/random mutation, this small genome would evolve into larger genomes. This model is perfectly acceptable with the non-teleological hypothesis, and the non-teleological hypothesis does not predict otherwise. However, this model — where a minimum genome gradually evolves into the biological complexity we see today, through gene duplication, genome duplication, natural selection, and random mutation — is not compatible with the front-loading hypothesis. This is because front-loading requires that the first genomes have genes that would be used by later, more complex life forms. Of the 250 or so genes required by life, none of them could encode proteins that would be used later in multicellular life forms (excluding the proteins that are necessary to all life forms). A front-loading designer couldn’t possibly hope to stack the deck in favor of the appearance of plants and animals, for example, by starting out with a minimal genome.
Look at it this way. With a minimal genome of 250 genes that are involved in metabolism, transcription, translation, replication, etc., evolution could tinker with that genome in any way imaginable, so that you couldn’t really front-load anything at all with a minimal genome. You couldn’t anticipate the rise of animals and plants. Such a genome would not shape subsequent evolution. If the last common ancestor of all life forms had a minimal genome, and if you ran the tape of life back, and then played it again, a totally different course of evolution would result. But if you loaded LUCA with genes that could be used by animals and plants, you could predict that something analogous to animals and plants would arise. If you loaded this genome with hemoglobin, rhodopsin, tubulin, actin, epidermal growth factors, etc. — or homologs of these proteins — something analogous to animal life forms would probably result over deep-time.
Given that you couldn’t really front-load anything with a minimal genome consisting of about 250 genes, under the front-loading hypothesis, it is necessary that LUCA contain unnecessary (but beneficial) genes that would later be exploited by more complex life forms. Non-teleological evolution does not require this. It has no goal, unlike front-loading. It tinkers with what is there — and if a minimal genome was all that was there, it would tinker around, eventually producing endless forms most beautiful as Darwin so famously put it. On the other hand, front-loading is goal-oriented: a minimal genome does not allow one to plan the origin of specific biological objectives.
Thus, under the front-loading hypothesis, we would predict that important proteins in eukaryotes, animals, and plants will share deep homology with unnecessary but functional proteins in prokaryotes.
Non-teleological evolution does not predict this. Non-teleological evolution could explain that observation, but it does not predict this. And this is the important point to understand. There is nothing in non-teleological evolution that requires multi-cellular proteins to share deep homology with unnecessary prokaryotic proteins — but front-loading demands this. There is nothing in non-teleological evolution that requires that LUCA have a genome larger than the minimum genome size — but for front-loading to occur, this must be the case. I conclude, then, that this prediction is made by the front-loading hypothesis, but it is not made by non-teleological evolution, and so front-loading is certainly testable.
P.S. Often in lengthy essays like this, typos will be a bit frequent. I’m sure all of you will forgive any typos
Also note, again, in this thread I do not intend to discuss whether the predictions of front-loading have been confirmed to any degree. I intend to discuss whether these are indeed valid predictions of FLE.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.
Edited by Genomicus, : No reason given.
Edited by Admin, : Reduce image width, change image background to white, remove link to image, center caption. It's a poor quality image, even at full size.

Replies to this message:
 Message 116 by Dr Adequate, posted 02-24-2012 12:58 AM Genomicus has replied
 Message 117 by Dr Adequate, posted 02-24-2012 2:17 AM Genomicus has replied
 Message 118 by PaulK, posted 02-24-2012 2:20 AM Genomicus has replied
 Message 158 by Taq, posted 02-24-2012 1:45 PM Genomicus has replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 108 of 216 (653725)
02-23-2012 11:01 PM
Reply to: Message 83 by Wounded King
02-22-2012 11:00 AM


Re: The best of error minimizing codes?
Massimo Di Giulio (2001) argues that this actually makes approaches such as Freeland's tautological since the nature of the genetic code has had a significant effect on the distribution of amino acid substitutions making it far from an independent measure of the optimisation of that code.
Quite right. Freeland et al. (The Case for an Error Minimizing Standard Genetic Code), 2004, responds to this argument thusly:
Attempts to obviate this problem by measuring similarity directly from estimates of the frequencies with which amino acids substitute for one another within real proteins do suggest the code to be close to a global optimum (Ardell, 1998; Freeland et al., 2000a), but have been criticized as tautologous given a correlation between the code and patterns of substitution (Di Giulio, 2001b), though the flow of causality in this correlation is yet to be determined. Encouragingly, a recent attempt to derive a multidimensional measure of amino acid similarity that is truly independent from the code supports the counter criticism: the more sophisticated our representation of similarity, the better the code appears (Gilis et al., 2001).
For the record.
Edited by Genomicus, : No reason given.

This message is a reply to:
 Message 83 by Wounded King, posted 02-22-2012 11:00 AM Wounded King has seen this message but not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 109 of 216 (653727)
02-23-2012 11:08 PM
Reply to: Message 85 by Dr Adequate
02-22-2012 11:21 AM


Re: The best of error minimizing codes?
Wait ... scientists are disagreeing about something? This has hardly ever happened before.
Actually, it happens all the time. Investigators do disagree on the origin of the type III secretion system, for example (i.e., there is a good bit of disagreement on whether the TTSS and the bacterial flagellum are sister groups or whether the TTSS descended directly from the flagellum).
According to your hypothesis (if you fudge it a bit) it would be jolly nice if the genetic code was globally optimal. But no-one has proved this to be the case.
Not meaning to quibble over semantics here, but science isn't so much about proving things as it is about providing evidence. And evidence has been presented by various researchers that the genetic code is very close to being globally optimal for error-minimization - it is an extremely optimal code.

This message is a reply to:
 Message 85 by Dr Adequate, posted 02-22-2012 11:21 AM Dr Adequate has replied

Replies to this message:
 Message 147 by Percy, posted 02-24-2012 5:49 AM Genomicus has seen this message but not replied
 Message 152 by Dr Adequate, posted 02-24-2012 6:10 AM Genomicus has not replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 110 of 216 (653728)
02-23-2012 11:16 PM
Reply to: Message 94 by Wounded King
02-23-2012 6:42 AM


Re: Catching up
I just performed the same search, on the NCBI site limiting the search to the Prokaryota, and there were indeed hits below the 1e-05 level, a grand total of three of them and all from the same thing, a putative paired box domain protein from the protobacteria Methylocystis. I imagine that the sequence similarity is why this is identified as a putative paired box domain protein in the first place.
The region that comes up only covers 23% of the submitted amino acid sequence and has a maximum identity of 30% and less than 50% for positive (BLOSUM62 compatible) sites. If this was what he got then I'm not surprised Genomicus didn't want to go into any more detail about the results since they are so weak.
That the matches have >30% identity isn't that bad at all, when you factor in E-values. You can have a database match to your query sequence that has little over 20% sequence identity, but if the match is accompanied by a good E-value, it's still significant.
Perhaps more relevantly all of the lower hits are from tranposon and insertion sequences, suggesting that what we are seeing is a convergent signature for DNA binding interacting activity rather than an ancestral relic of a front loaded primeval PAX gene lineage. To support this the SMART database of protein architecture identifies 3 bacterial hits for the PAX domain and all 3 are from transposase sequences.
This may very well be the case. But from a non-telic perspective, do you think we'll ever find significant PAX6 sequence/structural matches in prokaryotes?

This message is a reply to:
 Message 94 by Wounded King, posted 02-23-2012 6:42 AM Wounded King has seen this message but not replied

Replies to this message:
 Message 114 by Dr Jack, posted 02-23-2012 11:52 PM Genomicus has replied

  
Genomicus
Member (Idle past 1941 days)
Posts: 852
Joined: 02-15-2012


Message 111 of 216 (653729)
02-23-2012 11:23 PM
Reply to: Message 95 by Chuck77
02-23-2012 6:56 AM


Re: A summary
Can anyone someone lay his position out for me in laymans terms? Is he arguing that evolution was front loaded by a creator/deity?
I'm arguing that evolution was front-loaded by some intelligence(s), the identity of which I haven't got the slightest idea - and for all we know, this intelligence might be extinct by now.
The FLE hypothesis is not a theistic argument, and the existence of gods (or lack thereof) has no bearing on this hypothesis. It is, however, a teleological, ID hypothesis.

This message is a reply to:
 Message 95 by Chuck77, posted 02-23-2012 6:56 AM Chuck77 has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024