Deep Homology and Front-loading

Well, try to either have your cake or eat it. If the homologues are sufficiently useful that natural selection would conserve them, then they will in fact be conserved, and their conservation is in fact predicted by the ToE.

Either:

(1) The histone homologues are useless to prokaryotes, in which case FLE does not predict their conservation and can't even explain it either, since no mechanism for conserving them is known (conservative selection being ineffective for this purpose by hypothesis) ...

... or:

(2) The histone homologues are useful to prokaryotes, in which case the ToE predicts their conservation by natural selection.

Edited by Dr Adequate, : No reason given.

Yes, but that's not the issue here at all. I said that the FLE would predict that proteins crucial to eukaryotes will share deep homology with proteins in prokaryotes.

Well, so does Darwinian evolution. How could they not?

The only way I can think of you could not have such homology is if you had a LUCA front-loaded with two disjoint sets of genes, one for prokaryotes and one for eukaryotes. Which would be a front-loading hypothesis.

Edited by Dr Adequate, : No reason given.

1) Darwinian evolution (that is, the current paradigm) does not predict that crucial proteins in eukaryotes will share deep homology with functional but unnecessary proteins in prokaryotes.

So the emphasis is on functional but unnecessary?

But then you could say the same of front-loading. What is there in FLE that says that by the modern era, the only homologous proteins won't be those absolutely necessary to both lineages?

2) Unlike front-loading, Darwinian evolution does not necessarily predict that eukaryotic histones will share deep homology with prokaryotic proteins.

How likely, though, is a complete turnover of genes?

A cool example of this is T-urf13, which, IIRC, evolved from rRNA sequences and other non-coding sequences.

Could you clarify --- do you think that this, and things like it, actually happened?

---

You haven't addressed my point about front-loading. What is there in the concept of FLE as such which stops the LUCA from having two non-intersecting sets of genes, one for eukaryotes, one for prokaryotes? Now, I know the evidence is that the LUCA wasn't like that, but if there's nothing to stop a designer from doing something like that, then there's no prediction, is there?

Edited by Dr Adequate, : No reason given.

If I understand your question correctly, the answer is: if you loaded the first genomes on earth with proteins "in waiting" that will find their intended function only later, when complex life forms have arisen, then, in the modern era, you can trace this homology back in time. Naturally, these proteins "in waiting" will carry out a function such that it isn't likely that they will be weeded out by natural selection.

But the bit of that that made sense is Darwinian.

Let's try it again.

You make a big deal out of the fact that things that are essential to eukaryotes are merely useful to prokaryotes.

(I must have missed the place where you proved that, but let's take this as true for the sake of argument.)

But where does FLE predict that some things that are essential to eukaryotes are merely useful to prokaryotes? Is not FLE consistent with the idea that everything that is essential to eukaryotes is also essential to prokaryotes?

Well then, if FLE is consistent with that, then it does not predict the opposite of that.

Well, it's likely enough to happen.

But again I have to ask if you believe what you're saying.

If you believe that Darwinian evolution is sufficient to produce a complete turnover of a genome --- then this would be the time for me to point out that you do not deny that Darwinian evolution occurs. You apparently believe in Darwinian evolution plus front-loaded evolution. So if you believe that a complete turnover of the genome is "likely enough to happen" given Darwinian evolution, then you believe that it is "likely enough to happen". In which case, where is your prediction? If you admit that given FLE, it is "likely enough to happen", then FLE has no prediction that it wouldn't happen.

Yes. That's what the evidence indicates happened in the case of T-urf13.

But ...

... you admit that Darwinian evolution did that ...

One of the reasons that it's hard to argue with you is that it's difficult to know what it is that you think. I mean, I know that you're trying to slip ID in there somewhere, but ...

... I'll come back to this point when I know what you wish to say.

Well, in this case, we'd have an either/or prediction. In other words, FLH predicts either that the LUCA will have functional and unnecessary genes or that the LUCA will have non-intersecting sets of genes.

But then you're screwed.

Don't you see?

We can't actually look at LUCA, can we?

We can only look at what's in front of us, the modern genomes of modern organisms.

Now, when you've broadened your hypothesis to saying: "A front-loaded LUCA could be like this, in which case it would explain this; or, alternatively, a front-loaded LUCA could be like that, in which case it would explain that" ... then you've lost the point you were originally meant to be gaining.

We can't look at the genome of LUCA. We can just look at the genomes of the organism that are presently available to us. Now, if all FLE predicts is that we will either see evidence that something is true, or we will see evidence that the opposite is true, then that's not a point in favor of FLE, is it?

You say you have an "either/or prediction". Well, how does that differ from the "either/or prediction" of Darwinism in terms of the phenomena that we can actually observe?

Edited by Dr Adequate, : No reason given.

Let's take a look at calmodulin (CaM) as an example. Calmodulin binds to calcium and is found universally among eukaryotes. This fact, coupled to the observation that its sequence identity is exceptionally conserved across taxa, suggests that eukaryotes require calmodulin for their existence. In a world without calmodulin (or a functional analog), would eukaryotes exist? Probably not. Now, it appears to me that you're saying "is not FLE consistent with the idea that calmodulin is also essential to prokaryotes?"

No, you're missing the point.

You can produce as many cases as you like of things that are essential to eukaryotes and merely useful to prokaryotes. Knock yourself out. It doesn't help

But is there anything in FLE that predicts that such a thing should exist? Would it not be compatible with FLE that everything that's essential to eukaryotes should also be essential to prokaryotes? Where does FLE rule that out?

Well, first of all, you said "turnover of genes;" I didn't think you meant a complete genome. And, come to think of it, the terminology "turnover of genes" is kinda vague in this context. Care to elaborate?

Well, suppose that prokaryotes started off with a certain set of genes, let's call them genes 1 ... 1000. Then gene 1001 arises by the mechanisms you suggest, and substitutes for gene 1 and displaces it, and then gene 1002 arises by the mechanisms you suggest, and substitutes for gene 2 and displaces it ... and so on until every gene in LUCA has been substituted by a gene arising by the mechanisms you postulate.

We can just imagine that happening. But is it likely?

I'm not sure how you're defining Darwinian evolution.

Evolution by known mechanisms: reproduction, mutation, natural selection, lateral gene transfer, recombination ... and so forth.

No, because FLE has an objective, while non-telic evolution does not. Thus, non-telic evolution could stitch together a protein from non-coding segments of a genome, "just happening" to produce a new protein. In short, if FLE relies on the cobbling together of proteins to reach an objective, it is extremely unlikely that a specific, intended outcome will be produced, given that there are trillions of different possible configurations for pieces of DNA sequences. Non-telic evolution has no goal, so it can cobble stuff together, and stumble on novel functions.

* sighs *

But will you make this clear?

You believe, do you not, in random mutation and in natural selection?

Therefore, you believe in "non-telic evolution" don't you?

So don't you believe that "Non-telic evolution [...] can cobble stuff together, and stumble on novel functions"?

To make it clear:

You apparently believe that Darwinian evolution happens. You apparently believe that Darwinian evolution allows for a complete turnover of genes, such that there would be no homology whatsoever between eukaryotes and prokaryotes. Well, in that case your model does not predict that there would be any homology between eukaryotes and prokaryotes, because your model incorporates Darwinian evolution.

You're trying to have your cake and eat it again. If Darwinism allows this sort of thing to happen, then your model also allows this sort of thing to happen, because your model permits Darwinian evolution.

Yes, we can. Not directly, of course, but indirectly.

But as I have pointed out before, our ability to look "indirectly" at LUCA depends crucially on accepting Darwinian evolution. That is the theory which allows us to infer LUCA. If we deny Darwinism, we have no basis on which to say what LUCA looked like.

Well, Darwinian evolution "predicts" that either the LUCA will have a minimal genome, or the LUCA will have more than a minimal genome. Confirmation of any of these "predictions" is compatible with Darwinian evolution. It should be noted that something that predicts every scenario isn't really predicting anything, and this seems to be the case with non-telic evolution when it comes to the gene content of the LUCA. However, FLE is not compatible with the former scenario (that of a minimal genome). Thus, if we find that the LUCA actually did have a minimal genome, FLE will have been falsified. Which means its testable, and that it makes a specific prediction Darwinian evolution does not make.

* sighs *

But again, we don't get to look at LUCA.

Let's look again at the data, according to you.

You claim that:

(1) There are certain proteins that are essential to eukaryotes.

(2) These proteins have homologues in prokaryotes such that prokaryotes would not actually drop dead if deprived of these proteins, but would be significantly disadvantaged if deprived of them.

How does FLE predict this state of affairs? How would I, or you, or anyone, reason from "LUCA was front-loaded" to this conclusion?

Edited by Dr Adequate, : No reason given.

Edited by Dr Adequate, : No reason given.

Huh? It's not FLE that rules that out, it's basic observations that rule it out.

* sighs, bangs head against desk *

You don't seem to have got the hang of this whole prediction thing.

Look, suppose you said that giraffes confirm FLE because FLE predicts giraffes.

And then I come back saying: "But is there anything in FLE that predicts that giraffes should exist? Would it not be compatible with FLE that there should be no giraffes? Where does FLE rule that out?"

And then you riposte with: "Huh? It's not FLE that rules that out, it's basic observations that rule it out."

These "basic observations", of course, being that we can see giraffes.

Don't you see that in order for a theory to predict X, it has to rule out not-X? You can't say: "My theory predicts X ... OK, I grant you it doesn't rule out not-X ... but we can see that X". Well in that case your theory hasn't made a prediction. We've just seen X.

I'm not suggesting anything like that happened. I'm saying that some key eukaryotic proteins, under the non-telic model, could have evolved through patching together different chunks of non-coding sequences.

But unless you claim that it is a prediction of Darwinism that there should most likely have been a complete genomic turnover, then you must acknowledge that it is a prediction of Darwinism that there should be homologies.

If that's how you're defining "non-telic evolution," then yes. But when I say "non-telic evolution" I mean this whole non-teleological framework which says that planning and foresight was not involved in the origin of genetic diversity on earth (with the exception of human intervention).

But if you admit that Darwinian evolution can and does take place, then none of your predictions can be based on the premise that it can't and it didn't.

Therefore, you must admit that FLE permits anything that ToE permits. In which case FLE can never be more specific in its predictions than ToE.

And, in this context, what does "Darwinism" mean? Random mutation coupled to natural selection (and similar mechanisms)?

Yeah, all known mechanisms, plus common descent.

I would be the first to admit that "Darwinism" is not the best word for that, but for want of a better, I'll go with that.

1. Front-loading, by definition, involves loading the first genomes with proteins that are unnecessary for prokaryotes but are necessary for the origin of eukaryotes and multicellular life forms.

No. That is not the definition of front-loading.

Possibly I'm not being clear enough.

If you were clear enough, then I think that you yourself would reject your own ideas.

Edited by Dr Adequate, : No reason given.

I sometimes find it useful to think of it as a court case, this is something concrete that we can all grasp.

You are the counsel for the prosecution.

You: If John stole the money, then the money would be missing. The money is missing. Therefore John stole the money.

Me: But that doesn't show that Tom or Dick or Harry didn't steal the money. To pin the crime definitively on John, you'd have to show that if John didn't steal the money, then the money would still be in the till.

You: But it is a matter of basic observation that the money is not in the till!

That doesn't get us anywhere. It is common ground that someone stole the money, and that it is therefore missing. But in order to pin the crime on John you have to prove that if it wasn't for John the money wouldn't be missing. Otherwise the fact that the money is missing isn't evidence against John.

Yes, but you're saying that FLE doesn't rule out the possibility that all essential proteins in eukaryotes will be essential in prokaryotes.

And I should like you to answer that point rather than another one.

The non-teleological model predicts both (a) that eukaryotic proteins will share homology with prokaryotic proteins, and (b) that eukaryotic proteins will share homology with non-coding regions of prokaryotic genomes. On the other hand, FLE exclusively predicts the former.

Where does it do that?

Again, if by Darwinian evolution you simply mean random mutation and natural selection (and other mechanisms), then yes, of course Darwinian evolution happens. But my prediction is based on the two different models for the origin of life's diversity: the teleological, front-loading model and the non-teleological model.

But so far as I can see, you believe in both.

In which case your range of predictions includes anything that Darwinism says is possible, and so cannot have greater specificity than Darwinism.

No, I disagree, precisely because FLE doesn't permit the LUCA's genome to consist of little more than a minimal gene set.

But, one more time ...

WE CAN'T SEE LUCA.

Any predictions have to be about what we can see in the present. If you admit Darwinian mechanisms, then anything that Darwinian mechanisms allow us to see in the present is allowed by your hypothesis. Which therefore cannot be more specific in its predictions than Darwinism.

Actually, by definition, FL does involve the loading of a genome with proteins that the basic prokaryotic cell architecture doesn't require but the eukaryotic cell plan does.

No. That is not part of the definition of front-loaded evolution.

Edited by Dr Adequate, : No reason given.

I answered your point by bringing up the example of ERVs.

No, that's where you ducked my point.

I think I've already answered this question, but here goes: Front-loading basically doesn't work if it must rely on some of the important eukaryotic proteins evolving purely through the chance stitching together of random DNA sequences. This is because the odds of this process generating a specified target is next to nil.

But that is still not an answer to my question.

For example, what's to stop LUCA from having two parallel sets of genes, one for prokaryotes and one for eukaryotes?

Now, please note once more that it is not to the point to observe that this the evidence we have now shows that this didn't happen --- in order for there to be a prediction, you need to show that the very concept of FLE precludes doing it that way.

Okay, I think I understand what you mean with this one. All FLE predictions (that I am aware of) include things that are possible under "Darwinism." However, there are some possible scenarios in Darwinian evolution that isn't allowed by FLE.

Which? Mutation still happens, doesn't it? And natural selection? And lateral gene transfer? And recombination?

Anything that can happen under Darwinian mechanisms can happen under the FLE hypothesis, since it does not deny the existence of those mechanisms.

True or False: We can detect this needless complexity by comparing the genomes of different species, allowing us to track back over deep time.

False.

Elaborate, if you please.

Well, in that you didn't actually answer it.

I already explained this: we have an either/or prediction in this case. Under front-loading, either the LUCA has more than a minimal genome, or LUCA has two parallel gene sets. Darwinian evolutionary models also generate either/or predictions.

True. But your claim was that (in this respect) FLE had greater specificity than Darwinism. If you now admit that FLE could have been done so that there'd be no homology at all between eukaryote and prokaryote proteins, then your point has vanished. We might see homology, we might see none. Or we could see any degree of homology, depending on how much overlap there was between the two gene sets. Also, the overlap could have only included absolutely essential proteins. Your "either/or prediction" is that in the case of any given protein, either we'll see homology between eukaryotes and prokaryotes or we won't. That seems to cover everything. Adieu predictive specificity.

You will note that I said "there are some possible scenarios in Darwinian evolution that isn't allowed by FLE." At this point, I wasn't talking about mechanisms.

Ultimately, this seems to be a distinction without a difference. Darwinian evolution allows every scenario that is allowed by the mechanisms. If you admit the mechanisms, you admit the possibility of every such scenario.

Really? Why?

Well, we can't see LUCA. You have no reason to suppose that, just because the histone genes are not absolutely essential to modern prokaryotes, they were not absolutely essential to LUCA. Maybe in this respect it was more like a eukaryote.

(BTW, I have no idea where you're getting the information about these proteins. Presumably this came up on another thread. Could I have a reference please? Thank you.)

Edited by Dr Adequate, : No reason given.

If so, then you might want to elaborate. Because reading it from here, it looks like you're basically saying "is there anything in common descent that predicts ERVs should exist? Would it not be compatible with common descent that ERVs do not exist? Where does common descent rule that out?" Ignoring, of course, the simple observation that ERVs do in fact exist, and from this observation we can generate a prediction.

Given that ERVs exist, we can predict something else, namely the pattern of their distribution. But what you're doing is given that histone homologues exist in prokaryotes, you're predicting the existence of histone homologues in prokaryotes.

It's just the same as the giraffophile designer I was talking about on the other thread. Given the hypothesis of a designer, we can't predict giraffes. So you throw in the premise that the designer wanted to make giraffes. Hurrah, now we can predict giraffes! Such specificity the hypothesis has! Yay!

In the same way, the designer might have made all sorts of decisions about what LUCA should be like. You need to add onto the FLE hypothesis the additional ad hoc supposition that the front-loading was done so as to produce the outcome we see. And then, lo and behold, your hypothesis predicts the outcome we see.

I could do the same with Darwinism. Let's add to the known mechanisms of Darwinism the hypothesis that the sequence of mutations and the selective pressures were just such as to produce the observed outcome. Whoopidie-do-dah, now I have just as much predictive specificity as you do. But I've done it the same way --- by adding in, ad hoc, extra hypotheses that fit my theory to the facts but do not follow from my theory.

Is this your question?

The question was: "Would it not be compatible with FLE that everything that's essential to eukaryotes should also be essential to prokaryotes? Where does FLE rule that out?"

Now, your solution seems to be to add in, ad hoc, the hypothesis that the designer didn't design things that way. Also apparently he likes giraffes but not unicorns.

1. The same thing could be said for Darwinian evolution. There's nothing in Darwinian evolution preventing the LUCA from having parallel gene sets, is there? 2. More importantly (I suppose a case could be made that it's unlikely from a Darwinian standpoint to have parallel gene sets in LUCA), the predictive value of the FLE does not lie only in the homology aspect, but in the fact that this homology is shared with prokaryotic proteins that are unnecessary for life.

Except that it doesn't predict that, does it?

To see that this is so, consider the case in which the homologous proteins were in fact all essential. Would that be cause for us to say: "Front loading definitely didn't occur, then. If there was a designer, there would definitely be inessential proteins in prokaryotes homologous to essential proteins in eukaryotes. We now know for certain that life had no designer and that LUCA was in no way a product of intelligence"?

No, of course not. It would give us no warrant to say that whatsoever. In which case it is not a prediction of FLE or of any other form of ID that such a homology should exist. If seeing ~X does not falsify theory T, then theory T does not predict X.

Do you agree that the basic prokaryotic cell plan does not require all the proteins necessary for the eukaryote and multicellular life? If so, then you must admit that it is possible that the LUCA did not require eukaryotic proteins. That this possibility was not manifested, however, is what front-loading argues. In other words, under non-teleological evolution, it is possible for the LUCA to have been structured such that it did not require key eukaryotic proteins. But if this is found to be the case, and moreover it is discovered that the LUCA did not in fact have key eukaryotic proteins, FLE is effectively falsified? Make sense?

I can't make sense of this as an answer to what I wrote.

My point was this. You suppose that LUCA had inessential proteins. You point to histones as an example of this. That LUCA had histone-like proteins you (rightly) deduce from the fact that they (or their homologues) are present in both prokaryotes and eukaryotes. But you also seem to be claiming that we know they were inessential to LUCA because we know that they are inessential to modern prokaryotes. But that is not a legitimate deduction. In this respect, LUCA might have had more resemblance to a modern eukaryote than to a modern prokaryote. Or not, of course. Who can say?

I'm not quite sure what information you're referring to.

About the role played by histone homologues in modern prokaryotes.

No, I'm not. It goes like this: given that protein X is required for eukaryotic existence, I predict that protein X likewise exists in prokaryotes, which is expected from front-loading.

But that isn't a prediction of FLE as such, is it? If prokaryotes lacked histone homologues, you'd just shrug and say that they dropped out of the prokaryote lineage just as (I presume you would have to say) the ability to synthesize peptidoglycan dropped out of the eukaryote lineage.

If it wouldn't damage FLE if they weren't there, then FLE doesn't predict that they'll be there.

Sigh. I keep answering this question.

That's as may be; but you keep getting the answer wrong.

Is it true or is it false that proteins required by eukaryotes are not required for the existence of carbon-based life (more specifically, DNA-based life)? If that is true, then we have our FLE prediction: proteins universally required by eukaryotes will share deep homology with proteins in prokaryotes that are not required for the existence of life.

Well, no. You've already admitted that FLE is consistent with having two completely separate suites of genes, one for eukaryotes, one for prokaryotes. In which case, FLE does not predict any homology.

I think you're missing a crucial point. You say "consider the case in which the homologous proteins were in fact all essential." I assume by that you mean that these proteins are essential to both eukaryotes and prokaryotes. But it is an objective fact that eukaryotes require proteins that are not required by all types of prokaryotic life forms. And from here is where we get our prediction.

Or: "I think you're missing a crucial point. You say "consider the case in which giraffes did not exist." But it is an objective fact that giraffes exist. And from here is where we get our prediction." Quite so. You got your "prediction" from seeing what was true and then retconning your hypothesis.

Look, it's very simple. It would be consistent with front-loading that all the homologous proteins were essential to both eukaryotes and prokaryotes. The world isn't like that, but there would be nothing about FLE as such that would prevent the world from being like that. In which case FLE does not predict that the world is not like that.

You keep stumbling on this point. Obviously if you constrain your hypothesis with reference to what you know post hoc to be true, then you can get it to predict true things. But it goes round in a circle. "FLE could have produced that. But we don't observe that. Therefore, FLE didn't produce that. Therefore the front-loading was done in such a way as not to produce that. Therefore FLE predicts that we won't see that."

As I've pointed out, I could do exactly the same thing. "Darwinism could have produced unicorns. But there are no unicorns. Therefore the mutation and selection events were such as to produce no unicorns. Therefore, Darwinism predicts an absence of unicorns". But Darwinism as such predicts no such thing. I had to throw in some extra hypotheses which were tailor-made to account for the absence of unicorns --- an ad hoc hypothesis to explain a post hoc observation.

Well, we've got a lot of key proteins in eukaryotes that are not found in prokaryotes. So we can actually make a prediction from FLE that isn't ad hoc, namely that we'll find homologs of these proteins in prokaryotes.

Very well.

Consider ubiquitin, a highly conserved protein found in all eukaryotes, which marks down proteins for proteasomal degradation. This article lists it among the proteins unique to eukaryotes. (Later on, we might look at some of the others.)

Now, according to WP: "The proteasomal degradation pathway is essential for many cellular processes, including the cell cycle, the regulation of gene expression, and responses to oxidative stress." It seems, then, that ubiquitin is indispensable to eukaryotes.

And yet according to this article: "No ubiquitin homolog has been found in prokaryotic genomes sequenced thus far".

Prokaryotes instead use a protein known as prokaryotic ubiquitin-like protein, which is functionally analogous but non-homologous.

So, if you're right about what FLE predicts, then goodnight FLE. But wait! I can just hear you now. "It is a matter of objective fact that ubiquitin has no prokaryotic homologue. Therefore, FLE doesn't predict that there should be one ..."

I find it suspicious that when I said "Evolution predicts that if the bacterial flagellum evolved, a number of its components will share similarity with proteins that are more ancient than the bacterial flagellum" you made no issue of it. Yet, now that we are discussing FLE you're bringing up the possibility of homologs being lost in various lineages. In other words, if you want to bring up that aspect of all of this, then you've got to admit that Darwinian evolution actually doesn't predict that the bacterial flagellar components will share homology with more ancient systems.

I never said it did. You did. There is in fact no a priori reason why they should still be hanging around. They might well have been lost instead. There's nothing in Darwinism per se that says this shouldn't have happened. If you want to admit that you were wrong about this too, go right ahead.

Edited by Dr Adequate, : No reason given.

What do you mean by "no LUCA"? Do you mean that modern life would have evolved from a pool of different species? I have no problem with that ...

In which case, once again, FLE does not predict any deep homologies at all.