front loading: did evos get it backwards

The full text of the paper is freely available online here.By 'type' the paper seems to be discussing protein superfamilies and PFAM protein domains.It is an obvious prediction of common descent that many modern lineages will not contain genes their common ancestors possessed. That this has turned out to be the case is not surprising. Simply looking at the diversity we see in modern bacteria and unicellular organisms would suggest there was probably a wealth of genetic diversity which may have been available to the latest common ancestor of plants, animals and fungi which is not represented in these more recently derived lineages. It is likely that similar losses occurred in particular lineages when the eukaryotes evolved.I'm not sure you should really be adding full stops to quotations when you are actually truncating the quote.You quote Loomis saying "appears to lead to LOSS [my caps] of genes." adding the full stop and leaving out ... You are right about specialisation involving gene loss but to overemphasis loss and entirely excise diversification seems somewhat misleading.TTFN,WK

becomes ... Did someone declare opposite day and just not tell me? Or is it just that you have as poor a grasp of what I said as of what the paper says?TTFN,WK

I object to you taking something I say about the latest common ancestor of certain lineages and representing it as concerning the earliest common ancestor.If you can't differentiate those two concepts you are seriously out of your depth.TTFN,WK

I don't agree. What the paper shows is that what diversity there was in the latest common ancestor has been retained more by Dictyostelium than any other group, it doesn't say that the ancestor had more types of gene than any modern group. In this respect I think Loomis statement is an exaggeration going beyond what the data will support.Indeed if we look at Figure 6 from the paper we can see that the number of PFAM domains which must have been found in the common ancestor of plants, animals and Dictyostelium, following the papers logic that Dictyoselium split from the metazoa after the split form plants, is 938. We get this from taking the genes common to animals and plants (which includes the numbers in boxes within those shared domains) and the ones shared by Dictyostelium and plants (the boxed number in the blue plants domain). The metazoa however have 1560 PFAM domains, 502 of which are unique to the metazoa. Even if we include the fungi and say that the LCA's complement was 1092, which is all of the genes shared by any 2 or more groups, it is still smaller than the complement of the metazoa. Even if we add in the SFAM superfamily data there are still only another 261 families to play with, not enough to make up the shortfall to the metazoans.In the paper the authors conclude ... This seems much more moderate language than the quote you provide from Loomis, he may simply have been exaggerating the impact of the research as happens all too often when research is being presented for media consumption. I just can't see how the data presented in the paper can support the claim that 'The cells which gave rise to plants and animals had more types of genes available to them than are presently found in either plants or animals'. Perhaps Loomis was relying on other data outwith their research, who can say?The statement may well still be true, there could easily have been gene types in the LCA which haven't survived in the organisms studied in the paper or indeed in any modern lineage. We will probably be able to expand the shared number of gene types as more organisms are sequenced, but at the same time we are likely to be expanding the number of novel or unique gene types. The number of gene types ascribed to the LCA might grow to exceed that for individual class lineages, as it already does for the fungi, but I think it is unlikely. Having said that the quote is exaggerated there has still obviously been considerable gene loss of particular domain types among different lineages since the latest common ancestor. The data shows at least 34 PFAM domains which have been lost in the metazoan lineage. I don't see anything supporting your scenario however, especially since there is a large disjunction between types of gene as identified by PFAM domains and the actual number of genes in terms of discrete protein coding sequences. It is worth noting that a discrete protein coding sequence is now becoming a somewhat obsolete definition of a gene, but it will do in this instance.Even if the trend since the LCA had been a net decrease in gene number it would not go against our theories of how evolution operates or known evolutionary mechanisms. However, it would be contrary to what our current models of evolutionary history would predict; such models are based upon an assumption of parsimony. If you want to make an argument against parsimony as an assumption in phylogenetics then you should feel free, although maybe a new thread would be more suitable for that. This is in no way supported by the evidence. A significant role for gene loss is certainly undeniable, but I don't see how it is a problem for current evolutionary theory. I certainly don't see any basis for claiming gene loss is the primary mechanism of evolution. Because the genes involved are not solely involved in the generation of the complexity you are talking about, especially when we are not talking about individual genes but about protein domains which can be shared by hundreds of genes. If one ancestral shared protein with a given domain subsequently diversifies into multiple genes would you not consider that a basis for increased complexity?TTFN,WK

Did you actually read the paper? Do you think it shows more gene types in the LCA than in extant groups? Do you have an opinion? Or did you simply decide that since it sounded like what you wanted to hear it was automatically good science and got a free pass from any actual scrutiny? I remember the paper but how you describe it is pretty well scientifically illiterate. I went into some detail on your presentation of this paper on the Criticizing neo-Darwinism thread, see my response in Message 263. I have no problem with these conclusions, a brief glance at endoparasitic organisms would be sufficient to confirm the large scale loss in some lineages, but it is a far cry from being evidence for front loading. This isn't just scientifically illiterate, it seems to be actually illiterate. Did you hit send a bit too soon?TTFN,WKEdited by Wounded King, : No reason given.Edited by Wounded King, : No reason given.

Really? Care to provide any substantive argumentation about the comparative morphology of ancient single celled organisms and metazoa? Or even the comparative genetics? You haven't shown anything to substantiate this. The paper you presented in the OP certainly doesn't. No one other than perhaps creationists posits the starting point being an organism with a complex master genome. All we know is that there is research showing that some common ancestral lines, i.e. the metazoa in the case of A. millipora and the eukaryotes in the paper from the OP, are more complex than was previously estimated. As I pointed out in my post this is because we use a parsimony based model to estimate these things, obviously the assumption of parsimony will tend to underestimate the extent of a common ancestral genome more the fewer modern organisms are used to estimate the common ancestral genome.Would you like to address this point and suggest how we could operate under any assumption other than parsimony?But the fact that we are refined the data for these LCAs and expanding the estimated size of the LCA genomes involved does not equate to the sort of ludicrous front loading scenarios you posit.Your case is certainly not helped by your equivocating between gene types and genes. As has been pointed out repeatedly gene types only refers to large overarching superfamilys and protein domains, each one of which can represent from one to a thousand or several thousand genes. Additionally the importance of gene regulation over gene complement is being more and more appreciated as we find out more about gene regulatory networks and especially since the discovery of further layers of regulation in the higher order structures of DNA and non-coding RNAs. And there is not a shred of evidence contradicting this. You still don't seem to grasp that an estimated LCA genome more complex than a previous prediction is just that, it doesn't suddenly become more complex than all modern species just because you want it to. Can you tell us how the morphology of the putative LCA compared to that of primitive plants or metazoa? Beyond it sharing basal eukaryotic traits and some suggestions from the data in the OP I don't see that you would have anything to go on, and a lot of those proteins didn;t seem to have any morphological role at all, and why should they, single celled organisms tend to put their complexity into biochemistry more often than morphology. Similarly what do you think you can say about their genetic complexity, other than the latest estimates using a wider selection of organisms indicate a higher 'gene type' complement than previous estimates?The most important issue to address is that of parsimony, it is the reason why we would expect to underestimate the diversity of an ancestral genome on the basis of only a few descendant lines. I'd rather you address that point than anything else in this post.TTFN,WK

No matter what you think, parsimony is the issue. The OP paper's findings don't contradict parsimony in any way. The fact that after looking at a wider variety of organisms, by including Dictyosteliumn, they estimate a larger gene type complement for the LCA is a result of parsimony.By parsimony we assume that identical or very high identity genetic features only arose once. If such a feature is shared by 2 species we assume it was present in their common ancestor. If we find 1 species out of a number which has a unique gene then we assume that the gene arose in that 1 species lineage rather than being lost in all of the others. Those are the fundamental principles of maximum parsimony, building phylogenies on the basis that they involved the least change neccessary.Obviously then if you add another species into such a maximum parsimony phylogeny you cannot reduce the estimated gene complement of the LCA, since it is already as parsimonious as it can be. All adding a new species can do is increase the estimated genetic complement of the LCA, as in both the cases you have brought up, or leave the original estimated size as it is. I would suggest that given how few species most of the common LCA genomic estimates are based on, and the ever increasing number of sequenced genomes, we must expect to see increases in the size of the estimated LCA genomic complement.Do you disagree with this analysis? If so why? No it isn't, or if it is you haven't provided any research to suggest it is, certainly not either of the papers you have referenced here. If you think they do make this case then give us your reasoning don't just give us your conclusions. Rubbish. There is no single front loader model, I have seen several suggested and they would make quite different predictions. If you have a particular favoured actual testable model then tell us what it is and what you think it predicts. Is it the supergenome type model? You think that the LCA for either all life or for the LCA of each created kind/baramin had all the genetic diversity represented in all its ancestors?Simply saying 'Front loading' tells us nothing because there are several alternative 'Front loading' scenarios, i.e. John Davison's prescribed evolutionary hypothesis is a 'Front loading' hypothesis but doesn't require an ancestral supergenome. I have even seen people argue that simple de novo creation of an ancestral cell with the potential to evolve can be considered 'Front loading', see here.Whatever the shortcomings of 'Front loading' as a specific term the other contention, that this contradicts neo-darwinian theory, is simply wrong. It suprecedes previous estimates of the LCA genomic complement for exactly the parsimony related reasons I described previously, it doesn't predict the opposite at all. If you have a reasoned argument why it does then make it, if you haven't then stop saying it does. Again you say something scientifically nonsensical. The latest common ancestors aren't the most primitive organisms. This is exactly what you did before, mixing up the concept of an earliest common ancestor, the most primitive organism, with that of the latest. Bald assertion, none of your references support this, you have stretched hyperbole to the breaking point. I have made some arguments in the criticising neo-darwinism thread addressing this ... It would be naive in the extreme to assume that just because an organism has a bigger genome it should be a bigger organism, or even necessarily a more complex one. Having said that it should be noted that in some cases of polyploidy, i.e. whole genome duplications, the resultant species are larger, the prime example being the frog Xenopus laevis. Nonsense, there are numerous examples of endoparastic organisms with reduced genomes, which would correspond to your genetic/morphological parallelism. There are also multitudinous examples of comparatively simple organisms with comparatively simple genomes. To go from the finding that LCAs may be more complex than expected to saying they are as complex or more complex than extant modern species is a leap from science into science fiction. Did A. milipora suddenly have an increased genome size just because it shared a gene with 'higher' metazoa? No. A completely unsupported assertion as you would realise if you understood parsimony. Show us the evidence and explain the reasoning which gives you LCAs with genomes as complex or more complex than all the following lineages. Don't just refer us off to some papers, because none of the papers you have referenced so far support your conclusion, if you still contend they do then make the blooming argument and demonstrate how they do. Maybe it is since there is no clear definition of what 'Front loading' is or what we should expect to see from it.TTFN,WKEdited by Wounded King, : Some grammatical foibles

I didn't mean that statement to cover all creationists, since creationist beliefs are highly heterogeneous. People who argue from 'front loading' are predominantly creationists, there may be a few UFO cultists as well but mostly they are creationists.Having said that ... ... makes no sense at all to me are you saying that a planet itself is the common ancestor, are you being pointlessly metaphorical? Are you making the 'common creator' argument but mixing the words up to make it confusing and impenetrable?TTFN,WK

That depends on what you consider a feature. If you look at morphological traits then you are correct if we consider similar traits, the question is whether something similar is the same as something identical. We think that proteins are homologous because of their shared sequence identity not merely because they perform the same function. Therefore similar functional morphological traits, as we see from convergent evolution, need not have a similar genetic basis. Similarly parallel evolution may not have a similar genetic basis, or perhaps I should say identical genetic basis as the same gene can be affected in a multitude of different ways and different mutations on the same gene could be considered a similar genetic basis.If you have some clear examples of parallel or convergent molecular evolution I would be very interested. There are several cases, such as armor changes or spine loss in Sticklebacks which show that the same genes can be identified in parallel evolution, but offhand I can't think of any examples of identical mutations.Your argument stands for morphological phylogenies however, which is why we don't ideally base phylogenies on single traits but on a wide range. The principle is still one of parsimony but the fuller the picture the more outlying convergent traits between distant species require more changes to be incorporated into a common ancestor.To explain further, if we have very similar characteristic which evolves multiple times, lets take for example trap jaws in spiders, see here for a phylogeny of spiders showing the occurence of trap jaws, then a parsimonious explanation should involve the least changes possible. The basis of the original tree is genetic sequences rather than morphology. You can see that for the trap jaw trait to be the result of common ancestry, even in the more closely related Ponerinae and Ambyloponinae clades, we would be requiring the loss of the trait at least twice. Two losses of the trait is no more parsimonious than two gainings of the trait, which we require in the absence of common ancestry, when we go to the more distantly related clades you would need as many as 6 independent losses of the trait compared to 3 independent gains under the assumption that it isn't commonly derived.Perhaps I simplified too much and should have caveated my statement with, 'all other things being equal'. No, because there is a world of difference between being more complex than the previously estimated LCA genome and being as complex or more complex than any existing descendant species. If you had understood any of what I have been saying about parsimony, or even understood the concept itself, you would see that parsimony will tend to underestimate the size of the LCA genome, especially when it is only estimated from a few fully sequenced genomes. This doesn't mean that the true LCA genome would have been as complex or more complex than the genome of any existing descendant species. What I have emphasised time and again is that with parsimony we would only expect to see the LCA genomic complement expand as more species are added.Now were extensive convergent genetic evolution or even protein evolution, to occur then you would be right to suggest that in fact our parsimonious reasoning could lead us to overestimatethe gemonic complement of the LCA. However, as with the trap jaw trait in my example as we added more species, in this case their genomes, to our phylogeny the less parsimonious and overestimated LCA would become.As yet I am aware of no compelling evidence for significant parallel or convergent genetic evolution producing identical genetic bases for traits. I would consider substantive evidence of such a phenomenon being widespread to be quite good evidence for some sort of 'hidden' information form of front loading.The 'super genome' form of front loading produces lots of non-parsimonious tress as shared genetic traits from distant species would require multile independent losses in various species not sharing that trait despite its presumably having been present in the supergenomic LCA. Good point, what I am saying is that there aren't even these commonalities between 'front loader' models. The predictions of a 'supergenome' type front loading are completely opposite to those of a 'hidden information' type front loading argument or somone like the writer at telic thought who thinks that 'front loading' consists merely of the de novo creation of an organism with the potential to evolve further complexity. In fact only the 'supergenome' form of front loading seems to agree with your prediction of a LCA as complex or more complex than any descendant species, the others would tend to agree more with the standard evolutionary models. Enormous hyperbole to describe your own totally unsubstantiated interpretation of data. How on earth could anyone get that from you writing, 'the most primitive organisms of the latest common ancestors'? Except you are once again storming ahead of the evidence. The evide hasn't failed to 'pan out'. As I seem to keep having to reiterate a more complex LCA than previously estimated is by no means an LCA as complex or more complex than any of its descendant species. If you think you actually have some evidence showing an as complex or more complex genome then present it. Don't just keep going, 'but look how surprised these evos were!!!!!', it isn't an argument or at least it isn't a good argument. Because, as you say, there is a tendency to expect simpler morphologies to correspond to simpler genomes, and previous estimates have predicted simpler genomes for the metazoan LCA, that much is quite correct. It's where you go from there, leaping ahead of the evidence, that is problematic. That's what the current models of the metazoan LCA had predicted.I think I need to point out that it would by no means contravene evolutionary theory to have a common ancestor to a clade like the metazoa which was more complex than its descendants. What it would contradict are all of our current models of the evolutionary history of life on earth, which are principally based on parsimonious assumptions. It would contradict evolutionary theory if the LCA were to have fully developed genetic networks for the development of structures which the LCA did not posses. I know that is how you have tried to skew the A. millepora (I think before I was spelling that millipora, like the water filters) research but it just won't go that far. In my reply to you on the criticising neo-darwinism thread, see my Message 263, I describe known alternative roles for the genes 'associated with highly differentiated nervous systems', which don't have anything to do with nervous systems but with very early stage embryonic development.TTFN,WK

I've been on holiday and haven't had a chance to look over this thread again yet, I hope to get to it in a day or two.Just for now though I would point out that ... Is not supported by the following extract from the paper, 'three regions consistently mutated in all monkeys studied', does not speak to identical genetic mutations occurring, so while they may have described this as convergent it by no means necessarily fits the criteria we have been talking about. I haven't had a chance to read the paper yet, as I don't have access at home, but the abstract given does not support the specific contention. Similar genes isn't the question. I'd be perfectly happy to stipulate that similar genes are probably involved. It's whether there is a specific shared genetic basis at the sequence level that is the question. So as you predicted what? As you predicted you still predict the same thing? Is this to give the impression that people are now doing this even though they aren't?As I said, hopefully in a couple of days I 'll be in a position to address this thread at greater length.TTFN,WK

That isn't so much simple as vague. Could you maybe narrow it down a bit? Do you mean 'What does NeoDarwinism predict the latest common ancestor of the metazoa should be like genetically?'?TTFN,WK

Do you still stand by this claim? If so can you actually tell us what some of this evidence is. At the moment it just seems to be a quote from that article not backed up by the actual research paper.You said earlier that it was fine if I disagreed with Loomis statement but do you agree with it? If so why? what evidence supports it?TTFN,WK

So before we actually discuss the topic you want to discuss something totally different. Why not start another thread on findings that would contradict ND? Don't just keep repeating the same claims, why not show us some specific instances where 'front loader's specifically predicted findings like this? What specific hypotheses? I already suggested what I thought would constitute some strong genetic evidence for front loading. Definitely not, horizontal gene transfer (HGT) is one finding which is not supportive of ND in any substantial way, but it certainly isn't contradictory of it and once a gene has been introduced by HGT it seems to behave in accordance with ND population genetics. Similarly an ancestral genome at some point which was larger than some or even all descendant lineages is not something which supports or contradicts ND, but is still consistent with it.As has already been pointed out the predictions of ND concern genetic evolution with respect to fitness and the spread of traits through populations. Even expanding this to include different evolutionary scenarios for the origins of genetic novelty does not produce anything the research you have brought up would contradict. All the research does is expand the minimal predicted genome of the latest common ancestor of the groups in question.You go on to reiterate ... Would you care to try and substantiate this from the paper yet? You seem happy to keep repeating the claim over and over again without apparently knowing if it is true or exactly what it means. Please either address the criticism of this point I have pointed out to you several times or stop making the same misleading claim. To refresh your memory ... Care to address this yet, or will you admit that 'complex nerve function' is not the be all and end all of these genes. None of the papers you are touting show that simpler organisms don't have simpler genomes, all they show is that their genomes are not as simple as the previous simplest estimates of them. You have yet to produce any evidence showing simpler organisms don't have simpler genomes than their 'complex' modern counterparts. I have already detailed several data from the paper the OP is discussing which show that modern metazoan lineages have more gene types than the estimated LCA genome the paper posits, see Message 25.Havign said that, as I pointed out previously, some LCA could have a more complex genome and not contradict ND, but it would have to have had a simpler ancestor itself at some point. We can accept almost any level of fluctuation in size and genetic complexity throughout lineages evolutionary history but they must have originated at some point from a genetically simpler form. Just pointing at one point on the lineage and saying, 'look how complex this is' shows us nothing other than the complexity at that particular point. it certainly doesn't produce a pattern of genetic change which contradicts ND. Don't these 2 scenarios produce 2 potentially contradictory expectations of what we would predict to see? If not then how do the scenarios' predictions differ? Can you actually cite any specific prediction made?TTFN,WKP.S. Please start spelling theoretical correctly! I'm not a grammar or spelling Nazi but when you use a word so often it gets annoying when you persistently misspell it

Please no-one fall for Randman's subtle trap. The largest known genome belongs to an amoeba, Amoeba dubia.Of course this organism isn't 'earlier evolved' as such, even if its 'form' was.TTFN,WK

Why didn't you? Surely ... is better served by actually pointing out the fallacy than by asking a purportedly hypothetical question which you already know the answer to?TTFN,WK