|
Author
|
Topic: The GUToB
|
peter borger
Member
Member # 722
|
posted 05. June 2003 21:50
Hi Mesk,
You ask me:
"If ACTN3 possessed important functions for most of mammalian evolution (as it does, for instance, in mice), and those functions were only lost during the evolution of the human lineage, is there any paradox here for evolutionary theory?"
First of all you say here "evolution did it". That is your axioma. I do not have an axioma. I just look at the data and give an unbiased interpretation of the data. You also look at the data, but you give it a evolution-based interpretation. In my opinion, standard evolutionary theory is not explanatory and thnks to molecular biology becomes more and more an outdated theory. It should be upgraded. However, if ACTN genes have important FUNCTIONS in whatever genome they will be kept stable in the genome since inactivation will be selected AGAINST (=GUToB). If in a similar genome the genes are more or less redundant they are allowed to get inactivated (as observed for dozens of genes in the human MPG, including ACTN3, lots of olphactory receptor genes, GLO, etcetera). Furthermore, I have pointed out my comments on the ACTN3 gene in my previous letters: What we most likely observe in these genes in primates is NRM similar to what is observed in lamin. Or 1g5. Or whatever gene. A lot of genes demonstrate NRM.
Besides, if you want evolution to have done it why can it not be so that evolution is subject to rules and laws? Like everything else in the universe. That would be my question to you.
You and I agree that it is possible that the 577X allele is present in chimpanzees at a low frequency, and this would be consistent with our current hypothesis that the 577X allele has been maintained for a long period of time by balancing selection. However, it should be claer that this mutation can also be present due to NRM.
Furthermore, it has been shown that these pointmutations in the ACTNs gene are in disequilibrium. Do you observe double mutants? That would advocate my hypothesis of NRM. Or do you propose selection on exons?
Furthermore, you say that my assertion that we do not know of the interchromosomal translocation mechanisms to be patently false. Since 'Such a result could be achieved in a number of ways by well-understood molecular mechanisms - for instance, intrachromosomal duplication followed by an interchromosomal translocation event'. If so please describe the latter phenomenon in a molecular way.
And if the 2R hypothesis is by no means the only mechanism, please elaborate a bit on the other utter random mechanism that give a [(A)(B,C,D)] topology. That would interest me.
You also say that "Modern evolutionary theory is more than capable of explaining the genesis of gene families."
I think I have to further disappoint you since it is not. Standard theory is also NOT able to explain the src-family. If you wanna discuss it be my guest.
And you say "I strongly recommend that you familiarise yourself with the extensive literature on this topic before you make any further claims along this vein."
I heard this before. The usual response when standard theory can't hold.
Finally you say that "there is an association between gene duplication and redundancy - in that recently duplicated genes are more likely to be redundant than other genes - but this fact alone does not explain the high degree of functional redundancy in genomes." It seems that network buffering or functional degeneracy plays a significant role here. So, on this planet we have an evolution in favor of the most rigid organisms? Evolution towards organisms that are unable to change? Sound a lot like a direction to me.
You seem to be puzzled 'Why the reliance on the duplication-redundancy association as an indicator of the validity of evolutionary theory? But isn't it obvious that from an evolutionary stance duplication is the sole and only source of genes? Genetic redundancies also must have their origin in gene duplication. At first, any duplicated gene is redundant and will be be lost, since there is no selective constraint on such genes. The organisms having duplicated DNA regions will be selected against since they require more energy and time to duplicate their DNA (as observed in bacteria). And if they are able to keep it in their geneome due to selective constraint they will be kept pure (purified) as observed for the high abundant genes (for instance tRNA's). The result is that they will never change. Now we find that there is not an association between redundacies and duplications and thus the source of such genes is an evolutionary riddle. Maybe the evo-fairy did it?
You say that "stronger predictions of evolutionary theory - the congruence of morphological and molecular phylogenies, for instance - would be better targets." They can be easily demonstrated not no be valid all time. It can also be alternatively explained from the GUToB as NRM that give the illusion of common descent. Besides it simply depends on what you are looking at. If you are looking at genes you probably will find a more or less congruence since similar DNA sequences will give similar NRM(Although this can be demonstrated not to be valid for several genes). But what about chromosomal make up? The chromosomes of primates do not advocate an utter random mechanism and common descent. Superficially it seems like evolution, that's true, but a close up scrutiny reveals GUToB.
best wishes
Peter
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 06. June 2003 00:01
Peter, are you *ever* going to give a detailed demonstration, theoretical or with data, that posulates a specific model of NRM and shows how it leads to a false conclusion of common ancestry?
If not, I would appreciate it if you'd just say so, so that I don't have to bother checking this thread any longer.
This is central to your theory, yet hasn't been demonstrated.
Performing this demonstration is essential, as it allows one to quantify levels of NRM necessary to fool tree-building, and thereby predict what various experiments will find regarding mutations. Without that, you can only end up in a did-not did-too shouting match with people who create evolutionary models, except the people with trees of common descent can attach significance and goodness-of-fit to some of their measurements, while you can only attach rhetoric.
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 06. June 2003 00:23
Hi Pim,
Pim, you say that NDT did not assume them to be random but observed them to be random with respect to fitness. I know all about that, but it probably not true in genes that change in respons to external stimuli (in 'evolutionary' arms races for instance).
And I am aware that they are not random wrt location in the genome, or even wrt time does not negate that these changes seem to be random wrt fitness (utility in the present environment). However, the NRM hypothesis is an alternative to the common descent hypothesis.
Furthermore, you say that my examples do not show that there is an illusion of common descent. In fact one of the worked out examples showed otherwise. It pertain the 1g5 genes an demonstrated that NRM are present in these genes. You may interpret them as the result of common ancestry, but I think it cannot hold. Let's have another look at the 1g5 genes in drosophila. If we only had analysed the sequences in the Italian, Peruvian and USAsian populations your immediate conlusion would be: common descent (based upon the shared pointmutations in positions 161, 178, 211, 334 and 346. However, since we have additional mutations (peru and japan AND australia and USA) and shared mutations in the intron immediate common descent is unlikely (it pertains a neutral region). So, my point is that you can not jump to the conclusion of common descent based upon a couple of shared mutations. If you are going to make appropriate conclusions you better exclude a common mechanism.
You say -and I agree- that if NRM is refering to mutations non random wrt location or time but not fitness then NRM is purely Darwinian. But you should agree with me that it may have severe implications for phylogenetic analysis. If the position of a mutation is determined by the properties of the DNA and the nucleotide determined by the biochemical apparatus than shared mutations do not say anything about common descent. But rather all about common origin.
You say that I may have misread the PNAS article, since it does not pertain the same positions but rather in the same areas or hot spots.
However, figure 2 in the PNAS paper demonstrates the hotspots as point mutations and the authors say that 'a similar profile of homoplasy at nps 16093-16390 was reported. (Meaning that the point mutations arose independently in exactly the same location (=NRM1).
In addition, you like me to show that such mutations would lead to common descent. It is, however, fairly easy to understand that if such mechanism also operate at the level of the genome, chimps living in the same region would suffer the same point mutations in the same genes, since they have the same original -but shuffled- MPG. And GUToB says: same DNA sequences plus same biochemistry plus same background equals shared mutations. So, do we really require a worked out example?
And you are interested to find out why common descent infered from DNA seems to match common descent infered from anatomy and fossils. Grosso modo, it may have a functional origin. There is some evidence for that. I will look into it in more detail. Shared point mutations can be explained as above.
Best wishes,
PB
IP: Logged
|
|
gordon
Member
Member # 781
|
posted 19. June 2003 12:53
After reading your comments in the 'manifesto' thread, I thought I would check out this one.
Mr.Borger: quote:
But you should agree with me that it may have severe implications for phylogenetic analysis. If the position of a mutation is determined by the properties of the DNA and the nucleotide determined by the biochemical apparatus than shared mutations do not say anything about common descent. But rather all about common origin.
If this were so, it seems to me that any and all phylogenetic analyses would produce more or less random results.
Phylogenetic analyses are performed by sequencing loci in a few or many living creatures. First and foremost, all the creatures being examined have to have the gene or locus in question. As such, if I understand your position, all creatures possessing the identifiable gene or locus will, by definition, have a similar sequence, and so mutations would be 'directed' by the physical properties of the DNA sequence.
Therefore, all creatures possessing the locus/gene should, according to you, be equally likely to have generated/accumulated the same distribution of substitution as any other creature.
So, upon sequencing, as there should be nothing other than pure random activity producing mutation in this locus/gene, we shouold expect not to find any identifiable patterns.
According to what you are saying.
This, however, if monumentally erroneous. If your idea were representative of actual mechanisms, should we sequence and analyze ten genes in ten different species, we could expect to see 10^10 different topologis (if I remembered my probabilties correctly).
That is hardly the case. I am familiar with Dr.Caporale's book and the citations she provides on the issue of phylogenetic impacts of such "directed" mutational hot spots, and what those papers say is not quite what you imply. It has been some time since I read them, but if I recall correctly, it is explained that most phylogenetic analysis programs take into account the potentially 'noisy' data and correct for it and that sufficiently long stretches of sequence data can compensate for any confounding signal.
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 23. June 2003 01:36
Hi Gordon,
It seems to you that that any and all phylogenetic analyses would produce more or less random results. And as a matter of fact they do.
The data presented in the journals are only a minority that are in agreement with common descent. It should be noted that a subdisciplin of phylogenetics works on reconciliation of trees: to bring the speciestrees in accord with gene trees. As discussed, the proposed IL-1beta gene to reconcile the trees is not present in the human genome (Yep, it got lost. Loss of genes is easy too explain, so I don't object. However, first evolutionism has to postulate a putative gene and than it is not present. Very convincing for a sceptic).
I have proposed GUToB as an alternative to common descent based upon phylogenetic analysis. It holds that shared mutations are related to initial shared DNA sequences indepent of inheritance (although once acquired such mutations can be propagated or back-mutate). There is increasing evidence for this vision.
You are quite correct that "phylogenetic analyses are performed by sequencing loci in a few or many living creatures. First and foremost, all the creatures being examined have to have the gene or locus in question. As such, if I understand your position, all creatures possessing the identifiable gene or locus will, by definition, have a similar sequence, and so mutations would be 'directed' by the physical properties of the DNA sequence."
Not 'directed' probably but rather 'pre-determined'. As demonstrated for T4 by Lynn Ripley. The mechanism here is an imperfect hairpin, and could also explain other shared mutations. Furthermore, radiation introduces the same mutaions on the same spot by an as yet unknown mechanism but DNA damage related. In general it is now agreed upon that the DNA sequence contributes to where the mutations are introduced. And it may even determine the type of nucleotide.
Thus, a group of creatures possessing initially the same DNA elements will end up to have accumulated a similar distribution of substitution as any group of creatures possessing that initially the same DNA sequence. This is exactly what is observed for a group of genes. Another group of genes does not change 'at all' over time, while yet another changes with such high rate that is can easily be taken as a seperate group of originally created MPGs (for instance the GH gene distinguishes the primates from all other MPGs). Or they are subject to a directing mechanism. For arms race genes this can easily be conjectured, but for instance the GH gene (that is highly distinct in the primates from other organisms [only 30-35% sequence identity] is not involved in an arms race (to my knowledge) and could thus be determined by something external.
It is probably not possible to discriminate between common descent (assuming random introduction of mutations and the subsequent inheritance; but which is most likely a false assumption) and a common mechanism. However, there is increasing evidence for common mechanisms modulating mutation rates and positions.
You say that "If your idea were representative of actual mechanisms, should we sequence and analyze ten genes in ten different species, we could expect to see 10^10 different topologies (if I remembered my probabilties correctly)."
These experiments have been carried out already and the conclusion is grosso modo in accord with common descent. But it is also grosso modo in accord with a common mechanism. Therefore we have to look at extremes. I did that for the ZFY gene which is in accord with a common mechanism rather than common descent.
You are familiar with Dr.Caporale's book and the citations she provides on the issue of phylogenetic impacts of such "directed" mutational hot spots, and according to you those papers say is not quite what you imply. That may be correct, but these papers do not solve the problem. Furthermore, one is only to detect NRM1 after extensive sequence analysis and intra- and inter species comparisons. Still than, shared mutations are regarded as ancient and inherited. That is why analyses carried out by Ripley are so straight forward. She observed exactly the same mutations that can not have been inherited. Similar finding have been observed for bacteria. Apparently, organisms can independently arrive at exactly the same DNA sequence/order. Due to a non-random mechanism. Easy as that.
You say that most phylogenetic analysis programs take into account the potentially 'noisy' data and correct for it and that sufficiently long stretches of sequence data can compensate for any confounding signal.
As long as such analysis are based on one or two sequences per organism I am not impressed. It requires a lot more to detect all NRM positions in even a small stretch of DNA. As demonstrated from the data I have previously shown in this thread: If you miss a couple of sequences for Drosophila subtypes you wouldn't identify the NRM positions. And you wouldn't have identified that the Ig5 gene is subject to a NRM generating mechanism. That the authors do not discuss this, tells me that NRM mechanism are not allowed in evolutionary biology. But that will have to change soon.
Best wishes,
peter
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 23. June 2003 19:48
For Gordon,
Of the 293 independent mutations identified within the lac-d sequence in the F plasmid in E. coli, 63% of the point mutaions are located at 19 MEDIUM level hotspot sites. (Schaaper et al, Genetics 129:317)
Of 120 base substitutions identified within the rspL sequence on a multicopy plasmid of E. coli, 63% are located at 9 hotspot sites.(Yoshiama et al, J Mol Biol 307:1195.
Interestingly, in the genome-integrated part of this plasmid 70% of 1555 base substitutions are confined only to TWO hot spot sites (Ann Rev Genetics 2002, p292).
In conlusion, a detailed look (meta analysis) of multiple genomes reveals NRM1 all around. It is rule rather than exception. It elegantly explains the shared mutations between species, since they have the same original MPG.
Best wishes,
PB
[ 23. June 2003, 19:48: Message edited by: peter borger ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 25. June 2003 00:05
Peter,
I explicitly showed that the two halves of the ZFY gene sequences show essentially the same tree, and that the tree is highly similar to the classical one inferred from morphology. I used phylogenetic analysis to discover that one of the genes in your list was not in fact ZFY but a related gene ZFX.
So when you say that you showed that changes in ZFY are due to common mechanism instead of common descent, you are flat-out wrong. Propose a model for common mechanism that can explain why the common descent trees from different parts of the gene end up the same, or stop making claims that have already been demonstrated in this thread to be false.
Repeating something ad nauseum doesn't make it true.
In case you have forgotten, reread page 4.
Also, you claim that only a minority of genes are in agreement with common descent. We can test that. We have the Drosophila, mouse, and human genomes available. Of genes that are shared between the three, we can get three possible branching patterns if we construct a history of common descent:
code:
(1)
----------------Drosophila
|
|----------Human
|---Mouse
(2)
----------------Mouse
|
|----------Human
|---Drosophila
(3)
----------------Human
|
|----------Drosophila
|---Mouse
What fraction of genes that are shared should fall into each tree, given your model of common mechanism? I would expect about equal numbers of each under your model.
Under common descent, I would expect (1) in essentially every case where the genes were statistically distinguishable from each other, and whose similarity was statistically above that of noise.
Added in edit: fixed mislabeling of branching pattern (2).
[ 25. June 2003, 01:52: Message edited by: Rex Kerr ]
IP: Logged
|
|
gordon
Member
Member # 781
|
posted 25. June 2003 12:38
Peter B. wrote: quote:
It seems to you that that any and all phylogenetic analyses would produce more or less random results. And as a matter of fact they do.
This is not a fact at all! This demands support, as it is a serious charge. quote:
The data presented in the journals are only a minority that are in agreement with common descent.
I find this to be a fantastic claim. Please provide documentation for this, as again, it is a serious charge. quote:
It should be noted that a subdisciplin of phylogenetics works on reconciliation of trees: to bring the speciestrees in accord with gene trees.
What subdiscipline is that?
quote:
I have proposed GUToB as an alternative to common descent based upon phylogenetic analysis. It holds that shared mutations are related to initial shared DNA sequences indepent of inheritance (although once acquired such mutations can be propagated or back-mutate). There is increasing evidence for this vision.
Frankly, I do not think so, and reading through this thread, it appears that there is none at all. It appears that the only “evidence’ you have is your conspiratorial claim about only “descent-friendly” analyses getting published.
I am quite familiar with the phylogenetics literature, and have performed phylogenetic analyses myself. And I have seen nothing like the ‘random’ tree topologies that you imply are the norm. What phylogenetic analyses have you done, as you say you have, and what were the results? Are they published? Can they be duplicated?
quote:
You are quite correct that "phylogenetic analyses are performed by sequencing loci in a few or many living creatures. First and foremost, all the creatures being examined have to have the gene or locus in question. As such, if I understand your position, all creatures possessing the identifiable gene or locus will, by definition, have a similar sequence, and so mutations would be 'directed' by the physical properties of the DNA sequence."
Not 'directed' probably but rather 'pre-determined'. As demonstrated for T4 by Lynn Ripley. The mechanism here is an imperfect hairpin, and could also explain other shared mutations. Furthermore, radiation introduces the same mutaions on the same spot by an as yet unknown mechanism but DNA damage related. In general it is now agreed upon that the DNA sequence contributes to where the mutations are introduced. And it may even determine the type of nucleotide.
You are restating the obvious and already acknowledged. If these and similar mechanisms are at work in homologous genes, then, again according to your vision, any and all phylogenetic analyses should produce random results, since any species possessing the similar DNA sequence will be equally likely as all others to possess substitutions at any correlating sites.
Simply claiming that there is a conspiracy to suppress non-descent friendly analyses seems a bit desperate.
quote:
Thus, a group of creatures possessing initially the same DNA elements will end up to have accumulated a similar distribution of substitution as any group of creatures possessing that initially the same DNA sequence. This is exactly what is observed for a group of genes.
This is just what I wrote, except for the last part, which, frankly, seems made-up. And if this were the case, then, as I pointed out and you claimed was the case, phylogenetic analyses should – all of them – produce random results. That is, say we examine 5 primate species: human, chimp; gorilla; orang; gibbon. Running an analysis should produce, if I remember my bifurcating tree probabilities, 5^4 alternate topologies, or some 600 possible combinations. Of course, putting realistic constraints on this but still allowing for the MPG factor, we could probably expect some several dozen outcomes.
Is that the case?
You have claimed evidence for the MPG at the now defunct link in the first part of this thread. Prior to the link’s demise, I downloaded the file and ran some simple analyses. No random patterns at all. Unless you can provide either some personal data or some primary source references, I kindly suggest that you drop the conspiracy charges as evidence for your position.
quote:
Another group of genes does not change 'at all' over time, while yet another changes with such high rate that is can easily be taken as a seperate group of originally created MPGs (for instance the GH gene distinguishes the primates from all other MPGs).
Or, more reasonably, one group is under tighter selective constraints than another? quote:
It is probably not possible to discriminate between common descent (assuming random introduction of mutations and the subsequent inheritance; but which is most likely a false assumption) and a common mechanism. However, there is increasing evidence for common mechanisms modulating mutation rates and positions.
If it would not be possible to discriminate, why then do you so confidently proclaim inheritance descent to be “most likely a false assumption”? Especially in light of the fact that inheritance has been demonstrated to be a mode of transmitting substitutions from one generation to the next.
quote:
You say that "If your idea were representative of actual mechanisms, should we sequence and analyze ten genes in ten different species, we could expect to see 10^10 different topologies (if I remembered my probabilties correctly)."
These experiments have been carried out already and the conclusion is grosso modo in accord with common descent. But it is also grosso modo in accord with a common mechanism. Therefore we have to look at extremes. I did that for the ZFY gene which is in accord with a common mechanism rather than common descent.
So, outliers prove the case? That does not seem to warrant serious consideration.
quote:
You are familiar with Dr.Caporale's book and the citations she provides on the issue of phylogenetic impacts of such "directed" mutational hot spots, and according to you those papers say is not quite what you imply. That may be correct, but these papers do not solve the problem.
No, they point out that there really is no problem when one takes these phenomena into consideration.
quote:
Furthermore, one is only to detect NRM1 after extensive sequence analysis and intra- and inter species comparisons. Still than, shared mutations are regarded as ancient and inherited. That is why analyses carried out by Ripley are so straight forward. She observed exactly the same mutations that can not have been inherited. Similar finding have been observed for bacteria. Apparently, organisms can independently arrive at exactly the same DNA sequence/order. Due to a non-random mechanism. Easy as that.
Of course organisms can arrive at the same substitutions. Non-random with regard to locus, not planning or direction. Phage and bacteria also, of course, have much less DNA than do the more complex organisms, and so are more likely to accumulate ‘identical’ mutations. That is what I alluded to with reference to one of Caporale’s citations – larger data sets can overcome such ‘noise’ (homoplasy).
quote:
You say that most phylogenetic analysis programs take into account the potentially 'noisy' data and correct for it and that sufficiently long stretches of sequence data can compensate for any confounding signal.
As long as such analysis are based on one or two sequences per organism I am not impressed. It requires a lot more to detect all NRM positions in even a small stretch of DNA. As demonstrated from the data I have previously shown in this thread: If you miss a couple of sequences for Drosophila subtypes you wouldn't identify the NRM positions. And you wouldn't have identified that the Ig5 gene is subject to a NRM generating mechanism. That the authors do not discuss this, tells me that NRM mechanism are not allowed in evolutionary biology. But that will have to change soon.
I am sorry that you are not impressed, but that has little relevance as to your scientific claims. What you call “NRM positions” geneticists call “polymorphisms.” You may not read about “directed mutations” or “non-random mutations” because, sensu stricto, they have been shown not to exist, and according to your usage of the terms, they do not influence phylogenetic analyses as you would have us believe. The authors do not discuss this as there is no reason to.
IP: Logged
|
|
gordon
Member
Member # 781
|
posted 25. June 2003 12:53
Peter,
The paper you cite deals with only a 210 bp segment of the gene with known ‘hot spots.’ The Yoshiama paper did not come up in a search, and I do not get the Annual Review, so I cannot address them.
As for the localization of substitutions to various mutation-prone regions in plasmids, I fail to see what this has to do with confounding phylogenetic analysis, causing trouble for evolution, or supporting NRM in any way. Substitutions are more likely to occur in certain loci. Pardon the expression, but big deal. The paper you cite deals with only a 210 bp segment of the gene with known ‘hot spots.’
What about the rest of the plasmid? What about the nucleoid DNA? That specific regions, again, are prone to acquiring substitutions, it strains credulity to consider this as anti-”Darwinian” evidence of support for what amounts to front-loading.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 25. June 2003 20:14
Incidentally, the number of topologies for a binary tree with N branches (N+1 nodes) is (2N)!/(2^N*N!)
E.g. for 2 nodes (one branch), we have 2/(2*1) = 1.
For 3 nodes (2 branches) we have 24/(4*2) = 3 (and they are (A,(B,C)), (B,(A,C)) and (C,(A,B))
For 4 nodes (3 branches) we have 720/(8*6) = 15, namely D(A(BC)) D(B(AC)) D(C(AB)) C(A(BD) C(B(AD)) C(D(AB)) B(A(CD)) B(C(AD)) B(D(AC)) A(B(CD)) A(C(BD)) A(D(BC)) (AD)(BC) (AC)(BD) (AB)(CD)
Beyond that it gets rather hairy to write out.
(5 nodes therefore gives 105 rather than 5^4=625 possible topologies.)
[ 25. June 2003, 20:15: Message edited by: Rex Kerr ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 25. June 2003 22:00
Listen guys,
You simply don't get it. If mutations are dependent on DNA sequence and biochemistry (and this is highly probable considering the latest data in the references I cited) than we are probably unable to dicriminate between common descent and common mechanisms. Commond mechanism are more similar in shared MPG and these will group together. That's all I say.
Morover, Gordon, how can I refer to sequences not appearing in literature? If you do a search for Roderick Page, you will find the literature on reconsiliation of gene trees with species trees. I think it is his specialty.
I will work out a model demonstrating my claims for Rex.
best wishes,
Peter
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 25. June 2003 22:11
quote:
--------------------------------------------------------------------------------
Another group of genes does not change 'at all' over time, while yet another changes with such high rate that is can easily be taken as a seperate group of originally created MPGs (for instance the GH gene distinguishes the primates from all other MPGs).
--------------------------------------------------------------------------------
Or, more reasonably, one group is under tighter selective constraints than another?
Wanna discuss the genetic redundancies (again)?
NB: Redundant genes do not have an association with gene duplication and they do not change faster than essential genes (refs in previous mails). It is clear evidence for a multipurpose genome.
best wishes,
Peter
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 25. June 2003 22:19
Gordon, you summarise the whole point here:
"As for the localization of substitutions to various mutation-prone regions in plasmids, I fail to see what this has to do with confounding phylogenetic analysis, causing trouble for evolution, or supporting NRM in any way. Substitutions are more likely to occur in certain loci. Pardon the expression, but big deal. The paper you cite deals with only a 210 bp segment of the gene with known ‘hot spots.’"
KNOWN hotspots. The other shared mutaions can also be hotspots. It is UNKNOWN. The whole phylogenetics is based on the asssumption that mutations are random. But they are not. You didn't exclude NRM either, did you? To exclude NRM in homolgous genes requires a tremendous amount of sequencing both inter and intraspecies. The refernces I cited demonstrate that meta analysis of sequences reveal NRM.
best wishes,
Peter
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 25. June 2003 22:31
quote:
peter borger:
It seems to you that that any and all phylogenetic analyses would produce more or less random results. And as a matter of fact they do.
The data presented in the journals are only a minority that are in agreement with common descent.
Peter, you have made many extraordinary claims in this thread, yet you have consistently failed to provide any hard evidence in support of them. This is hard to understand given that many of your claims - including the one above - could be easily tested using publically available data.
If your claim above is correct, then the vast majority of genes should generate sequence-based phylogenies which significantly differ in their topology, both from one another and from the most widely accepted phylogenetic trees. I challenge you to support this claim be performing the following experiment, which should take up much less than a week of your time: identify a large number of genes (say 100) which have recognisable homologues in human, mouse, fly and worm (four organisms with complete or nearly complete genome sequences). Do not apply any biased system in identifying these genes - simply pluck them out at random from the databases. Confirm that each gene has a single homologue in each of the four species, since the potential phylogenetic complications of gene families are well known.
Now generate phylogenetic trees for these genes using a variety of standard phylogenetic analysis tools - there are other members of this forum who have much more expertise in this area than I, and I'm sure they would be more than willing to help you perform this task.
Then present the results of your analysis for all of us to see - give us the method you used to select genes, the names of each selected gene, and the topologies of the phylogenetic tree(s) in each case. If your claim is correct and phylogenetic trees are essentially random, then the trees should be distributed evenly among all of the possible topologies. If your claim is false (and common descent is true), then virtually all of the trees should conform to a single topology - and that topology should be consistent with the most widely accepted phylogenetic tree, which is based on both morphological and molecular data.
Put the data where your mouth is, Peter. Let's see some evidence, rather than yet more unfounded claims of conspiracy and deception.
I look forward to examining your results.
Mesk.
[ 25. June 2003, 22:33: Message edited by: Mesk ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2003 00:32
Dear Mesk,
Conspiracy and deception?
That's what it is? I put forward a new idea. And thus I also have to show why and where standard evolutionary theory is wrong.
Time for something new. It is GUToB: Non-Random Mutation in a Multi-Purpose Genome (NRMinMPG). I and others have provided several lines of evidence for that. It's up to you to rebut NRM and their effects on phylogenetics and common descent. One has to exclude NRM before one publishes data and phylogenies. That's how science works: exclude possible alternative explanations.
Regarding your 100 (or so) genes. The first that springs in mind (at random) is H3. According to histon H3 I could be either a fish or a bird. And if I am well informed the multitude of H3 genes are kept stable in the genome by purifying selection (whatever it is).
Best wishes,
Peter
IP: Logged
|
|
|
Printer-friendly view of this topic