|
Author
|
Topic: Peter Borger: Shared mutations: Common descent or common mechanism?
|
Moderator
Administrator
Member # 1
|
posted 23. June 2006 06:39
Shared mutations: Common descent or common mechanism?
by Peter Borger
Abstract Mutations are a fact of life. Darwin gave mutations, which he called natural variation between individuals, a key role to explain the origin of species. The origin and nature of mutations is one of the most fundamental questions of biology, and are a hot topic in origin debates. If mutations are merely a matter of chance, then the alignment of mutations in distinct species that do not reproduce together qualifies as independent molecular evidence of common descent. We know now, however, that mutations are not utterly chance driven phenomena as the DNA context may determine to a considerable extent where mutations occur. If mutations are modulated because of biophysical mechanisms the question is not whether rules and laws determine where mutations are introduced, but rather – do non-random mutations affect phylogenetic analysis? The DNA analysis of the 1G5 gene in Drosophila melanogaster demonstrates that over 70 percent of the mutations that are shared between subpopulations of species that do not interbreed are independent of common descent. Likewise, over 50 percent of the mutations in the GULO pseudogene that are shared between humans and the great apes are mutational hot spots also found in guinea pigs – they exactly match the mutations that set humans and primates apart from the rat and line up independent of common ancestry. This paper advances a new hypothesis to understand alignment of mutations in homologous DNA sequences of separated species as the result of a common mechanism operating in similar genomes, and provides the first biological evidence that the location where a mutation will occur and the type of mutation (transition or transversion) are largely predetermined. The consequence is that we may not be able to discriminate between common descent and this common mechanism.
To read the entire paper, click here.
[ 23. June 2006, 07:04: Message edited by: Moderator ]
IP: Logged
|
|
Shi
Member
Member # 1923
|
posted 23. June 2006 15:23
Peter, Interesting paper. I agree with you that there are cases of non-random mutations. But you cannot refute, as you intend to do, the claim that most mutations are random. The facts are most consistent with the following view. The microevoulution from the first fish to many subspecies of fish involves mostly random mutations. In contrast, the macro evolution from fish to the first amphibian mosly involves non-random mutations of divine and purposeful origin.
If non-random mutations is more common than random and if shared mutations does not indicate common decent, we would expect a chaotic phylogenetic relationship. Indeed, you used the examples of drosophila in your paper to show that shared mutations has nothing to do with phylogenetic relationship. So, if you are correct, your thesis would predict a vastly different phylogenetic tree of species from the present tree or no tree at all. But the present tree is largely correct, consistent with common sense, and works in practise. This proves that your thesis that non-random mutation is more common and invalidates common decent is untenable. Human decented from yeast more recently than from bacteria is more consistent with facts than your view that human shows no more parental relationship to yeast than to bacteria.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 24. June 2006 13:51
Peter Borger, exciting work, thanks. Sorry it took me a while to get around to properly reading it. I have been very busy at work lately. In truth I am baffled at the lack of activity on this thread.
Let me just simplify what you said into layman's terms, to confirm that I understand.
First the fruit fly. If I understand correctly, the mutagenic tree (is that a correct term?) that one gets from the 1G5 gene in fruit fly analysis, a tree that assumes common ancestry and random mutation, produces a tree that is unexpected based upon the continents that the flies were taken from.
But more importantly, the analysis of the GULO gene, the vitamin C gene, suggests that, if random mutation is correct, the guinea pig is much more closely related to humans than it is to rats. Is that correct?
I seem to recall you suggesting that in the cytochrome C molecule, the rattle snake is surprisingly like humans as well, is this correct?
I get a little confused when you speak about mutational hot spots. At first glance it appeared that you are simply saying that there are locations within the DNA that are, well, mutation magnets. These places invite random mutations. However, such would only be semi-non-random, and would not truly be non-random. It also seems that you are suggesting that there is a specific "mechanism", yet undefined, which is causing these mutations. If this is so and becomes accepted as so, the naturalists will dig in and try to find that mechanism, to which I say "good on em".
In general, if there is a common mechanism, differentiating between that common mechanism, and common ancestry would be impossible when analyzing a single gene. However the result of a common mechanism should be that as different genes are studied, you get different mutogenic trees. Is that correct? I understand that gene studies leave some amount of confusion of whether man is more closely related to the great ape or to the chimpanzee. Am I correct?
Three more questions:
1 - Have you submitted this document for publication in a peer reviewed journal?
2 - Do you mind if I spread your link around to a couple of other forums?
3 - Somehow I was of the impression that you are Austrailan, yet your paper says that you live in Germany. Was my impression totally bogus, or are you an Australian living in Germany? Just curious.
IP: Logged
|
|
John A. Davison
Member
Member # 1425
|
posted 24. June 2006 16:22
Mutations never had anything to do with creative evolution. Even the genes that are likely to mutate have been predetermined as well as the alleles to which they are likely to transform. So are the sites on chromosomes most likely to break and recombine. Selection played no role in any of this amd neither did obligatory sexual (Mendelian) reproduction. Just as ontogeny proceeds driven entirely from within so did phylogeny take place (past tense) the same way. The molecular trees are meaningless and bear no resemblance to anatomically undeniable evolutionary relationships. All this is perfectly acceptable within the confines and potentialities offered by the Prescribed Evolutionary Hypothesis. All that allelic mutation is good for is to ensure ultimate extinction if something else doesn't cause it sooner. That is all that we see today. The only role for the environment, now as in the past, is to provide a stimulus. Everything else was and still is endogenous in origin.
"A past evolution is undeniable, a present evolution undemonstrable."
John A. Davison
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 26. June 2006 03:16
Peter,
I haven't had time to examine your 1G5 data, but your evidence for "non-random mutation" in the GULO gene is actually due to a very basic misunderstanding of phylogenetics. The sites at which humans, chimps, orangs and guinea pigs have one base but rats have another do NOT represent convergent evolution in the first four taxa - they simply represent mutations in the rat lineage. (Seriously, Peter, draw a tree, sit down and think about the data for five minutes - this is glaringly obvious.) So sites 01, 22, 31, 58, 79, 81, 109 and 157 are perfectly congruent with even the simplest model of common descent. Most of the other sites are examples of multiple mutations at the same site. This is slightly unusual, but note that many of these are located within a CG dinucleotide motif (commonly known as a CpG motif), which are well-known to have a high mutation rate due to DNA methylation. And despite the recurrent mutations at these sites, they can still be mapped unambiguously onto the standard phylogenetic tree.
As best I can tell, there is only a single site within the entire sequence that is truly incongruent with the standard phylogeny: site 55, in which the ancestral C base is changed to T in two separate lineages (the human and orangutan lineages). This incongruence can be explained in several ways: (1) independent mutation to T in the two separate lineages; (2) a change to T in the common ancestor of all primates, followed by a reversion mutation to C in the chimp lineage; or (3) a long-term polymorphism maintained across species boundaries, which was subsequently fixed as different alleles in the chimp lineage compared to the two other primate lineages.
So there is one sole remaining genuine incongruence in the GULO data-set, which can be easily explained by any of three plausible mechanisms under the standard evolutionary model. It seems to me that the evidence for anything at all unusual occurring within this stretch of DNA is looking pretty shaky.
Mesk.
[Added in edit - that last incongruent mutation is actually pretty interesting. In the common ancestor of the primate sequences a mutation occurred that changed the base next-ddor to it from an A to a G, resulting in a CpG motif in all primate lineages. This has then changed to a TpG in two separate primate lineages (human and orang). Given that CpG to TpG mutations are by far the most common mutation (by an order of magnitude) due to basic biochemistry, this makes the "two independent mutation" theory much more plausible - without having to rely on mutations of the "non-random" variety.]
[ 26. June 2006, 03:27: Message edited by: Mesk ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 04:20
quote:
The sites at which humans, chimps, orangs and guinea pigs have one base but rats have another do NOT represent convergent evolution in the first four taxa - they simply represent mutations in the rat lineage.
Show an outgroup to support your claim. The bats for instance, they are likely to have the GULO inactivated.
peebee
[ 26. June 2006, 05:14: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 04:28
quote:
The sites at which humans, chimps, orangs and guinea pigs have one base but rats have another do NOT represent convergent evolution in the first four taxa - they simply represent mutations in the rat lineage.
This is what Kreitman, editor of JME argued as well (Of course he rejected the paper, as did the other five or so Darwin-dominated journals). However, the Japanese authors do not agree on that. I obtained the manucript from them personally and they argue for hot spots in guinea pigs, which is in accord with my idea of convergent genetic changes in similar sequences (NRM). It is hard to digest for a Darwinian I presume, but I simply proof my claim.
A theory demises when it has to.
Goodbey NDE,
peebee
[ 26. June 2006, 07:09: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 04:38
ANYWAY,
due to my analysis, the GULO gene CANNOT be taken as evidence for common descent of the primates. All pro-evolutionary books, journals and internet sites must now retract the GULO sequences as proof of common descent (let's see whether they will do that). That's the scientific method.
What is more, three years ago I predicted on the EvC site that the GULO gene is subject to NRM, even before the publication of Ohta's paper. I predicted so, because GUToB says there is NO common descent. Besides, I already had observed positions 55 (C->T) as a NRM position then. [In actuallity, the prediction was also made here on the ISCID site. To be precise in the GUToB thread in April 2003. The Ohta and Nishikimi paper was published in October 2003.]
I also predict that additional inactivated pseudogenes show a similar pattern of alignment due to NRM. Darwinians will immediately claim it is due to common descent. I don't agree.
From GUToB I can make very risky predictions that come true...
GUToB rules,
Peebee
[ 26. June 2006, 06:38: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 04:46
quote:
The facts are most consistent with the following view. The microevoulution from the first fish to many subspecies of fish involves mostly random mutations.
How do you know?
Most likely it is reshuffling of preexisting genetic elements. And if you look into the karyotyping you will find (like we find it for the primates) that the shuffling is predetermined. Point mutations are irrelevant to (sub)speciation, this has also become evident from dog genetics (As discussed).
peebee
[ 26. June 2006, 06:26: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 05:04
quote:
So, if you are correct, your thesis would predict a vastly different phylogenetic tree of species from the present tree or no tree at all.
No, you are mistaken. A similar MPG means same sequences, same biochemistry. The result is a tree as we currently find it. With of course minor exceptions to this rule that proof my point and falsifiy common descent.
peebee
[ 26. June 2006, 05:04: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 05:20
Bruce,
Three more questions:
1 - Have you submitted this document for publication in a peer reviewed journal?
Yep, about 5 or 6 or so. All rejected.
2 - Do you mind if I spread your link around to a couple of other forums?
No
3 - Somehow I was of the impression that you are Austrailan, yet your paper says that you live in Germany. Was my impression totally bogus, or are you an Australian living in Germany? Just curious.
I used to live in Australia, but I am originally Dutch (not so proud of after yesterday's soccer game). Presently, I am living in Germany.
peebee
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 06:48
quote:
(Seriously, Peter, draw a tree, sit down and think about the data for five minutes - this is glaringly obvious.)
You mean so "glaringly obvious" that Ohta and Nikishimi missed it? These guys (in particular Nikishimi) are (is) the expert(s) in the evolution of the GULO genes.
peebee
[ 26. June 2006, 07:05: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 07:01
quote:
Added in edit - that last incongruent mutation is actually pretty interesting. In the common ancestor of the primate sequences a mutation occurred that changed the base next-ddor to it from an A to a G, resulting in a CpG motif in all primate lineages. This has then changed to a TpG in two separate primate lineages (human and orang). Given that CpG to TpG mutations are by far the most common mutation (by an order of magnitude) due to basic biochemistry, this makes the "two independent mutation" theory much more plausible - without having to rely on mutations of the "non-random" variety.]
This is ONLY one of the currently known non-random mechanisms. Because it is such obvious mechanism we know about it. The more obscure mechanisms still have to be elucidated. If I had not followed up on the GULO genes nobody would have ever found out about the NRM positions that give the illusion of common descent.
peebee
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 26. June 2006 08:32
Hi Peter,
Do you have any further outgroup sequences? These could be used to test these ideas further. However, you claim that even without an outgroup, you can detect discordancies within this data-set. This is patently false; most of the discordancies you highlight in the data-set are straightforwardly interpreted as mutations in the rat lineage, while the others are sites with nultiple mutations that nonetheless map unambiguously onto a simple tree. As I said, there is only one site within the entire data-set that doesn't fit with a straightforward common descent by mutation model, and this involves a CpG to TpG mutation that makes independent recurrent mutation a plausible scenario. There is nothing here providing strong evidence against common descent for the GULO gene.
quote:
This is what Kreitman, editor of JME argued as well (Of course he rejected the paper, as did the other five or so Darwin-dominated journals). However, the Japanese authors do not agree on that. I obtained the manucript from them personally and they argue for hot spots in guinea pigs, which is in accord with my idea of convergent genetic changes in similar sequences (NRM). It is hard to digest for a Darwinian I presume, but I simply proof my claim.
There are certainly sites within the data-set that have undergone multiple mutations within the data-set (sites 55, 76, 97, 131 and 156), and these could realistically be termed mutational hot-spots. However, such sites are by no means inconsistent with the standard evolutionary model. Note that only one is incongruent with the standard phylogeny (position 55) - the others map perfectly well to the tree, with no need to invoke independent identical mutations in separate lineages. In other words, they provide no evidence for non-random mutation.
quote:
You mean so "glaringly obvious" that Ohta and Nikishimi missed it? These guys (in particular Nikishimi) are (is) the expert(s) in the evolution of the GULO genes.
Show me a single piece of evidence indicating that Ohta and Nikishimi would disagree with the statement that "sites 1, 22, 31, 58, 79, 81, 109 and 157 indicate mutations in the rat lineage."
quote:
If I had not followed up on the GULO genes nobody would have ever found out about the NRM positions that give the illusion of common descent.
You have done no such thing. The GULO gene is interesting in that it contains five sites (out of 165) that have been subject to multiple mutations within a comparatively brief period of evolutionary time. However, evidence that certain sites have higher mutation rates than others does not, in any realistic sense, provide evidence against common descent. Instead, evidence against common descent would require finding a high frequency of sites that are incongruent with the standard phylogeny. There is one such site in this sequence (and I can tell you from personal experience that the frequency is generally even lower in other genes). This is not solid evidence against mainstream evolutionary theory, and the journals had perfectly sound reasons to reject your manuscripts as such.
Anyway, Peter, we've been over this before back at ARN, and I won't spend too much of my time refuting the same arguments over and over again. My main interest here is not to convince you (since that is clearly impossible), but prevent you from dazzling any readers with a non-scientific background into thinking that your ideas have any substantial merit. So let me open this up to anyone other than Peter - are there any readers out there who are not clear about how Peter's arguments here are mistaken? I'd be happy to clarify my response for you.
Mesk.
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 26. June 2006 09:24
quote:
This is patently false; most of the discordancies you highlight in the data-set are straightforwardly interpreted as mutations in the rat lineage,
MESK, I know your opinion. You are a staunch Darwinian and it is your job to defend it. I know all of your work on the ACTN genes and you know my opinion on that gene as well: NRM. In 2003 I have pointed out all my arguments why the ACTN3 gene is not in accord with current evolutionary thinking. I see no point in reiterating them here.
Otah and Nishikimi are clear in their paper: the GULO gene in guinea pigs has many, many hotspots for mutations. This is what they write in the abstract:
quote:
A comparison of the remaining human exon sequences with the corresponding sequences of the guinea pig nonfunctional GULO gene revealed that the same substitutions from rats to both species occurred at a large number of nucleotide positions (emphasis PB).
FROM: Inai Y, Ohta Y, Nishikimi M. Department of Biochemistry, Wakayama Medical University, 811-1 Kimiidera, Wakayama 641-0012, Japan.
Comparing the retrieved sequences they reported in this journals with their previous report [Biochim Biophys Acta. 1999 Oct 18;1472(1-2):408-11.] I discovered that these hotspot mutations coincide with the mutations that line up in primates and give the illusion of common descent. I did not coin them hotspots, the Japanese experts did. You can of course always send a letter to Ohta and Nishikimi to complain about their findings or interpretations.
Listen, MESK, I will do another prediction. When the entire GULO genes are compared in these organisms (and not only exon X as I did) we will still find that around 50% of the point mutations line up as the result of NRM. That would proof that I am right. What is your prediction from standard theory?
And don't call my analysis "patently false", again.
peebee
IP: Logged
|
|
|
Printer-friendly view of this topic