Macroevolution: Its all around us...

Well they would be relevant as a confounding factor to phylogenies based on genetic data, in the same way that morphological convergence can be a confounding factor in constructing phylogenies based on morphological data.At the moment there is no evidence that there is any convergence above the amounts expected, let alone a significant amount, in the majority of sequences used to construct phylogenies.Randman's initial contention was, as I understood it, that morphological convergence in aparticular trait might be tied to convergence of the genetic sequences underlying that trait. Which, if it were the case, might produce problematic data for phylogenetic analyses and undercut the evidence, in favour of common descent, from the congruence of morphological and genetic phylogenies.This relies on a lot of very unlikely factors, such as ignoring the fact that no-one is likely to start relying solely on the HOX genes or any other key developmental genes likely to affect morphology as their source of data for constructing phylogenies.But in theory it could produce confounding data.TTFN,WK

Excellent. Thank you. This is what was needed to put a dozen posts into some perspective.

Thinking about what has been written it seems to me that the effect would depend on:1) The location of the microsatellites - if the locations in 2 species were not very close they would affect different regions of DNA2) The proportion of mutations that occur within "rnage" of the microsatellites. If this is small then there will be little effect.Any comments ? I'd appreciate some sort of quantititive estimates of the relevant factors (even if I am wrong about what they are !)

From the Vowles and Amos paper So they identified 5537 microsattelites uniquely homologues to human microsatellite regions, and these were initially identified by homology in a region covering a considerably larger region than that of the microsatellite or its region of 'influence'.I don't know to what extent this corresponds to chromosomal synteny. This depends rather upon the model of mutation you are looking at. Certainly it is highly unlikely that a protein coding region is going to be within the range of influence of a microsatellite, except perhaps a small (N)2 dinucleotide repeat, but it is not totally impossible that a regulatory element could be near to a substantial microsatellite.One of the aims of this paper was to investigate whether the rate of mutation was higher or lower than expected in sequences flanking the microsatellites, what was found was that both were the case, but at different positions relative to the microsatellite.Part of the paper estimates that around 30% of the genome will be under the range of effect of an 'ACAC' repeat. Taking into account other lengths of repeat and dinucleotide sequences the number is likely to be much haigher than that. It is worth bearing in mind however how small a proportion of the genome is actually protein coding.The best way to get real answers to your questions aboutthe number and nature of mutations would be to read the paper.TTFN,WK

So can we safely assume that phylogenies constructed from highly conserved and/or transcribed regions will be largely unaffected by this potential source of error?

Quite true -- particularly if you are willing to apply the term "theory" to a highly conjectural game of smoke and mirrors...

Thsi quote is from the paper My impressions are as follows:The biggest effect appears to be an overall trend towards conservation of sequences - so differences would tend to be preserved rather than being overwhelmed by "convergent" changes.Given that the microsatellite placement (and the sequence of the microsatellite itself) would have to be similar for genuine convergence. I don't see much room for significant convergence between organisms that were not already very similar genetically. Species might tend to stay "in step" after an evolutionary branchpoint but that should just cause the branchpoint to be placed a little later than it actually occurred.(Added in edit)
And since it seems to be non-coding DNA any effects will be limited to recent divergences because non-coding DNA is not sufficiently conserved to be used in the study of anything elseThis message has been edited by PaulK, 06-23-2005 02:36 PM

You mean ToE?

Randman-First off, stop being so reactionary, and please take the time to read my posts. You state: I NEVER denied convergent evolution of DNA. I support it, at least for the types of DNA sequence that there is evidence for convergent evolution. I simply feel that your extension of the specific evidence you have to theories of common ancestry and evolution in general is overreaching. It seems that rather than deal with the actual criticism I made you'd accuse me of denying evidence, evidence that I supported and even thanked you for bringing to the discussion. Why? What is the specific prediction that "special creation theory" makes that suggests that the results of convergent evolution would produce closer sequence similarities between chimp and human than between human and mouse?You can't just say "the data fits special creation perfectly". You have to explain how and why the data fits special creation, and how special creation predicted that data trend. Is it really "an area of intense study"? How many labs are working on the problem? How many papers have been published on the subject?You've provided a great piece of evidence NOT that non-coding DNA exhibits convergent tendencies, but that non-coding DNA flanking specific types of microsatellite repeats exhibit convergent tendencies. Do you understand the difference? I never suggested you argued a sole explanation. Again, perhaps you should take more time reading posts before responding to them. In fact, I acknowledged that you listed several commonalities. Are you interested in having a decent discussion, or just pointing accusatory fingers? No, my line of reasoning rests on the fact that different types of DNA sequences evolved differently. You can't simply take evidence from one class of DNA sequence and immediately extend it DNA in general - that is where I believe the root of your "overreaching" lies.And there IS much else to say, since you by saying I am incorrect you are asserting that there is, indeed, evidence for convergent evolution of coding or pseudo-coding genes.Please provide that evidence (hopefully you realize that the reference you've already given does not give such evidence).

I don't want to be overly reactionary, and sometimes other posters responses affect my tone towards those not being uncivil. Sorry. I assumed with the above-comment that you were denying convergent DNA was real. I don't know how many labs are working on it. WK here made the comment that it was a "hot topic" and a cursury review seemed to verify that.Do you think that is incorrect? I guess I disagree here. One of the bedrock principles of common descent is that all DNA is essentially the same thing, but with different compositions, etc,..but exhibiting such similarity that it is used as an argument for universal common descent. I don't really deny that it can be used for evidence of common descent.Along with that, if we discover that DNA has convergent properties within it, then to me it is reasonable to think that holds true across the board. So if human DNA exhibits a predisposition to mutate according to a certain pattern, it is likely that all DNA has within it, an embedded predisposition to mutate according to certain patters.Imp, it is too early to know the extent of this phenomena, and the mechanisms involved, but considering the fundamental role DNA plays in reproduction and any presumed evolutionary processes, the fact DNA displays any convergency at all and is not just random is quite revealing, and it's effects should not be discounted until further understood.One way to look at the patterns certain DNA is predisposed towards is to see those patterns as an embedded design.This message has been edited by randman, 06-23-2005 01:48 PM

Hi WK-Doesn't "highly unlikely" seem a bit strong, especially since the paper is examining such short repeats? As you mention, (AC)2 repeats produce significant influence (and the influence seems to increase greatly at (AC)5) - is there any reason they wouldn't be in the coding sequence of the gene? There are plenty of genes with short microsatellites within their coding sequences, even more with microsatellites in introns, no?Here is an interesting paper on the subject (more or less) from a few months ago (though it did not focus on AC micosatellites, and actually analyzed sequence at the amino-acid level): What are your thoughts?

Hardly. The problem you are citing when attempting to apply this across species is similar in many respects to that of denying that one can arrive at phylogeny on the basis of endogenous retroviruses because when two species have endogenous retroviruses in exactly the same position, they could have very well been infected separately, and by some incredible coincidence, the point of insertion may have been the very same in both cases. Granted, there will be hotspots which reduce the the number of probable insertion points, but there are still a great many of insertion points, and trying to argue in the fashion belongs under the heading of sophistry, not science.As further evidence, I cite: The statement you quoted was with regard to coding and pseudo-coding sequences, not the satellite regions where the DNA is highly repetitive -- and which was the subject of the paper which you refered to. I think it is going to be rather difficult and quite possibly unproductive to argue with someone who has such a habit of either misunderstanding or misrepresenting what someone else writes. The individual who you were responding to had done a masterful job of responding to you, and the only way that you could respond back was by twisting what he had said at numerous points. A pattern such as this suggests intent, not accident.This message has been edited by TimChase, 06-23-2005 02:05 PM

Wrong. It is fallacious to compare the reality of convergent DNA with endogenous retroviruses, as you have done.The simple fact is that if DNA contains within it, a certain predisposition to mutate towards a certain pattern, depending upon the existing pattern, and not to just mutate randomly, then that is an embedded characteristic of DNA. So there is a cause creating an effect.You seem to be arguing that the effect is via chance, and thus make a fallacious argument. There is no appeal to chance or coincidence in any of my posts here at all. You are misrepresenting the argument. Maybe you don't understand it?I'll just repeat it and maybe you will get it this time.There is embedded within certain patterns of DNA a predisposition to mutate according to a pattern and not just randomly.Right now, we do not know the extent of this phenomena, but may know more in a few years.But it is a reasonable assumption that if we see this characteristic in human DNA that it is likely true across the board for all species.Why is that so hard for you to grasp?As far as a statement, I provided a link to an entire scientific paper. I suggest you read it. WK originally provided the link, and it could be useful if you would take the time to learn a little about the subject, especially since you seem to be qualified to understand the terms and concepts.

Thanks for the apology - I do appreciate it (one of pet peeves is when people claim I made arguments that I did not). I guess I'm not sure precisely what you mean here. At the non-sequence level DNA is essentially the same thing. BUT, we are not discussing non-sequence level behavior of DNA here, we are discussing the influence of specific types of sequence, and thus the sequence level behavior of DNA. Thus we cannot simply say "all DNA is the same, therefore it must behave the same way", because it is not all the same.Do you see the difference?Just because AC repeats appear to induce or stabilize convergent evolution in nearby DNA in no way implies that the sequence of the Cytochrome C gene, for example, has undergone convergent evolution.Another way of putting it; repetitive stretches of DNA sequence have a higher rate of mutation than non-repetitive DNA sequence. Thus even though at the non-sequence level it is "all the same DNA", the sequence does indeed influence the biochemistry of the DNA.Hopefully that all makes sense...If it makes sense maybe you'll see why I think your statement- -is a simplification that leads to that leads to overreaching the evidence. That would be the case if any pattern was indicative of design, which I do not feel is the case. Just because certain DNA sequences form a pattern because they are mutable or stabilized at a molecular/biochemical level does not indicate design to me, especially since those sequence have no apparent function outside of themselves.In other words, I see chemistry as the result of the pattern where you see design - when minerals stabilize themselves into patterned crystals, I see chemistry at work, not design. What about you?

So you are willing to admit that retroviruses are a kind of smoking gun for evolution?Constructing primate phylogenies from ancient retrovirus sequences
Welkin E. Johnson and John M. Coffin
Proc. Natl. Acad. Sci.
Vol. 96, pp. 10254-10260, August 1999
Just a moment...Phylogenetic relationships among cetartiodactyls based on insertions of short and long interersed elements: Hippopotamuses are the closest extant relatives of whales
Masato Nikaido, Alejandro P. Rooney, and Norihiro Okada
Proc. Natl. Acad. Sci.
Vol. 96 pp. 10261-10266, August 1999
Just a moment...