"The Edge of Evolution" by Michael Behe

Indeed, the thread was [thread=-13798]. Unfortunately the original Nature paper and commentary are only available with a subscription.TTFN,WK

This bears no relation to how evolutionary theory has actually progressed. Things like drift and neutral theory were not considered sources of new genes, but they do explain genetic features which aren't explained by adaptive evolution alone. The fact that these changes have subsequently been found to have an effect on adaptive evolution is incidental. No, but it is hard to believe that you know what you are talking about. A genetic background is simply a common technical term for the genomic state to which a mutation is introduced.As to joining up the dots, isn't that what the creationists are always begging us to do? Thornton's work is one of the most joined up examples of molecular evolution we currently have.TTFN,WK

Yes, but they are mostly obvious things like synonymous mutations and functionally neutral SNPS, two things with a high level of overlap, but not neccessarily exactly the same. You might also consider some elements of what have been considered 'junk' DNA such as degenerate non-expressed pseudogenes. Um, the reference would be exactly Joe Thornton's research we are just discussing. The whole point of Thornton's research is that the adaptive evolution of a genome is carried out against the shifting background of the rest of the genome and much of this shifting is effectively neutral in terms of adaptation/fitness. That is why it is harder to construct a path backwards from a more derived protein to the ancestral protein, because you don't know what the permissive backgrounds for the differing steps were. Junk DNA along with everything else in the genome does constitute the genetic background in vivo, in vitro the background was in this case the rest of the protein coding region which remained unchanged and the genome of the Chinese Hamster Ovary cell in which the constructs were tested. One can easily argue that this is clearly demonstrated in almost any selection intensice experiment. If we grow up a clonal colony of bateria to identify mutations conferring resistance to some anitbacterial agent then all we are doing is trying to generate a wide panel of effectively neutral mutations which will become beneficial on introduction to the selective environment.TTFN,WK

I don't see how this is shooting himself in the foot. Unless your contention is that he failed to take into account the magical intervention of the intelligent designer during the experiment. The mutations clearly occured, they need not be simultaneous but the required maintenance of the earliest one in the absence of positive selection does make their joint ocurrence less likely, but obviously not impossible.Indeed Lenski's research clearly show that these mutations were not simultaneous since one of the mutations had already occurred by ~20,000 generations but the Cit+ trait didn't evolve for a further 11,000 generations. It also isn't clear that the potentiating mutation was selectively neutral, simply that it doesn't provide any ability to metabolise citrate (Blount et al., 2008).TTFN,WK

What do you mean here? That there are 20,000 protein coding genes in humans? That there are only 20,000 possible functional coding sequences? Once again simply picking a number out of the air with no real basis leads you to make ludicrous statements. Simply understanding the concept of third base wobble shows you should be cutting your improbablilty factor by about a third since in many cases changes at the third base do not affect the amino acid sequence produced.The number you need to be able to divide by is the number of all possible functional coding sequences for the function in question, if you want to calculate the chance of a specific function arising. But even then you are making the same mistake that has been pointed out numerous times by assuming that that particular function was a goal that had to be reached rather than just the particular endpoint that happened to be reached.TTFN,WK

I would, but I still don't know what you mean and it doesn't sound like you do either. Can you tell us where you got this estimate from? I can find sources quoting 20,000 as the number of distinct enzymes in a human body or cell, but thats all. That certainly doesn't fit with your idea that there are only 20,000 possible functional enzymes sequences since almost all other species have numerous functional enzymes with multiple structural differences from those in humans. But you are only taking all other enzymes in the human body into account, if that is the origin of your number, this won't include all the other possible sequences that perform the exact same function.If you can't tell us where this number comes from how can we tell you exactly how wrong you are?TTFN,WK

"The figures are correct", seriously? You just totally changed what the figure was supposed to represent but your argument is still supposed to make sense? The point is that, as has now been pointed out ad nauseam, the specific sequence coding for an enzyme is not the only possible sequence that codes for a functioning form of that enzyme. All you need to do is look up a few enzyme encoding genes in entrez and you can find numerous sequence variations in humans, in the form of single nucleotide polymorphisms (SNPs). So even these estimated 20,000 human enzymes have a wider range of functional variation in the human population than you are willing to concede, and we have little if any idea what the limits of that functional variation are.So you are wrong because the numbers you are using don't actually seem to refer to anything in reality, and also in that your approach relies on the need for specific sequences.TTFN,WK

This is just more empty babbling showing how little you know about biology. The mutation behind sickle cell anaemia is a single nucleotide substitution. Do you really think that sickle cell anaemia isn't seen by natural selection? There are many other examples, if you think single nucleotide changes cannot form the basis for significant natural selection then you clearly know nothing about evolutionary biology, molecular biology or developmental genetics.TTFN,WK

I think you are going over the top here. In the vast majority of cases such changes probably won't remain in the gene pool as low frequency alleles which aren't subject to selection are prone to be lost due to drift. That certainly doesn't mean that they can't remain in the gene pool, but it is far from being a sure thing.I find your hypothetical 19 neutral potentiating steps scenario almost totally ridiculous. Can you give us any evidence of such a thing occuring? If you look at Lenski's long term evolution experiment the evolution of citrate metabolism took more than 30,000 generations and that seems to have only involved 2 mutations, neither one of which may be neutral (Blount et al., 2008). Neutral changes can rise to fixation through hitchhiking, but for 19 to do so before a beneficial mutation can arise seems to be stretching things a bit. I can quite easily believe that in a dynamic evolving genome there will be proteins whose evolutionary trajectory has been substantially shaped by chains of interaction with, at the time, selectively neutral mutations with a potentiating effect, which is what Thornton's research shows, and that this could encompass dozens of selectively neutral changes. I find the idea of 19 seperate neutral mutations arising and being maintained at sufficiently high levels to allow 1 beneficial mutation to arise to be doubtful.The 2nd dice model seems to be exactly the sort of locking scenario that everyone wrongly accuses Dawkin's weasel program of being. You seem to ignore the fact that neutral mutations have no reason to be considered locking, neither do beneficial mutations in most cases but it is a more reasonable assumption. The Weasel still seems a better demonstration.TTFN,WK

Well clearly they do alter it by any sane standard. The extent to which they do and whether it constitutes a gain or loss of information would depend on how you are measuring information, something you still seem loathe to tell us.You seem to be saying that neutral snps won't be seen by selection, whcih is true but has no actual bearing on the discussion. A claim with no evidence to support it, compared to the masses of evidence for SNPs with observable phenotypes, including beneficial ones.TTFN,WK

It seems more likely to bring about an avalanche of people pointing out that this is a totally useless definition in terms of actually measuring information.TTFN,WK

No he didn't, please provide a quote to support this blatant misrepresentation.What Kimura believed was that the vast majority of evolution, not adaptive evolution, was neutral. In other words most single nucleotide polymorphism changes that represent the majority of variation in the genome do not produce functionally significant effects (Kimura, 1989). Similarly he may extend this to the grosser phenotype and suggest that a number of gross morphological variations may be essentially neutral. Such neutral variation is, according to Kimura, kept in the population by a dynamic balance of continuous mutation and extinction of polymorphic alleles and may become fixed by genetic drift. Kimura's 'survival of the luckiest' describes the mode, in a statistial sense, of evolution not the whole of it in the same way that Gould points out that the mode of life on earth is bacterial. Kimura does not deny the existence of adaptive evolution, he just points out that most genetic variation is not governed by it. Given your previous record I tend to doubt it, but feel free to show us precisely where he wrote this.TTFN,WKEdited by Wounded King, : No reason given.Edited by Wounded King, : No reason given.

Now I'm wondering if I was right. I have just started a new thread about genes generated de novo in the human lineage, New genes in the Human lineage, and I'm wondering if these would be a counter example. Surely the pre-existing sequence could be considered to consist of the product of multiple potentiating mutations? But the real question is whether these de novo coding genes are in fact beneficial. Perhaps rather they are themselves selectively neutral.Certainly we would have to wonder if such a de novo mutation gave rise to something like a highly sequence specific transcription factor. Does the putative pre-existing sequence for the NylB, Nylonase, gene in Flavobacterium consist of multiple neutral potentiating mutations? The maintenance in the antisense strand of the pre-existing sequence of so called 'Nonstop frame' suggests that perhaps these sequences are not selectively neutral. What could drive this maintenance is unclear though it has been suggested that it could be a strategy for imcreased evolvabilty specifically because it can give rise to such de novo mutations (Yomo et al., 1992).TTFN,WKEdited by Wounded King, : No reason given.Edited by Wounded King, : Updated thread link