Is convergent evolution evidence against common descent?

If you are claiming that birds don't have digits, which is suggested by saying the wing is constructed from exclusively etc..., then you are completely wrong.The following figure shows the digit morphology in a wild type chick wing and a reduplicted pattern in a wing with a transplanted ZPA. Maybe that wasn't what you were suggesting, in which case your figure was just rather misleading.TTFN,WK

The real hot topic is the extent to which we see convergent evolution in DNA sequences. Vowles and Amos (2004)document convergent evolution in sequences close to microsatellite regions.This is partially atributable to mutation biase such as those for C->T transversion but there seem to be a number of other effects yet to beaccounted for.Perhaps the most fruitful are of study would be studying the underlying sequences of DNA key to the development of analogous/ convergent structures for signs of convergence.TTFN,WK

Mostly, yes. In this particular case the 'traits' as such simply are particular sequences so to distinguish between traits and sequences is incorrect. There certainly are a number of rather generalised assumptions about DNA which have been shown to be overgeneralised. Thinking that all mutations are entirely random certinly is wrong. There are specific mutations which are favoured due to the chemical structure of DNA, even at the level of individual base pairs such as the C->T transversion I mentioned previously. As this research shows there are also effects of the local environment, in terms of the composition of nearby sequences of DNA, which can affect the character of mutations in that region. Predetermined would be an exaggeration, predisposed would be a better way to phrase it. In the same way that a certain environment or genetic characteristic may predispose someone to develop cancer but does not predetermine that they certainly will.TTFN,WK

It depends how far you want to characterise a trait. I can't see any way in which a specific protein coding sequence could be predisposed to develop almost de novo but within an alrady established gene there may well be hotspots which frequently mutate leading to specific traits, such as the many human developmental genetic syndromes which have been well characterised. Some of these developmental syndromes may be said to be physical forms which are predisposed to develop due either to chromosomal or sequence level mutational hotspots.So within a given sequence there may be a predisposition towards a certain mutation but this does not neccessarily correlate with a phenotypic trait, although it may in some cases.The microsatellite and repeat sequences studied ar very different in character to protein coding and gene regulatory sequences of DNA. These are perhaps the second simplest form of DNA one could find, i.e. repeats of a 2bp (ACACAC) sequence as opposed to a stretch all of one nucletiode type (AAAAA), and the amount of effect they might have on anything relating to phenotype is probably minimal. That is correct, although more general environmental factors may also be involved. Particular metabolic processes within the cell or particular proteins produced in the cells might also have predisposing effects on certain mutations.TTFN,WK

In this case I would think that both structure and function should be seen as 'surface' elements since they are so inextricably linked. For a foundational element one would need to look at the DNA coding for the particular protein as there are a number of viable sequences capable of coding for a specific primary sequence of amino acids or of a particular class of amino acids allowing a specific structure, such as a helix-loop-helix, to form.TTFN,WK

This is overstating things. DNA does mutate randomly, but not completely randomly. It also, in certain circumstances, tends to produce highly specific very simple repetitive nucleotide sequences.TTFN,WK

You really need some evidence to support this. It is a big leap from general predispositions to certain mutations in highly repetitive code to the same mutation occurring in 2 highly independent populations and then coming to predominate even while being selected against. It isn't theroretically impossible, and never has been, but it is still only marginally less unlikely taking into account non-random elements in mutation and therefore much less parsimonious than assuming common ancestry.It has never been physically impossible for totally convergent traits at the genetic level to arise independently, but the chances against it are far too high for it to be a preferable explanation to that of common ancestry.TTFN,WK

There is no reason to suppose that all of these memebers are going to have the trait in the same generation at least not without the intervention of a further entirely unknown mechanism. A predisposition is just that, mutations will occur, some will tend to occur more frequently, there is no reason to expect a sudden blip of enough members with a particular trait to appear to allow them to successfully spread that trait throughout the population. It is more likely than a trait which is predisposed against perhaps but to suggest that the likelihood is quite high seems unwarranted in the absence of any sort of analysis, other than that of you giving your opinion. In fact this is the situation where it is least likely that your novel trait will be fixed, it would be much more likely to spread through the whole population if the population is small and more subject to the effects of genetic drift. These sort of effects are pretty well documented, and many are linked to the founder effect I mentioned before. Many Ashkenazi jews have a predisposition to breast cancer for instance due to mutations in the BRCA1 and BRCA2 genes.To a large extent this is simply connected to inbreeding but there are also cases where a negative trait which is either linked systematically or geographically, i.e. close proximity of the genes on 1 chromosome, can 'hitchhike' to fixation, in the same way as neutral polymorphisms in terms of locus, provided that the positive selection for the first trait is strong enough. The problem is that these are considerably less plausible than straightforward morphological convergency and certainly less plausible than homology. As I suggested before these scenarios are by no means impossible, there is just a paucity of evidence to suggest that they are in any way preferable to the current thinking in terms of common descent. Perhaps such mechanisms will be discovered, but in their absence you are just constructing hypotheses out of very slight foundations.TTFN,WK

In this case you may need to put in a bit of legwork. There are at least three studies looking at DNA based phylogenies of Thylacine wolves. Unfortunately they don't seem to be available online except in abstract, or if they are my university doesn't have the neccessary access.The papers abstracts are linked from here.I'll go and see if I can find one of the papers at the library this lunchtime.Alternatively we could do our own little bioinformatics project using the Thylacine wolf DNA sequences on genbank, here.Or we could look at the protein sequences, here. If you lick on the Blink hyperlink it will give you listings of all the results of a Blast search for that protein. Looking at Cytochrome b you will see that the closest hits are all marsupials, even better if you click on 'Taxonomy report' which will show you the taconomic classifications of the various species along with their scores for similarity. you could also collate some Cytochrome b sequences yourself and use them to make a phylogenetic tree using a program like CLustalX.TTFN,WKThis message has been edited by Wounded King, 06-07-2005 05:57 AM

Thanks Para. I've corrected that now, some square brackets in the address had messed things up.TTFN,WK

That tree is based upon the nucleotide sequence of Cytochrome c. The original data under discussion was in terms of amino acids in the beta chain of hemoglobin.So while this supports the argument that different genes may give rise to very different phylogenies it does not support the claim that the hemoglobin data suggests that mice are more closely related to kangaroos.Let us try another little bio-informatics exercise. Here is the genbank protein entry for the Kangaroo hemoglobin beta chain. Then find similar sequences for Human, mouse and a monkey species.To save time I have found those sequences and put them into FASTA format. (edited by adminnosy to insert blanks to stop wide page effect) Cut those out and paste them into the ClustalW form at http://align.genome.jp/ , now make sure to remove any gaps at the ends of lines so that the amino acid sequences are all on one line. Now click on the 'Execute Multiple Alignment' button.You now have an alignment of the amno acids, as well as being able to put this data into a tree you can see the scores for the various pairs. Human against Kangaroo gives 73, Human against Mouse gives 80 and Mouse against Kangaroo gives 69. So clearly the mouse is in fact more divergent from the kangaroo than Humans are, and humans and mice are clearly more closely related than Kangaroos are to either, at least based on this data.TTFN,WKThis message has been edited by Wounded King, 06-07-2005 09:08 AMThis message has been edited by AdminNosy, 06-07-2005 01:36 PM

Is that the right link? You have referenced an earlier post of Randman's.TTFN,WK

Well I put the Mouse, Kanagroo and human in because those are the species for which the site gave data which you claimed showed that the mouse was more closely related to the kangaroo.I can put a Dunnart sequence in however, in this case the fat-tailed dunnart. If you add that to the previous sequences for alignment then you will see that the Dunnart trees out with the Kangaroo. Dunnart and mouse score 68.7, Dunnart and kangaroo - 80.1 with Mouse and Kangaroo scoring 69.2.So in fact the Mouse amino acid sequence is more similar to that of the Kangaroo than the Dunnart.TTFN,WK

Maybe you aren't familiar with how science works. Scientific research is published in journals. Only a small subset of those journals are available online and of those an even smaller subset are available to those without institutional access from a university or similar body.Not all of the relevant research is going to be freely available to you or to those debating with you. If you can't even be bothered to try and uderstand why looking for yourself at the primary literature is important then this whole exercise is pretty futile. That would certainly be the acid test, but it would take a pretty sophisticated major bit of developmetal genetics to do that. At the moment however all of the genetic evidence points to common descent and away from convergence at the genetic level. In the literature, in Genbank. Did you do your homework and run those FASTA sequences through CLustal yourself? I don't want you just to be taking my word for this stuff.TTFN,WK

Who taught you 'evolutionism'? Because none of my biology lecturers ever tried to indoctrinate me.TTFN,WK