Freethought & Rationalism Archive

Albion · 09-08-2004, 05:29 PM

That Meyer paper, including the Nature article, is being discussed in this E/C thread.

Wounded King · 09-09-2004, 03:20 AM

As has been suggested a model like this is only as good as its assumptions, and I'm not convinced that Behe and Snoke's assumptions are all that good.

Their stipulation of neutral selection is allowable given their assumption that some early stages in the development of a multi-residue sequence may not be selected for functionally, but the quote they use to illustrate this

Quote:

the early steps might have been selectively neutral because the new
function might not be manifested until a certain number of steps had already occurred.

Interestingly by the time they reach the conclusion this has metamorphosed into

Quote:

As Li (1997) points out,the evolutionary origins of such features must have involved multiple mutations that were initially neutral with respect to the MR feature.

The emphasis on the word must, instead of the word might, is mine.

has enough mights in it to make an assumption that this is neccessarily the case in general dubious, but in the specific situation they posit for the model it is allowable assuming one therefore realises the limited applicability of the model.

Another assumption is the rather loosely defined one that most non-synonymous mutations produce a null phenotype.

Quote:

most mutations that alter the amino acid sequence of a pro-
tein effectively eliminate function (Reidhaar-Olson and
Sauer 1988,1990;Bowie and Sauer 1989;Lim and Sauer
1989;Bowie et al.1990;Reidhaar-Olson and Sauer 1990;
Rennell et al.1991;Axe et al.1996;Huang et al.1996;
Sauer et al.1996;Suckow et al.1996).

Many of these papers, including the 'Axe, 1996' paper which Behe subsequently discusses, actually suggest a great degree of tolerance in the amino acid sequence of a variety of structural features. Axe shows that 23% of mutated core amino acid sequences were still functional even if every one of 13 core amino acids was substituted. Admittedly all that really seems required for the core structure is that a reisdue be hydrophobic so it might be a special case.

Quote:

As an example,consider a gene of a thousand nucleotides.If a
total of 2400 point mutations of those positions would yield a null allele,whereas three positions must be changed to build a new MR feature such as a disulfide bond,then would be 2400/3, or 800.

Where do these numbers come from? Why 3 mutations to produce a disulfide bond, surely all you really require is two cysteine residues within a certain proximity to each other? I'm less than convinved that the exact location (Behe notes the specific nucleotide positions) is always a neccessary prerequisite for disulfide bond formation either, an amino acid one way or the other is unlikely to make that much difference.

Where do the 2400 null mutations come from, what are they rendering non-functional? The parental function of the duplicate gene? The future function of the novel gene? Is it a total loss of function or just a loss of complete functionality? If it is a partial loss of the parental function then it is by no means appropriate as a factor and will not neccessarily lead to pseudogene formation. Loss of a binding site, for example, important to the function of the parental protein need not deny the possibility of a novel function to the novel protein, i.e. acting only on a subset of the initial targets of the parent protein for instance. It may be that in this case a null mutation is simply one leading to the formation of a pseudogene and therefore removing the gene from any maintenance by selection, in which case it is reasonable to assume the gene would most likely degrade subsequently and not evolve novel functionality.

I don't deny that the odds are in favour of null rather than novel functional mutations, but it would be nice if Behe made some effort to explain where he gets the numbers from exactly.

As an afterthought, the final conclusions of the paper seem not only highly tentative but pretty non-restrictive. If anything they seem to suggest that Li's 'hopeful monster' like scenario is possibly something of a strawman in terms of the evolution of multi residue feature, especially when there are mechanisms such as domain swapping for slotting in already evolved complex structural features.

TTFN,

WK

budgie · 09-09-2004, 11:43 AM

RBH, your clarification of copyright and fair usage is edifying. I will have to use the interlibrary loan system here to grab a copy from my alma mater - for personal use only.

Wounded King :notworthy :notworthy :notworthy

-jim

G-natrix · 09-25-2004, 06:31 PM

The PDF for Behe and Snoke 2004 is available at

http://www.proteinscience.org/cgi/co.../ps.04802904v1

Caveat: �*I was able to download this by clicking on the Full Text (PDF) link on the journal's home page, which I got to (ironically) through one of the ID sites (Discovery I think). �*I typically go through the library to get subscription PDF's but I wasn't in this instance so I assume anybody can go there and get the PDF. �*There may be a slim chance that you have to access it through the ID site

http://www.discovery.org/scripts/vie...%20-%20Science

Enjoy.

RufusAtticus · 10-11-2004, 10:59 AM

Bumping because our essay critical of Behe and Snoke (2004) is now out.

Theory is as Theory Does.

Nic Tamzek · 10-11-2004, 05:36 PM

A little more for people:

Theory is as Theory Does
by Ian F. Musgrave, Steve Reuland, and Reed A. Cartwright
http://www.pandasthumb.org/pt-archives/000480.html

[A critical review of this paper:

Michael J. Behe and David W. Snoke (2004). "Simulating evolution by gene duplication of protein features that require multiple amino acid residues." Protein Science, Published online before print August 31, 2004. DOI: http://dx.doi.org/10.1110/ps.04802904
http://www.proteinscience.org/cgi/co.../ps.04802904v1 ]

Musgrave et al.'s conclusion:

Quote:

Conclusion
We began this essay with a quotation from Behe complaining that a paper describing an evolutionary simulation (Lenski et al. 2003) had “precious little real biology�? in it. What we see here is that Behe and Snoke’s paper is acutely vulnerable to the same criticism. A theoretical model is useful to the extent that it accurately represents or appropriately idealizes the processes that occur in the phenomenon being studied. Although it is worthwhile to investigate the importance of neutral drift, Behe and Snoke have in our opinion over-simplified the process, resulting in questionable conclusions.

Their assumptions bias their results towards more pessimistic numbers. The worst assumption is that only one target sequence can be hit to produce a new function. This is probably false under all circumstances. The notion that a newly arisen duplicate will remain selectively neutral until the modern function is firmly in place is also probably false as a general rule. Their assumption that 70% of all amino acid substitutions will destroy a protein’s function is much too high. And finally, we have shown that their flagship example does not require a large multi-residue change before being selectable.

And ironically, despite these faulty assumptions, Behe and Snoke show that the probability of small multi-residue features evolving is extremely high, given the types of organisms that Behe and Snoke’s model applies to. When we use more realistic assumptions, though many bad ones still remain, we find that the evolution of multi-residue features is quite likely, even when there are smaller populations and larger changes involved. In fact, the times required are within the estimated divergence times gleaned from the fossil record. We can therefore say, with confidence, that the evolution of novel genes via multi-residue changes is not problematic for evolutionary theory as currently understood.

Wounded King · 09-01-2005, 04:22 AM

Reviving a blast from the past, the other "ID" paper which came out around the time of Meyer's helpful contribution to systematics and taxonomy.

A paper has been published in 'Protein Science' which both critiques the Behe and Snoke's paper (2004), and discusses an alternative model based upon theirs in part but with some assumptions more in line with the current literature (Lynch, 2005) and which suggests that neo-functionalisation is considerably more likely than in the scenario proposed by Behe and Snoke. There is also a response by Behe and Snokes and an editorial.

TTFN,

WK

Sven · 09-01-2005, 04:45 AM

Quote:

Originally Posted by Wounded King

A paper has been published in 'Protein Science' which both critiques the Behe and Snoke's paper (2004), and discusses an alternative model based upon theirs in part but with some assumptions more in line with the current literature (Lynch, 2005) and which suggests that neo-functionalisation is considerably more likely than in the scenario proposed by Behe and Snoke. There is also a response by Behe and Snokes and an editorial.

From the response:

Experimental studies contradict Lynch’s assumption of complete neutrality as a rule; the majority of amino acid substitutions decrease protein function.

This would be news to me. Funnily, they do not provide a reference for this claim, although they do so elsewhere.

ETA: I think this incident clearly shows the claim of an a priori bias against ID "research" to be a lie.

Steven Carr · 09-01-2005, 04:46 AM

Out of idel curiosity, how does the paper show any sort of evidence of any kind for any signs of action by any type of intelligent designer?

Wounded King · 09-01-2005, 05:01 AM

It doesn't. That didn't stop the DI claiming that it did though. Beha and Snoke recognise this problem themselves.

Quote:

We subscribe neither to triumphant views in some circles that our paper disproved Darwinism, nor to overwrought ones that it supports some grand anti-science conspiracy.

TTFN,

WK

09-09-2004, 11:43 AM	#23
budgie Senior Member Join Date: Oct 2003 Location: I Owe the World an Apology Posts: 890	Thanks to . . . RBH, your clarification of copyright and fair usage is edifying. I will have to use the interlibrary loan system here to grab a copy from my alma mater - for personal use only. Wounded King :notworthy :notworthy :notworthy -jim

09-25-2004, 06:31 PM	#24
G-natrix New Member Join Date: Sep 2004 Location: Corvallis, Oregon Posts: 1	Behe PDF available online The PDF for Behe and Snoke 2004 is available at http://www.proteinscience.org/cgi/co.../ps.04802904v1 Caveat: �I was able to download this by clicking on the Full Text (PDF) link on the journal's home page, which I got to (ironically) through one of the ID sites (Discovery I think). �I typically go through the library to get subscription PDF's but I wasn't in this instance so I assume anybody can go there and get the PDF. �*There may be a slim chance that you have to access it through the ID site http://www.discovery.org/scripts/vie...%20-%20Science Enjoy.

Thread Tools	Search this Thread
Show Printable Version	Search this Thread: Advanced Search

09-08-2004, 05:29 PM	#21
Albion Veteran Member Join Date: Jul 2002 Location: US east coast. And www.theroyalforums.com Posts: 2,829	That Meyer paper, including the Nature article, is being discussed in this E/C thread.

10-11-2004, 10:59 AM	#25
RufusAtticus Veteran Member Join Date: Nov 2001 Location: NCSU Posts: 5,853	Bumping because our essay critical of Behe and Snoke (2004) is now out. Theory is as Theory Does.

09-01-2005, 04:22 AM	#27
Wounded King Veteran Member Join Date: Mar 2003 Location: Edinburgh Posts: 1,211	Reviving a blast from the past, the other "ID" paper which came out around the time of Meyer's helpful contribution to systematics and taxonomy. A paper has been published in 'Protein Science' which both critiques the Behe and Snoke's paper (2004), and discusses an alternative model based upon theirs in part but with some assumptions more in line with the current literature (Lynch, 2005) and which suggests that neo-functionalisation is considerably more likely than in the scenario proposed by Behe and Snoke. There is also a response by Behe and Snokes and an editorial. TTFN, WK

09-01-2005, 04:46 AM	#29
Steven Carr Veteran Member Join Date: Jul 2001 Location: England Posts: 5,629	Out of idel curiosity, how does the paper show any sort of evidence of any kind for any signs of action by any type of intelligent designer?

Freethought & Rationalism Archive

The archives are read only.