|
Author
|
Topic: Data, interpretation and false positive inferences
|
charlie d.
Member
Member # 159
|
posted 16. October 2003 08:18
That's cool technology, rafe!
I just wanted to point out, however, that I don't really think Wang's results "falsify" Axe's. Axe's data do show that there are more constraints than expected in the neutral evolution of a protein within a single functional realm. This may genuinely have some implications in how we envision the evolution of certain proteins. Unfortunately, I don't think the enzymes he chose are the right ones for this argument, especially beta-lactamases can be considered not to be under constant selection, so temporary decreases in function, even severe, may be more tolerated. It would be more interesting to study the same effect in crucial metabolic enzymes, or things like cytochome C, whose function is required and whose essentially neutral evolution through extended evolutionary periods has been postulated and studied.
What Wang shows is that if different functions are allowed to be selected for, there are ways to go through deep valleys of nonfunctionality in some "dimensions". This I suspect would be OK with Axe (at least as far as I can remember from the discussion in his paper), but contradicts Dembski's simplistic interpretation of Axe's results as "prohibiting" evolution altogether (this kind of anti-darwinian "silver bullet", from the EF to the NFL theorems to this, seems indeed to be Dembski's idee fixe).
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 16. October 2003 21:37
Hi Rafe, stopped by for the chat and saw your reply.
Rafe writes:
quote:
if the author really wants to conclude that TEM has "extreme functional sensitivity", then he's entitled to do so, but as soon as the wang paper came out, axe's conclusion was falsified.
I'm afraid not. In fact, I'd say, that the Wang paper supports Axe's conclusions. Axe concluded that most substituations are destabilizing. You can see this by comparing Figure 6 in the Wang paper and figure 5 in the Axe paper. Wang found this as well, most of the substitutions were also destabilizing. The same subsitution in the paper that stabilized the protein kept comming up, which means there probably aren't a lot that can do this(this is something I will build on in a thread that I am participating in over at ARN once I get some more time on my hands, I'm currently really busy).
As has already pointed out, the "new" function in the Wang paper isn't new at all. In Figure 1b in the Wang paper you can see activity relative to Tem-1, which means that Tem-1 already had this "new" function, and all that is happening is optimization of an old function.
A truly objective reading of Dembski et. al. will show that what is being said is that a new function is unlikely to crop up without complete disruption of the folded structure (and by extention any function).
As to the technology you write about in your post, I'll take a look at it when I get some time (pretty busy at the moment). However, I did do a cursory reading of it and I think you are misunderstanding the Axe paper.
The Axe paper is not meant to show that all changes whatsoever will destroy function completely. That would be as problematic for Design, in my opinion, as it would be for Darwinian evolution. No. The Axe paper discusses this view:
quote:
These observations have led to the current view that functional constraints on sequence are minimal at these positions. Here, it is shown
that this inference assumes that the set of acceptable residues at each position is independent of the overall sequence context. Two
approaches are used to test this assumption.
In other words, if this is true we should be able to change the exterior positions a lot. We should be able to change all 171 non-active site positions and still maintain the original function. What Axe then shows, is that a small percentage of these are changed, and original function is disrupted. Thus "Extreme Functional Sensitivity".
[ 16. October 2003, 22:35: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 16. October 2003 22:15
quote:
A truly objective reading of Dembski et. al. will show that what is being said is that a new function is unlikely to crop up without complete disruption of the folded structure (and by extention any function).
Nelson, are you saying that for a beta-lactamase to evolve a new function, its structural fold has to be completely disrupted?
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 16. October 2003 22:17
Charlie,
Not really, the greater the change needed to produce a new function, the greater risk of catastrophy to the folded structure.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 16. October 2003 22:23
So, a completely new function would require a complete redesign of the structural fold. Correct?
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 16. October 2003 22:25
Not sure. As I said, more work should be done.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 16. October 2003 23:24
I see. So, there is in fact no conclusive evidence that "a new [protein] function is unlikely to crop up without complete disruption of [that protein's] folded structure". Fair enough. Thanks for clarifying this.
IP: Logged
|
|
rafe gutman
Member
Member # 134
|
posted 16. October 2003 23:56
quote:
charlie: I just wanted to point out, however, that I don't really think Wang's results "falsify" Axe's
sorry, when i say falsify, i'm more referring to dembski's interpretation of axe's data, rather than axe's own conclusions. i really don't feel like rereading the axe paper, but my initial impression was that axe's own conclusions weren't nearly as strong as dembski's. let me repeat dembski's statement:
quote:
But there is now mounting evidence of biological systems for which any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever (Axe, 2000). For such systems, neither direct nor indirect Darwinian pathways could account for them.
the fact that wang's mutants were functional falsifies this statement, in my opinion.
quote:
rafe: if the author really wants to conclude that TEM has "extreme functional sensitivity", then he's entitled to do so, but as soon as the wang paper came out, axe's conclusion was falsified.
nelson: I'm afraid not. In fact, I'd say, that the Wang paper supports Axe's conclusions. Axe concluded that most substituations are destabilizing.
so nelson, if axe concluded, in your words, that "most substitutions are destabilizing", do you feel that dembski is justified in interpreting that to mean "any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever". do you see the difference between the words "most" and "any"?
quote:
nelson: A truly objective reading of Dembski et. al. will show that what is being said is that a new function is unlikely to crop up without complete disruption of the folded structure (and by extention any function).
but that's not what dembski said, he said that "any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever". you keep trying to downplay these conclusions.
quote:
nelson: I think you are misunderstanding the Axe paper.
The Axe paper is not meant to show that all changes whatsoever will destroy function completely.
hmm, then why did dembski state that "any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever"? so far every statement you've made contradicts dembski's conclusion.
quote:
nelson: In other words, if this is true we should be able to change the exterior positions a lot. We should be able to change all 171 non-active site positions and still maintain the original function. What Axe then shows, is that a small percentage of these are changed, and original function is disrupted. Thus "Extreme Functional Sensitivity".
if this is what axe means when he says "extreme functional sensitivity", then i have no problem with axe, although i do disagree with his choice of the word "extreme". but of course, as i've already mentioned, this is not what dembski said in his summary.
let me ask you a question, nelson. if, instead of the 20 hybrids that axe screened, we screened 100,000, do you think that all 100,000 hybrids would be non-functional?
[ 16. October 2003, 23:58: Message edited by: rafe gutman ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 17. October 2003 00:29
Charlie wrote:
quote:
I see. So, there is in fact no conclusive evidence that "a new [protein] function is unlikely to crop up without complete disruption of [that protein's] folded structure". Fair enough. Thanks for clarifying this.
I think Axe's paper suggests such a thing.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 17. October 2003 00:46
Rafe writes:
quote:
sorry, when i say falsify, i'm more referring to dembski's interpretation of axe's data, rather than axe's own conclusions.
Why are you changing what you said? You stated:
quote:
if the author really wants to conclude that TEM has "extreme functional sensitivity", then he's entitled to do so, but as soon as the wang paper came out, axe's conclusion was falsified.
And you're calling me a hypocrit? Thats really shameful rafe.
Rafe quotes Dembski:
quote:
But there is now mounting evidence of biological systems for which any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever (Axe, 2000). For such systems, neither direct nor indirect Darwinian pathways could account for them.
If for in order to get a new function there is a risk of catastrophic disruption of the folded structure, and a small percentage of the exterior positions causes this, then I see what Dembski was saying.
Rafe writes:
quote:
so nelson, if axe concluded, in your words, that "most substitutions are destabilizing", do you feel that dembski is justified in interpreting that to mean "any slight modification does not merely destroy the system's existing function, but also destroys the possibility of any function of the system whatsoever". do you see the difference between the words "most" and "any"?
I think he's referring to the small fraction of the 171 non-active sites as "any". However those were two different sentences. As you yourself stated:
quote:
i think dembski calls it "extreme functional sensitivity" because he doesn't understand the paper. this is highlighted by his use of the word "slight" to describe changes of up to 30 amino acids.
Context, Rafe, context. I'm not going to argue semtantics here, but I usually give people benefit of the doubt, since I myself have been grossly misinterpreted by ID critics. By the way, you never answered this question:
quote:
The structural stability of the new cephalosporin-active enzyme can then be recovered by an otherwise silent mutation, Met182-->Thr (of the kind that Axe would have said are ultimately detrimental).
Do you agree, or disagree with this?
Also, was there non-functionality in the Wang mutants?
Rafe writes:
quote:
let me ask you a question, nelson. if, instead of the 20 hybrids that axe screened, we screened 100,000, do you think that all 100,000 hybrids would be non-functional?
Not sure, looks redundant to me.
[ 15. February 2006, 13:44: Message edited by: Nel ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 17. October 2003 02:53
It does look redundant. So, why is Diversa screening hundreds of thousands, instead of ~20, before giving up?
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 30. October 2003 17:10
Rex,
Here is one of the purposes of the technology
quote:
By performing GSSM on a gene encoding a protein, all possible single, double, and/or triple amino acid codon substitutions within a protein are created, removing the need for prior knowledge about the protein structure and allowing all possibilities to be tested in an unbiased manner.
Wang's paper pretty much provided us with a means to "remove the need for prior knowledge about the protein structure". Wang did pretty much the reverse of what Axe did, and came up with the same results. The technology would be nice to apply to this question, but I don't think it would make much of a difference.
[ 30. October 2003, 17:11: Message edited by: Nelson-Alonso ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 31. October 2003 19:07
![[Confused]](confused.gif)
I don't get your last post, Nelson, care to elaborate?
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 23. November 2003 14:32
Sure. Wang started with a selection-based approach to identify substitutions, which they then characterized structurally, whereas Axe started with a structure-based approach. Would you like a PDF?
[ 15. February 2006, 13:46: Message edited by: Nel ]
IP: Logged
|
|
|
Printer-friendly view of this topic