Stylus paper (from Biologic Institute) published

_{Paul Nelson

June 5, 2008

Intelligent Design}

Share: Facebook; Twitter; LinkedIn; Flipboard; Print; Email

Stylus Figure One

It’s open access: check it out.

Comments

the paper reports the results of a computer model running simulations, not actual experiments using actual proteins. As such it’s mildly interesting from an engineering point of view, but from the standpoint of the natural sciences…well, it isn’t natural science at all, not really.
Sort of like Avida huh?Mung_{June 6, 2008
June
06
Jun
6
06
2008
09:48 PM
9
09
48
PM
PDT}

Wow. This is huge. Cutting edge. And from the Biologic Institute no less, those evil purveyors of the not-so-secret theocratic schemes of the anti-science dogmatists at Discovery Insitute. What will they come up with next?Mung_{June 6, 2008
June
06
Jun
6
06
2008
09:45 PM
9
09
45
PM
PDT}

JunkyardTornado: I forgot: you are right that I have not commented on your original points, but I was really sidetracked by your following points! I agree that it is not comforting to make points and having nobody who comments upon them. . It happens often to me too,and I don't like it either. But it's how blogs work, and we have to accept it.gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
10:22 AM
10
10
22
AM
PDT}

JunkyardTornado: Please, relax. The problem here is not to be right or wrong, but to exchange our ideas. You may not believe me, but I don't want to be derogatory against anyone, including you, but I like some lively intellectual confrontation, and sometimes a little bit of irony. Only when my interlocutor manifests intellectual insincerity, maybe I can become a little bit aggressive. But it's a game, after all. I don't agree with all this reasoning about teams. I am absolutely pro ID, but that's only because I do believe in ID. I don't know who the people from the biological institute are, and if they are ID I am glad for that, but I have to agree with what they say. Anyway, I have always had contrasting feelings about the article we are discussing: I have always said it is interesting, but at the same time I feel there is something wrong in the general assumptions, and that it tends to generate sonfusion about important questions. So, I have tried to express more explicitly my thoughts, and you have been kind enough to stimulate me. Back to your comments. "OK there has to be a strict correlation, otherwise the model is irrelevant. Got it. 95% correlation? Not strict - therefore irrelevant." It's not a problem of quantity. The problem is that protein structure is not the same as protein function. They are two different concepts. Protein structure and folding contributes to function, but, for instance, the same active site can be built using completely different domain structures. So, you cannot derive function only from the genral similarity in folding structure, even if that can be a pertinent clue. "An english sentence is a machine because it has an objective meaning determined by the arrangement of its constituent parts. A Han character also has constituent parts that determine its meaning. New characters (i.e. new functions) are formed in Han as well by recombination of these constituent parts." Well, if you want to remain confused, there is nothing I can do, I will make a last try. I am not denying that an english sentence has a function, only that function is a symbolic function: the function of a sentence is to transmit information. In general, I would not call a symbolic functional information a machine, but if you like to do so, you're welcome. What is the difference between that symbolic function and the function of an enzyme? Simple. The function of an enzyme is different: it is not to transmit information. It is not symbolic. The function of an enzyme is to react with a substrate and catalyze a specific chemical reaction. No symbol in that. I would definitely call that a machine. DNA is another matter. Take a protein coding gene, for instance. It is not a machine. It is a piece of information. It is like the english sentence. It is symbolic. In itself, it is useless. But, when transcribed and translated by the appropriate molecular machines, it serves as the information source for a protein. In other words, the symbolic information stored in DNA becomes part of a working machine, which is the cellular machine for protein synthesis. Let's go to Han characters. It is a symbolic functional character, not a machine. Its parts contribute to its function, which is to convey a meaning. Its function is symbolic. The contribution of the various parts consits only in making the character recognizable to those who have to recognize it. You say: "New characters (i.e. new functions) are formed in Han as well by recombination of these constituent parts." No. You can't form a new Han character by recombination. You can form another existing character, but not a new one. A new character cannot exist unless we define it in advance. You could create a new form which could in theory be a new character, but it would mean nothing until a meaning is assigned to that form by someone. Why? Because a character is a symbol. It represents something else, but in itself it is only a geometric form, without meaning. Not so a protein. You can certainly, in principle, create a new protein, which has never existed before, and observe that it can fold in a new way, and express a new function. For instance, it could promote a chemical reaction for which no enzyme has been known before. You don't need to assign that function to the new protein. You just observe that the function is there. Finally, I have reviewed the arguments in Allen MacNeill's post, to understand why you insist in saying that I make the same arguments. Well, I don't understand. Allen says: "As such it’s mildly interesting from an engineering point of view, but from the standpoint of the natural sciences…well, it isn’t natural science at all, not really." That implies that engineering models cannot be really useful form the study of natural sciences. That's typical Allen MacNeill, and I absolutely disagree. I do believe in the utility of theoretical and engineering models in natural sciences. I just think that those models have to be judged according to their own merits, in other words we cannot accept any model, we have to verify how well a model models. Allen says: "But then, that’s just what one might expect from an institute founded by people who have never actually done any genuine empirical science." Typical AllenMacNeill, and I would say, rather irritating. I have never had such an adoration for empirical science. It is important, but for me theoretical science is even more important. Allen's affirmation is the same as saying that theoretical physics is useless, and that only experimental physics is important. That may be Allen's thought, but it is not certainly mine. Allen says: "The one interesting feature of the paper is the suggestion that it might be possible to sort through a huge number of possible protein configurations using a two-dimensional “reduction” equivalent to a Chinese ideogram." I agree, that is an original contribution, and it could be the basis for further thought. But that is an appreciation, and not a critic. Allen says: "We will have to wait and see if this technique is adopted by anyone who actually works with real proteins." Again, the anti theoretical and "I love experimental" attitude. I totally disagree. Allen says: "Personally, I think the idea that “meaning” in a three-dimensional protein can somehow be captured as “meaning” in a two-dimensional ideogram is pretty sketchy, but I’ll wait and see if it works…" Here I find the same confusion that you have shown. Allen seems not to be aware of the difference between "meaning" and "function" in the context of proteins. I don't agree. In a sense. here it is you who are in the same team with Allen (which, if we reason in terms of teams, should not be a surprise at all). Let's see what DLH has said about Allen's post: "I am surprised you are so dismissive of theoretical research that can provide a basis for predictive models and directing experiments in the incredibly large search field of genomics and proteomics." Excuse me, but I would say that it's exactly what I have said. Why in the world should DLH, who obviously agrees with me in many things, criticize me for what Allen says, which is the opposite of what I am saying? Maybe what you see as team loyalty is only loyalty to our sincere beliefs...gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
10:18 AM
10
10
18
AM
PDT}

Someone going by the username farslayer9 just gave it a one star rating, apparently after having thoroughly read, reviewed and digested this rather complex treatise. He or she also thought the paper deserved a low score on insight, reliability, and style. This only after one or two days after the article was published. Now c'mon! If you spent all that time trying absorb the contents of such a paper, shouldn't it have deserved some credit for piquing and sustaining your interest? If not, at least for originality and insight.JPCollado_{June 6, 2008
June
06
Jun
6
06
2008
09:38 AM
9
09
38
AM
PDT}

gpuccio: Yeah, comparison of the chemical formula for a protein to the protein was invalid. After I posted I realized it was just a pointless rhetorical argument with no merit, and thought about retracting it. But you jumped all over it. Good for you. Everything else I said was on target. In that sense, a Han character corresponds, at best, to a 3D structure, but as I said protein function is not a strict consequence of structure OK there has to be a strict correlation, otherwise the model is irrelevant. Got it. 95% correlation? Not strict - therefore irrelevant. The problem, IMO, is that a Han character is not, in itself, a functional machine. It is a symbol, whose function derives from a symbolical meaning. A symbolical meaning cannot be tested for its function. Moreover, no new function can be created, unless a new symbolical meaning for a new character is established in advance. An english sentence is a machine because it has an objective meaning determined by the arrangement of its constituent parts. A Han character also has constituent parts that determine its meaning. New characters (i.e. new functions) are formed in Han as well by recombination of these constituent parts. Why on earth, DLH doesn't feel compelled to take you to task for making esentially the same arguments as A. Macneil is beyond me. It is because you root for the same team. Its clear you didn't realize in your orginal post that it was from the Biologic Institute- thus your derogatory tone towards their research. Now of course you realize it but have to be consistent. Of course none of this has anything to do with my original points in my original post above, which you or no one else has commented on.JunkyardTornado_{June 6, 2008
June
06
Jun
6
06
2008
08:48 AM
8
08
48
AM
PDT}

JunkyardTornado: Ah, the old good disagreement: "I guess if someone wrote out the chemical formula for some protein you would say that was just a symbol as well and not a machine." Sure, a chemical formula is a symbolic storage of information, and not a machine. Are you doubting that? "DLH should have directed his comments he made to Allen_MacNeill to you as well:" No, I perfectly agree with DLH on the importance of theoretical research. And I really don't feel like agreeing with Allen ManBeill! But that does not mean that a theoretical research cannot be analyzed for its own merits. I have never criticized tha article because it is theoretical research, I have simply commented about its specific content. "So here you’re saying that their system is unnecessarily restrictive, because many more functions could exist than they allow for. So IOW, you think evolution is easier than their system implies." No. I thought I had been clear, but here it is again. I have made two different objections. One is that new unknown functions seem not to be available. That would make the system unnecessarily restrictive. The other is that "evolution" is guide by a specific measurement of proficiency, based on known and stored information. That makes the system a model of guided, designed evolution. So, I am sorry to disappoint you, but definitely I don't "think evolution is easier than their system implies.". I think their system, although interesting, is not a model of unguided evolution for at least two different reasons. Just to be clear, I absolutely don't think that evolution (non designed) is easy, nor that it is difficult: I just think it does not happen. "Put up the Han symbol for stop at an intersection and see if it objectively causes anything to happen. The meaning of a protein is determined by its context and how things in its environment react or respond to it - same with Han pictographs." You keep making a great confusion between symbolic meaning and function. I can't understand why. Have you ever enjoyed a good drive using the word "car"? Have you ever satisfied your hunger pronouncing the word "food"? A symbol is a symbol. It is not the thing itself. That should be an easy concept. A symbol transmits information, and depends on a context. Chinese characters transmit no information to me because, unfortunately, I can't understand them. A machine performs a task, and depends on a context. A protein needs the right environment and substrate, a car needs a street, and so on. So, both symbolic information and machines depend on a context. That's true. That does not mean they are the same thing. They are not. You should check your logic. The meaning of an enzyme, its name, its formula, the knowledge about chemical properties governing its strucutre and function, all of these things are symbolic information. They mean something only to intelligent beings who understand them. But the enzyme itsel, in the right context, works even if no one is aware of that. Unless you don't believe in the objective existence of the outer world, a machine is part of that outer world. It exists, it works, it is observable, its effects are measurable, and usually repetitively measurable. A machine does nt depend on intelligence for its working: it just depends on intelligence for its existence (it had to be designed). You say: "The meaning of a protein is determined by its context". No. The function of a protein is determined by its context. More exactly, the function of a protein has to be made possible by a specific context. Out of that context, a protein usually has no function. Its formula, its "meaning", can remain the same, but the function is no more available.gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
08:14 AM
8
08
14
AM
PDT}

gpuccio: I guess if someone wrote out the chemical formula for some protein you would say that was just a symbol as well and not a machine. DLH should have directed his comments he made to Allen_MacNeill to you as well: "I am surprised you are so dismissive of theoretical research that can provide a basis for predictive models and directing experiments in the incredibly large search field of genomics and proteomics." Also, regarding your comment: Moreover, no new function can be created, unless a new symbolical meaning for a new character is established in advance. So here you're saying that their system is unnecessarily restrictive, because many more functions could exist than they allow for. So IOW, you think evolution is easier than their system implies. In other words, meanings are subjective functions, they have to be recognized by an intelligent agent, or by a machine instructed by him. On the contrary, machines objectively do something Put up the Han symbol for stop at an intersection and see if it objectively causes anything to happen. The meaning of a protein is determined by its context and how things in its environment react or respond to it - same with Han pictographs.JunkyardTornado_{June 6, 2008
June
06
Jun
6
06
2008
06:44 AM
6
06
44
AM
PDT}

JunkyardTornado: Well, even if we did not say the same things, I think there was some common ground in our reservations about the software. If that's not so, I will be happy to disagree with you as usual. :-) The problem, IMO, is that a Han character is not, in itself, a functional machine. It is a symbol, whose function derives from a symbolical meaning. A symbolical meaning cannot be tested for its function. Moreover, no new function can be created, unless a new symbolical meaning for a new character is established in advance. In other words, meanings are subjective functions, they have to be recognized by an intelligent agent, or by a machine instructed by him. On the contrary, machines objectively do something, even if here again an intelligent agent is necessary to recognize that they do. There is a difference. A machine can be objectively tested. If a new machine evolves, and it does something new, that can be recognized. In other words, objective function can be objectively recognized and measured. In that sense, a Han character corresponds, at best, to a 3D structure, but as I said protein function is not a strict consequence of structure: proteins with very similar structure may have different function, while protein with completely different structure may have practically the same function. There are well known examples of both situations. Moreover, function can be directly measured, while the correspondence with a Han character must be assessed on the basis of a comparing algorithm which needs the knowledge of the reference information (the Han character itself). Just to be more clear, if we have different antibodies, we can measure their specific affinity with a given antigen, even if we have no idea of the 3D structure of the antibodies, of their sequence, or if any other information. Function is immediately relevant, direct and measurable. On the contrary, if we had to evaluate our antibodies for their correspondence with a reference antibody, we should know the 3D structure and possibly sequence of all of them, of the reference, and still we could draw wrong conclusions about their function. These are, in essence, my reservations. Anyway, I appreciate the paper for its originality, and yes, I am aware that the Biologic Institute is an ID institution. And so? We are not talking football teams or politics here. I need not endorse everything ID people do or say, and I need not criticize anything darwinists do. As I have often said, most of the evidence for ID, almost all, comes from darwinist research. The interpretation, obviously, is all another matter...gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
06:14 AM
6
06
14
AM
PDT}

gpuccio: "For once, I agree with you. The main limit of this model and software seems to be that “function” is represented by the correspondence with existing Han characters, and proficiency is measured that way... Anyway, the model could be interesting to explore the space of 3D folding and the consequences of variation on that, but it cannot say anything about the possibility of new functions arising from random variation. For the same reason, there cannot be any exploration of truly random paths or of the effect of selection on them, because “selection” is not based on true function, but on a predetermined pattern, and therefore on an artificial “measurement” which requires the previous input of the required information..." I don't have any problem with their analogy of Han symbols to protein function. Your objection seems to be the same as Allan MacNeil's. Their whole objective was to find a model more in line with reality than what is typically used (the "lattice" model I think they termed it.) The paragraph I quoted in my post above lays out their rationale somewhat convincingly. 4000 known enzyme function - 5000 Han Characters. The line and curve segments of Han characters map to the 3-d structure of proteins. The respective systems are also each hierarchical in nature, being built up from preused parts. A Han character corresponds to a meaningful function. What's the problem? The nature of my objection was completely different. I assume you understand the Biologic Institute is an ID organization.JunkyardTornado_{June 6, 2008
June
06
Jun
6
06
2008
03:10 AM
3
03
10
AM
PDT}

Sorry, I left a sentence unfinished. Should have been: Any system working this way, and that includes AVIDA and practically all GA, is just another instance of the "Methinks it's like a weasel" model, and while it can show that a designed random search can work (which we already know), it says nothing about non designed random searchs.gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
02:43 AM
2
02
43
AM
PDT}

JunkyardTornado: For once, I agree with you. The main limit of this model and software seems to be that "function" is represented by the correspondence with existing Han characters, and proficiency is measured that way. Now, in proteins there is no exact correspondence betwee, folding and function, only a general one, and anyway function has to be defined "functionally", that is as the ability to act as a machine and do something. Anyway, the model could be interesting to explore the space of 3D folding and the consequences of variation on that, but it cannot say anything about the possibility of new functions arising from random variation. For the same reason, there cannot be any exploration of truly random paths or of the effect of selection on them, because "selection" is not based on true function, but on a predetermined pattern, and therefore on an artificial "measurement" which requires the previous input of the required information (in this case, the database of characters). Any system working this way, and that includes AVIDA and practically all GA. I have alway thought that any serious GA should look for evolution of function from random noise in a context which can have all the restraints we want, but which must not incorporate in any way the knowledge of the function which will evolve. But obviously, such a GA would easily falsify darwinian evolution, so I don't think it will be easily created, or accepted, by the official academy.gpuccio_{June 6, 2008
June
06
Jun
6
06
2008
02:41 AM
2
02
41
AM
PDT}

"...the new [setup] provides approximate correspondence between the number of Han characters and the number of distinct protein folds or functions in the biosphere. The standard enzyme classification scheme, for example, covers just over four thousand known enzyme functions which depend upon a few thousand family-level structures. By way of comparison, the Unihan database indicates that roughly five thousand Han characters find use in a single language... Visual discrimination of so many characters requires structural complexity beyond that of alphabetic characters, approaching the complexity of protein folds in some respects. Figure 1 illustrates the rough similarity in the number of parts (the line or curve segments that form strokes compared to the elements of secondary structure) that compose whole characters and whole proteins. Finally, both worlds exhibit hierarchical structure, meaning that complex forms are built from successively simpler forms (see Figure 2), most of which find extensive reuse in a variety of combinations for a variety of functional ends."
-------------- Given the above, one might have thought that using their software, you could run scenarios like the following: Given a Han character, start applying mutations to it. Then the question is, with cumulative mutations, can you stay above a certain "proficiency level" (as they define it) for that character, and transition to any other Han character while also staying above a certain proficiency (i.e. legibility) level for the new character. However the software they've developed allows you to do nothing of the kind. Their software only works with a single Han character, which you trace in using their editing software. This seems utterly useless because any change from that archetypal character is going to be a reduction in proficiency to some degree for that particular character (as the actual Han character will have a proficiency of 1.0). Some comments on the experiment they ran - They took some initial Han character (which they didn't specifiy) and applied a million cumulative neutral mutations to it (where they defined neutral in this instance as .45 proficiency). They did this test 10 times to produces 10 variants of the initial character. Then with these 10 new characters, they applied 100,000 noncumulative mutations to each of them. IOW they would cause a mutation on one of these characters, gauge the change in proficiency, record it, then erase the mutation on that character going back to the previous state it was in, and try another mutation. Then they were able to gauge the average change in proficiency for a single mutation which was recorded in the chart in Figure 18. Now, what they did not do (thankfully) is just take some existing Han character and apply 100,000 noncumulative mutations to it and gauge the average change in proficiency, as that would have been really stupid (as any change would by definition be a reduction in proficiency.) Instead they used these 10 new generated characters, where each of these new characters presumably had some proficiency level between 0 and 1. (Thus its seemingly more meaningful to apply noncumulative mutations to these and gauge the result.) However, they do not even bother to mention what the proficiency of these new 10 characters were (for some reason). But their results from figure 18 were as follows (as pointed out by gpuccio):
"About 3.7% of these non-synonymous changes cause a ten-fold or greater drop in proficiency ...The most common effect is a modest proficiency reduction of 0-5% The shaded region in Figure 18 indicates the neutral zone, where the frequency-averaged effect on proficiency is zero. Point mutations falling within this zone, amounting to about 8% of the non-synonymous total, can accumulate indefinitely without net loss of function."."
But its not clear what these results say, if anything. As noted previously if your initial character is 1.0, every mutation will be a reduction in proficiency. However even if its .9 or .8, won't the vast majority of mutations to it have to be reduction in proficiency (for that character)? Alternatively, something with say .1 proficiency, (for a particular character) any change to it is highly likely to increase its proficiency for that character. And they didn't bother to say what the proficiency of those 10 characters were. But anyway my main objection as I stated previously is that their software only allows you to look at one particular character, not transitions between characters. Did I read this paper wrong.JunkyardTornado_{June 6, 2008
June
06
Jun
6
06
2008
02:03 AM
2
02
03
AM
PDT}

Allen MacNeill is a bigger fan of Avida. http://evolutionanddesign.blogsome.com/2006/07/06/doggies-are-better-than-weasels-2/Charlie_{June 5, 2008
June
06
Jun
5
05
2008
07:36 PM
7
07
36
PM
PDT}

Searching Pubmed for "axe-dd[Author]" gives: 1: Axe DD, Dixon BW, Lu P. Abstract Stylus: a system for evolutionary experimentation based on a protein/proteome model with non-arbitrary functional constraints. PLoS ONE. 2008 Jun 4;3(6):e2246. PMID: 18523658 [PubMed - in process] 2: Axe DD. Abstract Estimating the prevalence of protein sequences adopting functional enzyme folds. J Mol Biol. 2004 Aug 27;341(5):1295-315. PMID: 15321723 [PubMed - indexed for MEDLINE] 3: Axe DD. Abstract Extreme functional sensitivity to conservative amino acid changes on enzyme exteriors. J Mol Biol. 2000 Aug 18;301(3):585-95. PMID: 10966772 [PubMed - indexed for MEDLINE] 4: Axe DD, Foster NW, Fersht AR. Abstract An irregular beta-bulge common to a group of bacterial RNases is an important determinant of stability and function in barnase. J Mol Biol. 1999 Mar 12;286(5):1471-85. PMID: 10064710 [PubMed - indexed for MEDLINE] 5: Axe DD, Foster NW, Fersht AR. Abstract A search for single substitutions that eliminate enzymatic function in a bacterial ribonuclease. Biochemistry. 1998 May 19;37(20):7157-66. PMID: 9585527 [PubMed - indexed for MEDLINE] 6: Axe DD, Foster NW, Fersht AR. Free in PMC Active barnase variants with completely random hydrophobic cores. Proc Natl Acad Sci U S A. 1996 May 28;93(11):5590-4. PMID: 8643620 [PubMed - indexed for MEDLINE] 7: Bailey JE, Axe DD, Doran PM, Galazzo JL, Reardon KF, Seressiotis A, Shanks JV. No Abstract Redirection of cellular metabolism. Analysis and synthesis. Ann N Y Acad Sci. 1987;506:1-23. Review. No abstract available. PMID: 3324849 [PubMed - indexed for MEDLINE] PS It is also interesting to Search for: Allen-MacNeill Google Scholar "Macneill-Allen[Author]" and Searching Pubmed for "MacNeill A[Author]"DLH_{June 5, 2008
June
06
Jun
5
05
2008
06:51 PM
6
06
51
PM
PDT}

Allen MacNeill wrote: "But then, that’s just what one might expect from an institute founded by people who have never actually done any genuine empirical science." Do a PubMed search on Axe, D.Paul Nelson_{June 5, 2008
June
06
Jun
5
05
2008
05:45 PM
5
05
45
PM
PDT}

gpuccio at 2 The probability of "beneficial" to "harmful" mutations has been estimated to be at least lower than 1 in 10,000 and more like one in 1 in 1 million. I would not hold my breath. But come to think of it, what did those ancestral intermediate forms do who had to "hold their breath" until an essential and beneficial mutation came along? Allen_MacNeill, Axe et al. state:
As a full-featured implementation of the vector world, Stylus enables challenging evolutionary problems to be tackled in a model world having a level of realism that may allow informative comparison with biology. Despite the life-size genes and proteins it processes, computational performance is sufficient for useful results to be obtained with modest resources (e.g., the line-of-descent runs to produce sequences in Figure 17 used under 5 minutes of cpu time each on 3 GHz Intel Xeon processors). And because vector-world genes and proteins are described by the same text-file formats used for real genes and proteins, vector-world studies benefit from numerous existing bioinformatics tools. Finally, because the vector world is built on a structure–function relationship that is not only real but also visually intuitive, Stylus offers the attractive possibility of clarifying complex problems.
I am surprised you are so dismissive of theoretical research that can provide a basis for predictive models and directing experiments in the incredibly large search field of genomics and proteomics. PS both links worked for me. The Citation: Axe DD, Dixon BW, Lu P (2008) Stylus: A System for Evolutionary Experimentation Based on a Protein/Proteome Model with Non-Arbitrary Functional Constraints. PLoS ONE 3(6): e2246. doi:10.1371/journal.pone.0002246 Direct link: http://www.plosone.org/article/info:doi%2F10.1371%2Fjournal.pone.0002246;jsessionid=DD3DA8378EE09DEC97DEC597D6848408DLH_{June 5, 2008
June
06
Jun
5
05
2008
04:48 PM
4
04
48
PM
PDT}

P.S. The link posted in this thread doesn't work (at least it didn't work for me). I've downloaded a pdf of the full article, and will email it to anyone who wants a copy. My email is adm6atsigncornelldoteduAllen_MacNeill_{June 5, 2008
June
06
Jun
5
05
2008
03:05 PM
3
03
05
PM
PDT}

I've read the paper in pdf form. It's a report of a modeling study, not empirical research carried out in the field or the lab. Like Behe and Snoke (2004), the paper reports the results of a computer model running simulations, not actual experiments using actual proteins. As such it's mildly interesting from an engineering point of view, but from the standpoint of the natural sciences...well, it isn't natural science at all, not really. But then, that's just what one might expect from an institute founded by people who have never actually done any genuine empirical science. The one interesting feature of the paper is the suggestion that it might be possible to sort through a huge number of possible protein configurations using a two-dimensional "reduction" equivalent to a Chinese ideogram. We will have to wait and see if this technique is adopted by anyone who actually works with real proteins. Personally, I think the idea that "meaning" in a three-dimensional protein can somehow be captured as "meaning" in a two-dimensional ideogram is pretty sketchy, but I'll wait and see if it works...Allen_MacNeill_{June 5, 2008
June
06
Jun
5
05
2008
03:02 PM
3
03
02
PM
PDT}

Interesting. Just a couple of comments. I find it is fascinating that the search space and function space of real proteins is so huge and complex that, in order to test the general principles implied, it is easier to build a ficticious model of them based on human language characters, and in particular Chinese ideograms. The second observation is that the whole thing, although more sophisticated, looks very much like a variant of the famous "Methinks it's like a weasel". Look at the parts where it explains how proficiency is specicically measured. And how? With comparison to the Han archetypes! Finally, I would cite here this interesting paragraph, which will sound familiar to us in ID: "Combining the data from all ten output files (1 million total mutations), we find that just under 75% of the base changes were non-synonymous (causing a vector substitution). About 3.7% of these non-synonymous changes cause a ten-fold or greater drop in proficiency (Figure 18), similar to the proportion of amino-acid substitutions found to inactivate a bacterial ribonuclease (5% [29]). In the vector proteins, only a tiny fraction (0.4%) of the changes are so disruptive that proficiencies could not be calculated (this occurs when strokes in the vector protein cannot be mapped to strokes in the archetype—for example, when two strokes merge into one). The most common effect is a modest proficiency reduction of 0–5% (Figure 18). Again, this is consistent with the behavior of real proteins, which tolerate conservative substitutions in substantial numbers before function is lost completely." Maybe I missed something, but where are the beneficial mutations? Where is the increase in function? What were they expecting, new ideograms with new meanings? Anyway, the Han ideograms are cute!gpuccio_{June 5, 2008
June
06
Jun
5
05
2008
02:02 PM
2
02
02
PM
PDT}

An interesting post. Could someone explain this in more laymanish terms, so I know I'm 'getting' it?nullasalus_{June 5, 2008
June
06
Jun
5
05
2008
12:04 PM
12
12
04
PM
PDT}

You must be logged in to post a comment.

Leave a Reply