Recent paper with an ID spin? Abel and Trevors (2005).

Smoothie, you're giving yourself away.First you adamantly claimed that your belief in geocentrism was not related to religious belief but to science.Now you are arguing for the creationist position on CSI. And you'll probably try to tell us that that idea is based on science as well.Why is it that your "science" always seems to support the straight creationist belief system, and is dismissed by about 99.9% of scientists?(Sounds a lot like creation "science" to me.)

---

Religious belief does not constitute scientific evidence, nor does it convey scientific knowledge.

quote:

---

So SO you seem to have reached the same unanswered point in a seperate thread. Whose knowledge was increased by the information increase that constitutes the formation of humanity (or life on Earth more generally)?

---

I already answered this. An I answered it where I was supposed to. Please leave this thread and don't bother me anymore.

quote:

---

Smoothie, you're giving yourself away.
First you adamantly claimed that your belief in geocentrism was not related to religious belief but to science.

​

Now you are arguing for the creationist position on CSI. And you'll probably try to tell us that that idea is based on science as well.

​

Why is it that your "science" always seems to support the straight creationist belief system, and is dismissed by about 99.9% of scientists?

​

(Sounds a lot like creation "science" to me.)

---

This is my last response to you. You are clueless about any kind of discussion that is going on. Your only argument is an ad hominem. You are constantly accusing me of religious bias. That is all you have to say. Therefore, goodbye and never talk to me again.

Well how about an answer. Do you not have one? This is a debate forum you know.

---

Facts don't lie or have an agenda. Facts are just facts

You kids get off my thread/lawn! If you are already having this discussion in another thread then why bring it here as well?TTFN,WK

quote:

---

This is an assertion with no evidence to support it. A mutation is not a light switch. Genes do not exist in a simple binary state of on or off like a light switch. You are simply making an assertion, surely you realise that the very equations we are discussing show that mutations can increase functional information, at least in theory. You seem to be denying even the theoretical possibility of an increase in information, but it isn't clear on what basis.

---

Yes, it can be.Throw of a dice dictates a bug's life | New ScientistBacteria can alter ther diet when 300 permease proteins are found within the cell, the cell swithches it's diet with geentic regulation. The cell than gains the ability to digest sugar.Now you claim it has been shown that CSI can increase with random mutations and natural selection. Please show me where.

quote:

---

Again, a blank assertion with no evidence.

---

Why should I be proving a negative? You should give me evidence for your positive assertation.

quote:

---

This is why the Durston et al. measure of function is meaningless, it doesn't measure anything as they use it. The Hazen paper gives specific examples of how to measure the function of specific sequences and use that to weight the functional information. For Durston et al, where 'function' is only a proxy measure of conservation in PFAM families it is true that a single novel de novo point mutation will not increase the FSC because maximal FSC in their scheme is represented by residues with 100% conservation. So only a mutation in a population which brought the sequence further into line with the consensus sequence, or made it more similar to another sequence in the alignment, could increase the overall FSC of the whole alignment. This is entirely divorced from the actual function of the protein however, and is merely a measure of conservation.

​

It seems strange that the IDists who harp on about the importance of functional information don't want to produce a usable concept of function.

---

That is simply how they do it. Measuring the conservation in different sequences is how they measure functionality.

quote:

---

Yes, of course they are. Both in terms of biochemistry as a whole and in terms of specific amino acid sequences for individual proteins. Surely this is one obvious corrollary we can both agree on of Doug Axe's work. Some amino acid substitutions have larger functional effects than others and some amino acid positions are more sensitive to changes than others. All amino acid substitutions are far from equal. There are arguably some amino acids we could do without entirely, that is one of Dryden's main points when they say that you could produce the majority of known functional folds from a repertoire of only a handful of amino acids, in extreme cases possibly only 2.

---

That is true only after you ahve a protein. Before you have a protein all amino acides are equaly important, which is to say they are all not important in any way, since you have nothing to do with a bunch of ranom amino acids.

quote:

---

No it isn't if you see no difference then you are either blind or can't read. I spelled it out for you right in the paragraph you quoted. Durston et al. only look at the conservation within the PFAM alignment they are studying. BLOSUM does it on a large number of highly conserved aligned sequences to draw general functional relationships which they use to weight specific substitutions.

---

That's the same thing. One is based ony PFAM families the other is not.

quote:

---

They aren't improving, they are ignoring the already existing methods which have improved on what they are doing.

---

Well, you need time to improve something. Why not just be still for some time and let them do their work?

quote:

---

It does to the same extent as Durston et al. do, by conservation, but they also taks into account the functional effects of the substitutions at the various amino acid positions, measured in two different ways.

---

That is the same thing. How can you measure conservation if you do not take into account whic amino acid has just changed from the previous sequence you measured?

quote:

---

Well so is Durston et al. look at the program they wrote and you will see that calculating shannon entropy is part of the algorithm. That doesn't mean they are only looking at shannon information. As I have argued they have more connection to functionality than the Duston et al. method, and the Hazen method has more again.

---

I know it is, all of them are. I have not see any that is not based on Shannon's work. That is the basis for everything new in information theory. But Shannon's model alone is not enough.

quote:

---

And as I said, Axe's numbers are no more widely applicable than Durston's for the reasons I gave previously that you have yet to address.

---

I told you, it's a simple matter of extrapolation. There is nothing wrong here. You know very well that you can't just mutate proteins to infinity.

quote:

---

I agree that is what he is saying but eh is also actually sayin what it actually says in my quote. You can't just handwave away the fact that Szostak says that Functional information " is required to account for all possible sequences that could potentially carry out an equivalent biochemical function, independent of the structure or mechanism used". Which Durston et al.'s approach simply does not give us a framework for.

---

Yes it does. It's what Axe's work us used for. He did the work with proteins and changed them to where they stopped functioning as they should. This area of functionality, actualy the rpobability, is the number you need.

quote:

---

Any knowledge about the full set of possible sequences functionally equivalent to RecA.

---

Again, Doug Axe. Please explain, what do you find wrong with his work? Why can't his work be extrapolated to find the modificational possibilities of RecA?

quote:

---

Maybe you can, but that doesn't mean that such an extrapolation is reliable. Is 1 protein really a suitable proxy for all the possible proteins in existence? Not to mention it being a protein already mutated to be on the edge of functionality.

---

It's the best we got. Sure, you could make a better estimate, but not by much. I wouldn't be surprised if we even got a smaller number. You do know we can't get an infinity of possible modifications. Yet, that is what you are searching for. Or close to infinity. You simply won't get that. There is a limit.

quote:

---

Except he won't, he isn't doing an exhaustive screen of all possible functional sequences even in that one protein. He is pushing some of the limits of functionality in one protein and extrapolating from them to the entire functionality space of all proteins.

---

Umm... yeah, I know. It's how extrapolation works. What's wrong with this experiment, except nothing? Do you think there are soem proteins that can be coded for with infinite numbers of sequences?

quote:

---

Please explain how he does this. We aren't talking about 10 sequence variants of one protein, we are talking about 10 totally distinct 1ary sequences with functional equivalence. How does Axe's work even begin to address the existence of these functionally equivalent proteins?

---

Very simple. You take a protein, you mutate it, and you find where the edge is. When the edge is passed, your protein becomes non-functional. All the working structures are still considered functional. All those functions could have been designed, or as you believe evolved from the start. So the more deformable the protein is, and retains it's function, the more chance we have for that protein to evolve.

quote:

---

Saying this over and over again doesn't change anything. They quite explicitly say this is to remove indels because those indel regions could indeed inflate the FSC measure.

​

This is completely divorced from the concept of the entire functional sequence space that I am talking about, which would be every single possible sequence, alignable or not, which would fulfill the functionality that is being used as the F criterion in the analysis. Durston et al. assume that the PFAM family alignments are a sufficient proxy for this, but they are making a massive and obviously wrong assumption.

---

Well if they could inflate the FSC, than what are we arguing about? They are not wrong. OK, I understand you have another objections. Please explain what else is wrong.

quote:

---

Well those are two very distinct ideas. Evolutionary theory would generally hold that it would produce a flagellum gradually from elements that are similar. There are some cases of apparent radical de novo generation of new genes, but those are rare cases. As I said before your concept of biological function doesn't accord with either Szostak or Durston's.

---

I do not care, simply because I do not use their way of measuring biological functions. And I also know evolution is supposed to build a flagellum, but has it ever been observed? Not really. Do we have anything from which we can extrapolate from? Not that I know of...

quote:

---

Again you make your definitions up for yourself. Szostak clearly considers both of these suitable biological functions since he presumably approves the use of them as examples of such in the Hazen paper. Indeed the functional sequence that Durston et al. are happy to have pulled out in their analysis of Ubiquitins is a DNA binding site.

---

I'm not making anything up. Durston agrees that there is algorithmic information which is describing natural law. Simple chemical reactions are not biological functions.

quote:

---

That isn't what he said, I recommend you read it again.

​

What does this have to do with catalysis and binding affinity?He talks about ...

---

It is nto enough, that is the point. You can also use random measure of information liek RSC and get a measure from some biological function. But that measure is wrong. Simply because it does not measure the functionality.

quote:

---

But you dismiss biochemistry as mere natural law and not connected to function. You aren't talking about function as any biologist understands the term.

---

Because it is a natural law, like gravity. Ink will not fall on a piece of paper and produce a poem. Atoms of hydrogen and oxygen when they come together will form water. Amino acids will not form biological functions when they come together. There is no biological law to make them do that.

quote:

---

Except they do and the Hazen paper quantifies how much in at least one instance.

---

If that is true, than why does Durston, who des not agree with this, bases his work on Hazen's?

quote:

---

You don't want us to extrapolate, you want us to recapitulate the evolution of one specific functionality.

---

Ga ahead and extrapolate. I'm just asking you, from what are you going to extrapolate to what?

quote:

---

Unfortunately this doesn't actually let you measure anything objectively. And it is surely a measure of information rather than functionality? The two things are distinct but related in terms of functional information, but one surely cannot substitute for the other?

---

Of course it does. It measures specification, which can describe a function, and the amount of enteties like nucleotides, or aminoa cids, or proteins, needed to form that specification. That is the complexity and specification together, and it measures the information of biological functions.

Hi SO,Sorry to take so long to get back to you. To be honest I'm doubting the point in continuing this discussion, we don't seem to be progressing at all. From my side it seem like I raise points which directly address your claims, or rather those of Durston and Doug Axe, and you in turn simply repeat the claims and affirm that they are true because they are what Durston and Doug Axe said. You are getting confused here between genetic regulation and evolution. These genes aren't expressed in the absence of a certain environmental trigger, but they are present in the genome with all of the genetic sequences required for their expression. These cells always had the ability to digest sugar, they simply weren't expressing those genes when there wasn't any sugar in the environment. No I dont. I'm just saying that there isn't any evidence that they can't. At the moment I don't think CSI is a clearly enough defined concept for this to be done. Indeed, they measure it by an indirect proxy method which tells us very little about the range of possible functional sequences. Why when Hazen et al. demonstrate how it can be tied directly into actual measurable functions? Surely the functionality is the whole crux of this approach so I still fail to see why Durston et al. approached it in such a perfunctory manner. You seem to be trying to slip into a discussion of abiogenesis here, which is totally beside the point. No it isn't and I have described the differences in the conservation measures to you at least twice now. And I assume you know that that is a gigantic strawman and very far from being a cogent argument. The fact that you can't limitlessly mutate proteins and have them retain function does not mean that you can't substantially mutate them and retain function. In fact Doug Axe's own research showed that functionality could be maintained after 20% of a proteins surface residues had been altered. Except as Hazen and Szostak both emphasise, you need to effectively know all of the possible functional sequences for this to actually work, whereas Axe focuses onto a very narrow range of functional states by making the protein minimally functional to begin with. Because they are totally different proteins with totally different functions. Axe's work didn't even explore all the functional modificational possibilities of the protein he was studying. It may the the best that ID has but there is a wealth of comparative genetic data showing functionally similar proteins with lots of highly divergent genetic and amino acid sequences. I may post something to let us look at this in more depth at some other point. There is a very large amount of space between a handful of possible sequences and an infinite number, you are excluding the middle pretty strongly here. You clearly don't understand my point. There are totally distinct amino acid structures that can perform equivalent functions. How can Doug Axe's approach ever identify these since as you say they are local modifications of a specific existing proteins structure? That was never the point, as I have explained before. I didn't say they were using every possible method to inflate the FSC measure, I was pointing out that the fact that they assume that a PFAM alignment is a suitable proxy for all possible sequences with a given function will inflate the FSC measure. Well he certainly doesn't say so in his paper. This is arrant nonsense they are making direct measurements of biological function. To equate that to RSC is to essentially say, 'well sure you can evolve novel functions like that but they aren't real novel functions'. You tell me, you are the one who is saying that this is Durston's position. He avoids giving a usable definiton of biological function in his paper, as I complained before. From the various existing genetic sequences we are familiar with, the research into hypothetical ancestral sequences and their functions and from a growing body of research into the functionality of large scale artificially generated random and evolved sequences, i.e. SELEX.As I say we currently don't seem to be going anywhere, maybe instead of these mammoth omnibus post to each other we could focus on one specific point and actually discuss it in enough depth to reach some sort of conclusion. My personal favourite would be the question of how well we can estimate a total functional sequence space and how it aligns with Doug Axe's approach.TTFN,WK

quote:

---

Hi SO,
Sorry to take so long to get back to you. To be honest I'm doubting the point in continuing this discussion, we don't seem to be progressing at all. From my side it seem like I raise points which directly address your claims, or rather those of Durston and Doug Axe, and you in turn simply repeat the claims and affirm that they are true because they are what Durston and Doug Axe said.

---

Hi WK. It doesn't matter that you posted a bit later. Since I also don't visit as often. Because I have more important things to do.As for Axe and Durston. I don't claim they're re right, because they say so, I just say what they wrote in their papers. That's all.

quote:

---

You are getting confused here between genetic regulation and evolution. These genes aren't expressed in the absence of a certain environmental trigger, but they are present in the genome with all of the genetic sequences required for their expression. These cells always had the ability to digest sugar, they simply weren't expressing those genes when there wasn't any sugar in the environment.

---

Umm... I know that, that was my point alla long. It's like a light switch. It turns specific genes ON or OFF. So, tell me, how many times do you have to flip a light switch before you make the switch into something new? Never, obviously... Therefore, such changes are not evidence for evolution.

quote:

---

No I dont. I'm just saying that there isn't any evidence that they can't. At the moment I don't think CSI is a clearly enough defined concept for this to be done.

---

Actually, simple math says it can't be done. It's like saying that 2+2=5. We all know that is not true. You can't get anything more of than what you invested in the first place. A deterministic natural law, by definition can not add CSI into a system. Dembski in his book No Free Lunch, explains why in 2 ways. One is a more simpler, and another, a more complex mathematical proof.Here I will show you the simpler way. If you wan't to I can later on show you the more complex explanation.First we have a CSI j, and a detrministic natural law denoted by f. Natural laws are describet as functions. Simply because they act on a certain variable, and than give the same result every time.Just like 2X + 10 = 20. X will always be 5. In the same way, when you put water under 0C, you will always get ice.So now, you are claiming that this natural law "f", brought about CSI "j", without intelligent cause. That means that there was some element "i" in the domain of "f", that was acted upon by "f" and it brought upon "j".This is represented by the equation => "f(i) = j"This actually does not create new information, since "i" will always produce "j" when acted upon by "f". This simply means that the natural law has shifted the same amount of information from "i" to "j". The problem of where did the CSI come from is not resolved by this. Simply because we have to ask where did CSI in "i" come from? Because that is the same CSI as in "j". It just got shifted around by "f" acting upon it.Now we have this equation: "I(A&B) = I(A) + I(B|A)", let's call it "*". It explains that information in an event A and B equal information in the event A together with information B given that A is certain. Which basicly means that if A happens, B is sure to happen. Therefore, if we see that A happened, that means B happened too.Let us now use this equation in our example. Since we already know that "i" fully determines "j", with respect to "f", that means that "I(j|i) = 0". This means that if know all the information in "i", we will also know all the information in "j", when "f" acts upon "i". Which means that if "i" happens, "j" also happens, and whatever we learn from "i" we also learn from "j". And this means that we can learn nothing more from "j" than from "i". Meaning, information gained is equal to zero.Which means that CSI that was generated is not created by a natural law, it was simply shifted from some other place. All natural laws act like this. Therefore natural laws are precluded from creating CSI. They can only shift them around.If you want I cal also go through a more complex explanation in teh next post.No Free Lunch: Why Specified Complexity Cannot Be Purchased Without Intelligence - William A. Dembski - Google KnjigeRead pages 151-152 for more... Oh, and you said CSI is not well defined. Please tell me why not?

quote:

---

Indeed, they measure it by an indirect proxy method which tells us very little about the range of possible functional sequences. Why when Hazen et al. demonstrate how it can be tied directly into actual measurable functions? Surely the functionality is the whole crux of this approach so I still fail to see why Durston et al. approached it in such a perfunctory manner.

---

Their method is not supposed to tell us anything about all the possible functional sequences for that particular function. That is why they rely on Axe's work instead.

quote:

---

You seem to be trying to slip into a discussion of abiogenesis here, which is totally beside the point.

---

No, I'm not. I'm simply saying, that who cares about a bunch of amino acids. They are useless on their own. None of them is more important than the other. I mean, important to whome?

quote:

---

No it isn't and I have described the differences in the conservation measures to you at least twice now.

---

The only thing I can say I understood from you, is that you simply disagree with Durston's approach, simply because his method does not tell us about all the possible functions while measuring. And no, it's not supposed to. But than again, Axe's work comes into play here, and all gets sorted out.

quote:

---

And I assume you know that that is a gigantic strawman and very far from being a cogent argument. The fact that you can't limitlessly mutate proteins and have them retain function does not mean that you can't substantially mutate them and retain function. In fact Doug Axe's own research showed that functionality could be maintained after 20% of a proteins surface residues had been altered.

---

I said "infinity" just to make a point. I know no sane person thinks you can mutate them to infinity. The point I was trying to make is that there is a limit. Which means not all sequences will work.

quote:

---

Except as Hazen and Szostak both emphasise, you need to effectively know all of the possible functional sequences for this to actually work, whereas Axe focuses onto a very narrow range of functional states by making the protein minimally functional to begin with.

---

Well, if we extrapolate from what Axe has done, than I think we have no problem in determining the number of possible sequences.

quote:

---

Because they are totally different proteins with totally different functions. Axe's work didn't even explore all the functional modificational possibilities of the protein he was studying.

---

If he mutated it beyond functionality than that is a good enough estimate. You can't get a totally right number down to one single sequence.

quote:

---

It may the the best that ID has but there is a wealth of comparative genetic data showing functionally similar proteins with lots of highly divergent genetic and amino acid sequences. I may post something to let us look at this in more depth at some other point.

---

I have no problem with other data. If they show that proteins are even more deformable, and can still perform their function, than that's fine with me. Simply because my point remains. And that is, that there is a limit. And if we can estimate the limit, that there is nothing to stop us from measuring the functional information in the genome.

quote:

---

There is a very large amount of space between a handful of possible sequences and an infinite number, you are excluding the middle pretty strongly here.

---

I know, I was simply trying to emphasize on the amount. Not to be taken literally.

quote:

---

You clearly don't understand my point. There are totally distinct amino acid structures that can perform equivalent functions. How can Doug Axe's approach ever identify these since as you say they are local modifications of a specific existing proteins structure?

---

If his estimate includes those sequences, than it can. Otherwise it can't. That's obvious.

quote:

---

That was never the point, as I have explained before. I didn't say they were using every possible method to inflate the FSC measure, I was pointing out that the fact that they assume that a PFAM alignment is a suitable proxy for all possible sequences with a given function will inflate the FSC measure.

---

No, that is just the conservation rate among all the proteins that are found. I can't see this inflating the FSC rate.

quote:

---

Well he certainly doesn't say so in his paper.

---

He is basing his work on Abel and Trevor's work, which I think you know about. The one where they talk about three subsets of sequence complexity.

quote:

---

This is arrant nonsense they are making direct measurements of biological function. To equate that to RSC is to essentially say, 'well sure you can evolve novel functions like that but they aren't real novel functions'.

---

You completely misunderstood me. I just wanted to say that RSC can measure anything, not that it is good enough to use as a measure of biological functions. Simply because it can not tell apart from biological functions, and simple natural sequences. It will always give you out some measure.

quote:

---

You tell me, you are the one who is saying that this is Durston's position. He avoids giving a usable definiton of biological function in his paper, as I complained before.

---

If I recal correctly. He is using a conservation rate. The same thing as you are using in other programs to measure functional sequence complexity. The only difference is that you think this one is, well, false...

quote:

---

From the various existing genetic sequences we are familiar with, the research into hypothetical ancestral sequences and their functions and from a growing body of research into the functionality of large scale artificially generated random and evolved sequences, i.e. SELEX.

---

I for one, would disagree on that simply because SELEX is actually not so random. The amount of intelligent input is to much to simply call it natural selection in action. As noted below.

quote:

---

No empirical data or rational scientific basis exists for such a metaphysical leap. Certainly no prediction of biological self-organization has been realized apart from SELEX-like bioengineering. SELEX is a selection/amplification methodology used in the engineering of new ribozymes [133-135]. Such investigator interference hardly qualifies as "self-organization." All of the impressive selection-amplification-derived ribozymes that have been engineered in the last fifteen years have been exercises in artificial selection, not natural selection.

---

NCBI

quote:

---

As I say we currently don't seem to be going anywhere, maybe instead of these mammoth omnibus post to each other we could focus on one specific point and actually discuss it in enough depth to reach some sort of conclusion. My personal favourite would be the question of how well we can estimate a total functional sequence space and how it aligns with Doug Axe's approach.

---

Long posts are just fine with me. The more we talk about, the more we can learn.

And when I bring up objections to what they wrote you just re-iterate it, that isn't how a debate is supposed to progress. If you don't think they are right then why continue to repeat their claims? If you do think they are right then why not present further supporting evidence? No one ever said they were. I said that mutations weren't like like light switches simply turning genes on and off, I wasn't saying anything about changes in gene regulation during an organisms life.Mutations certainly can spontaneously disrupt a gene rendering it null, but such genes tend to degenerate subsequently they don't just sit in the genome waiting to be turned back on. In line with this there is no evidence that there are novel functional genes hidden in the genome just waiting for the right regulatory switch to evolve to allow them to be expressed. Ho wwould such genes be maintained? This is part of your problem, you incorporate so many assumptions into what you consider to be 'natural laws' and ignore the importance of context. There are a number of conditions in which water will not produce ice at 0C, the obvious one is at high pressure at 13.35 Mega Pascals(MPa) the freezing point of water is -1C it is -21.985C at 209.9 MPa. I don't disagree with this. I think that the information is that of the complex environment in which the system is evolving. It is information about the environment as embodied by the environment. In this case the medium really is the message.The Because ID proponents can't agree on a clear usable definition or demonstrate how to calculate it? Because the calculations require several large assumptions or estimates of key probabilities the reliability of which make it essentially a guess? But Doug Ace's work doesn't do this, and certainly not for distinct genetic sequences with equivalent functions. This is meaningless nonsense. The physico chemical properties of specific amino acids are important to anyone who wants to understand biochemistry or protein evolution at any meaningful level. I pointed out that Durston et al.'s apporach is crude because it fails to take the functional similarities of certain amino acids into account and essentially treats them all as interchangeable. Given that there are already several extant methods which do take into account the functional/physicochemical similarities of the amino acids. This make Durston et al.'s approach a big step backwards in terms of investigating functionality. This is why I think some more focused posts would bve beneficial. We are discussing several different issues and you seem to be getting them all mixed up. The differences in measuring conservation between Durston et al's method, Blossum and the other method are a quite distinct point from the fact that Durston et al. assume a PFAM alignment is a suitable proxy for all possible sequences with a given function. No it doens't and just keeping on saying that it does won't change that. Sure Doug Axe gives them a figure to use, but unless that figure is actually meaningful then he might as well have just made it up. A point hardly worth making which you yet try to make several times. No one has ever argued otherwise, hence my pointing out that this is a strawman. You may have no problem in coming up with a number, but there is no reason to think that that number has any relevance outside of the specific protein Axe was studying (if that given the already existing mutations befor his work even started). Axe's work, as I have now pointed out multiple times, never addresses distinct structures with similar functions, it doesn't even address the full functional space of the protein it is focused on. No it isn't. If he had tried single nucleotide substitutions and found 1 single nucleotide substitution that rendered the protein completely non-functional would that be sufficient to extrapolate that all proteins must be 100% conserved to maintain functionality? If not then why is his slightly larger scale approach sufficient to make sweeping generalisations of the complete functional sequence space for every possible function? the answer is that it isn't. Perhaps we should discuss Axe's work on a separate thread. I agree with you here, although obviously we might disagree on exactly what the criteria of functional information is since we don't agree on what constitutes a usable concept of biological function. I think the main point of contention is that you think we can make these estimates using Doug Axe's approach and I don't, or at least I don't think we can from what he has produced so far, not even for the protein he was studying. Well it doesn't, there is no way for his work to address these sequences as it stands. So it appears you now agree with me that it is insufficient? Yes? But it isn't, the measure of FSC also implicitly factors in the functionality component and if you are only using a small fraction of all the possible functionally equivalent sequences then your measure of FSC is going to be inflated because to get an accurate measure you would need an alignment with all possible sequences that satisfy the functional criteria. Would you not agree that such a data set would constitute the ideal situation? Obviating the need for estimates such as Doug Axe's work provides? But that doesn't address the issue either. This still doesn't give any reason why we should discount demonstrable measurable functional outcomes, such as enzyme catalysis rates or binding affinity, as measures of biological function. You yourself were quite happy to put forward a loss of binding affinity as a change in biological function, albeit a loss, when we were discussing anti-biotic resistance. Then you were talking about something totally irrelevant to this discussion. And how surprising it is that you quote Abel and Trevor's agreement with your position . Except of course when they say ... "All of the impressive selection-amplification-derived ribozymes that have been engineered in the last fifteen years have been exercises in artificial selection, not natural selection.", a sentiment your argument echoes, they miss the point that for such selection to work the functional change must have been produced by the random mutations. No one has ever considered selection to be a random process. Natural selection is no more random than artificial selection, although obviously artificial salection is usually more focussed on one specific trait than natural selection so it may act as a more intense directional pressure.The point you miss is that I am not putting SELEX forward as a model for evolution but as an example of a system where you can almost fully explore the functional space of a sequence or even function and produce some real measures of FSC. If you wish to consider it a product of ID that is fine, it still provides a baseline method against which we can compare our attempts to measure FSC in real proteins. You'ld hope so, I still get the feeling that having 3 concurrent topics runnning in one set of posts is leading to some confusion.TTFN,WK

quote:

---

And when I bring up objections to what they wrote you just re-iterate it, that isn't how a debate is supposed to progress. If you don't think they are right then why continue to repeat their claims? If you do think they are right then why not present further supporting evidence?

---

What other supporting evidence? We have been through it all already.

quote:

---

No one ever said they were. I said that mutations weren't like like light switches simply turning genes on and off, I wasn't saying anything about changes in gene regulation during an organisms life.

---

Some do some don't. Some bases get marked so they can get removed or inputed in the sequence when the organims needs it.

quote:

---

Mutations certainly can spontaneously disrupt a gene rendering it null, but such genes tend to degenerate subsequently they don't just sit in the genome waiting to be turned back on. In line with this there is no evidence that there are novel functional genes hidden in the genome just waiting for the right regulatory switch to evolve to allow them to be expressed. Ho wwould such genes be maintained?

---

In no way. But that is basicly my position. I am the one who is claiming new functions don't evolve, not you.

quote:

---

This is part of your problem, you incorporate so many assumptions into what you consider to be 'natural laws' and ignore the importance of context. There are a number of conditions in which water will not produce ice at 0C, the obvious one is at high pressure at 13.35 Mega Pascals(MPa) the freezing point of water is -1C it is -21.985C at 209.9 MPa.

---

No, I do not have a problem, you have a problem. You are nitpicking.The reason I used that example is because it was easy to understand. I used it because of simplicity, to demonstrate determinism. To model a real natural law we would need millions of functions. I used just one to explain how determinism works. To model your example, we would just have to add another function. The result is the same in any case.

quote:

---

I don't disagree with this. I think that the information is that of the complex environment in which the system is evolving. It is information about the environment as embodied by the environment. In this case the medium really is the message.

---

Great, so we aggree, that natural forces do not create new information. They just transfer it to another place.

quote:

---

Because ID proponents can't agree on a clear usable definition or demonstrate how to calculate it? Because the calculations require several large assumptions or estimates of key probabilities the reliability of which make it essentially a guess?

---

An object exhibits specified complexity when it conforms to an independently givven pattern and it's probabillity of occurance is bellow the UPB.I see nothing wrong with this definition.

quote:

---

But Doug Ace's work doesn't do this, and certainly not for distinct genetic sequences with equivalent functions.

---

We are supposed to measure the sequences that can be aligned. Obviously we do not need to take into account distinct sequences. Their informational content is subsumed in the smaller ones.

quote:

---

This is meaningless nonsense. The physico chemical properties of specific amino acids are important to anyone who wants to understand biochemistry or protein evolution at any meaningful level.

---

No it's not. Like I said before. Amino acids by themselves, when they are not folded in a protein have no bearing on any biochemical investigation. If they did, than so do atoms in rocks. You do know that chemicals are made from atoms? So by your logic, we can find out something about life, if we study atoms all by themselves. No we can't. That's why amino acids by themselves are useless to biochemistry. Now chemistry is another story. It deals with chemicals by themselves...

quote:

---

I pointed out that Durston et al.'s apporach is crude because it fails to take the functional similarities of certain amino acids into account and essentially treats them all as interchangeable. Given that there are already several extant methods which do take into account the functional/physicochemical similarities of the amino acids. This make Durston et al.'s approach a big step backwards in terms of investigating functionality.

---

If there are, well than, more power to them! I'm certainly not going to deny them that. Besides I never said Durston's method is the best. I was simply arguing from the start only that it existed. That is all.

quote:

---

This is why I think some more focused posts would bve beneficial. We are discussing several different issues and you seem to be getting them all mixed up. The differences in measuring conservation between Durston et al's method, Blossum and the other method are a quite distinct point from the fact that Durston et al. assume a PFAM alignment is a suitable proxy for all possible sequences with a given function.

---

To measure the conservation rate yes. There is nothing more I would add to it.

quote:

---

No it doens't and just keeping on saying that it does won't change that. Sure Doug Axe gives them a figure to use, but unless that figure is actually meaningful then he might as well have just made it up.

---

But you and I both know that he did not made it up. If you disagree with it than fine. But he did not just make it up.

quote:

---

A point hardly worth making which you yet try to make several times. No one has ever argued otherwise, hence my pointing out that this is a strawman.

---

To ephasize the point. Don't take it literally next time.

quote:

---

You may have no problem in coming up with a number, but there is no reason to think that that number has any relevance outside of the specific protein Axe was studying (if that given the already existing mutations befor his work even started). Axe's work, as I have now pointed out multiple times, never addresses distinct structures with similar functions, it doesn't even address the full functional space of the protein it is focused on.

---

1.) He does not have to adress distinct structures, because they are in a field of their own. They should be calculated on their own. Not within this structure. The probability of a different structure is obviously different.2.) Are you saying that he started mutating the gene from the middle or something? This is probablity we are working here with. We can't be sure that a result is correct to a single digit.

quote:

---

No it isn't. If he had tried single nucleotide substitutions and found 1 single nucleotide substitution that rendered the protein completely non-functional would that be sufficient to extrapolate that all proteins must be 100% conserved to maintain functionality? If not then why is his slightly larger scale approach sufficient to make sweeping generalisations of the complete functional sequence space for every possible function? the answer is that it isn't. Perhaps we should discuss Axe's work on a separate thread.

---

Well no, obviously not. But he did more than that. He mutated the protein many times and concluded that it has some ability to adapt to mutations and still work. But not much.

quote:

---

I agree with you here, although obviously we might disagree on exactly what the criteria of functional information is since we don't agree on what constitutes a usable concept of biological function. I think the main point of contention is that you think we can make these estimates using Doug Axe's approach and I don't, or at least I don't think we can from what he has produced so far, not even for the protein he was studying.

---

I think that biological functions are what genes code for. That should be the most general definition. From what I can see I can find no flaw in Axe's work.

quote:

---

Well it doesn't, there is no way for his work to address these sequences as it stands. So it appears you now agree with me that it is insufficient? Yes?

---

If we wanted to measure totally different structures than yes. But we are not trying to do that. As I said above, totally different structures, have totally different probabilities of arising.

quote:

---

But it isn't, the measure of FSC also implicitly factors in the functionality component and if you are only using a small fraction of all the possible functionally equivalent sequences then your measure of FSC is going to be inflated because to get an accurate measure you would need an alignment with all possible sequences that satisfy the functional criteria. Would you not agree that such a data set would constitute the ideal situation? Obviating the need for estimates such as Doug Axe's work provides?

---

No, since we are measuring the functional information for this specific function, and this specific protein sequence. Yes, with modifications. The protein can be modified. And from this point we measure the functional information.I understand your concern. And that is that there are other, totally different structures, that can do this same function. That's true, but we are not measuring them, now are we. We are trying to measure the probability of this structure, not the completely new one, which has different probability.

quote:

---

But that doesn't address the issue either. This still doesn't give any reason why we should discount demonstrable measurable functional outcomes, such as enzyme catalysis rates or binding affinity, as measures of biological function. You yourself were quite happy to put forward a loss of binding affinity as a change in biological function, albeit a loss, when we were discussing anti-biotic resistance.

---

Well we could under a general definition of biological information. And the only reson I defined it as an reduction in teh first place is because of a loss of specificity.

quote:

---

And how surprising it is that you quote Abel and Trevor's agreement with your position . Except of course when they say ... "All of the impressive selection-amplification-derived ribozymes that have been engineered in the last fifteen years have been exercises in artificial selection, not natural selection.", a sentiment your argument echoes, they miss the point that for such selection to work the functional change must have been produced by the random mutations. No one has ever considered selection to be a random process. Natural selection is no more random than artificial selection, although obviously artificial salection is usually more focussed on one specific trait than natural selection so it may act as a more intense directional pressure.

---

Well you see, that's the whole point. Natural selection has a constraint set on fitness. And therefore, it does not selct for new biological functions, but for fitness. Fitness and new biological functions are not correlated. That is why natural selection is as good as random search when it comes to building new biological structures.It has no teleology and it can not select those individuals that it needs to be something new. On the other hand, artificial selection, is teleological. It adds information by intelligently selecting the individuals that it wants.

quote:

---

The point you miss is that I am not putting SELEX forward as a model for evolution but as an example of a system where you can almost fully explore the functional space of a sequence or even function and produce some real measures of FSC. If you wish to consider it a product of ID that is fine, it still provides a baseline method against which we can compare our attempts to measure FSC in real proteins.

---

If you want to propose a better method than Axe's, than that is just fine with me. I have no problem with better methods in anything. I'm jsut saying that Axe's method is good also. Not that it is the best.