|
Register | Sign In |
|
QuickSearch
Thread ▼ Details |
|
Thread Info
|
|
|
Author | Topic: Information Changes in DNA by logical Analysis | |||||||||||||||||||||||||||
NosyNed Member Posts: 9004 From: Canada Joined: |
This has come up more than once. I had forgotten about this analysis and think it deserves a thread all of it's own.
Here is Dr. Adequate's comment( modified somewhat for this thread) lifted from a site he links to:
quote: Does anyone have any issues with any of these? If so why? Edited by NosyNed, : No reason given.
|
|||||||||||||||||||||||||||
Adminnemooseus Administrator Posts: 3976 Joined: |
Thread copied here from the Information Changes in DNA by logical Analysis thread in the Proposed New Topics forum.
|
|||||||||||||||||||||||||||
Wounded King Member Posts: 4149 From: Cincinnati, Ohio, USA Joined: |
Personally I don't have any particular issues with these axioms although I don't see how 'no DNA' is a workable null string. Generally a randomly generated sequence is more commonly used as a null state in bioinformatics, though I can see how this may confuse things since a random string obviously has some information by a number of metrics.
I see no problem with axiom 1 myself but in a number of discussions Percy has used Shannon information and allele number in a way which suggests to me that you would have to qualify it to specify that it also requires the same environment. In Percy's 'new allele => new information' example, see [thread=-9419], the number of alternative alleles and the frequency of the alleles can all change the amount of Shannon information conveyed by sampling one specific allele from the population. Using this approach the information content in a string can be changed by changing the environment rather than the string itself, i.e. in Percy's example the de novo mutation creating a new allele means that the pre-existing allele sequences can communicate more information. TTFN, WK
|
|||||||||||||||||||||||||||
Dr Adequate Member (Idle past 312 days) Posts: 16113 Joined: |
Personally I don't have any particular issues with these axioms although I don't see how 'no DNA' is a workable null string. I don't see what else one could mean by a null string, though perhaps that's because I'm a mathematician. A null string is one not containing any letters, and is conventionally written as λ.
Generally a randomly generated sequence is more commonly used as a null state in bioinformatics, though I can see how this may confuse things since a random string obviously has some information by a number of metrics. Well, I think everyone would have to admit that λ contains no information by any sensible metric; which is why I introduced it into the argument.
|
|||||||||||||||||||||||||||
Wounded King Member Posts: 4149 From: Cincinnati, Ohio, USA Joined: |
Well, I think everyone would have to admit that λ contains no information by any sensible metric; which is why I introduced it into the argument. I agree, I just don't see how it could actually be useful to state this. What use is it to define this null string? TTFN, WK
|
|||||||||||||||||||||||||||
NosyNed Member Posts: 9004 From: Canada Joined: |
I see no problem with axiom 1 myself but in a number of discussions Percy has used Shannon information and allele number in a way which suggests to me that you would have to qualify it to specify that it also requires the same environment. In Percy's 'new allele => new information' example, see Evolving New Information, the number of alternative alleles and the frequency of the alleles can all change the amount of Shannon information conveyed by sampling one specific allele from the population. Using this approach the information content in a string can be changed by changing the environment rather than the string itself, i.e. in Percy's example the de novo mutation creating a new allele means that the pre-existing allele sequences can communicate more information. This might require the statement that the two strings are compared in the same environment to be checked for "identicalness" or not.
|
|||||||||||||||||||||||||||
Dr Adequate Member (Idle past 312 days) Posts: 16113 Joined: |
I agree, I just don't see how it could actually be useful to state this. Well, the axiom is used in the second part of the proof, where it's shown that mutations can't all conserve information.
What use is it to define this null string? What use is it to have a number for zero?
|
|||||||||||||||||||||||||||
Wounded King Member Posts: 4149 From: Cincinnati, Ohio, USA Joined: |
OK, looking at the SkepticWiki page I see how the argument goes and where the 'no DNA' state fits in.
I think this argument fails to address the common creationist approach that we see here quite often where they equate maximum information to conforming to a specific sequence, presumably the ideal pre-fall sequence or similar. Obviously with such an approach you can never have more information than the 'perfect' pre-fall state. You can see often this line of thinking in Smooth Operator's arguments. TTFN, WK
|
|||||||||||||||||||||||||||
slevesque Member (Idle past 4668 days) Posts: 1456 Joined: |
I have the impression this thread was done just for me
But, as I've asked on the other thread: Who has read Gitt's book on information ? I can officially say I haven't, but anyone willing to discuss this will have to answer this question. EDIT: This is just to know how much acquired knowledge we each individually have on the subject. Edited by slevesque, : No reason given.
|
|||||||||||||||||||||||||||
PaulK Member Posts: 17827 Joined: Member Rating: 2.3 |
I doubt that anyone has bothered to read Gitt's book. I've looked at some stuff on the web, and it looks as if Gitt information is completely irrelevant. (There's no good reason to think that there is any "Gitt information" in DNA, nor is here a good way to measure it).
(I should also add that it doesn't look as if Gitt has much understanding of information theory. )
|
|||||||||||||||||||||||||||
NosyNed Member Posts: 9004 From: Canada Joined: |
I have the impression this thread was done just for me What on Earth would give you that idea? Actually, triggered by your use of information maybe but since it keeps coming up we might as well have a focus for it.
But, as I've asked on the other thread: Who has read Gitt's book on information ? I can officially say I haven't, but anyone willing to discuss this will have to answer this question. Why worry about that yet? Let's have a look at the argument put forward here as it is very simple. From that we might arrive at things we'd want to see in Gitt's argument. For example: if Gitt's views mean anything either the argument put forward from the 3 axioms here is wrong (which we can determine without reference to Gitt) or Gitt disagrees with one of the three axioms. How about we review this argument first to see if it is wrong on it's own or not? Then, only if needed, we can go to Gitt to see what he says.
|
|||||||||||||||||||||||||||
Dr Adequate Member (Idle past 312 days) Posts: 16113 Joined: |
OK, looking at the SkepticWiki page I see how the argument goes and where the 'no DNA' state fits in. I think this argument fails to address the common creationist approach that we see here quite often where they equate maximum information to conforming to a specific sequence, presumably the ideal pre-fall sequence or similar. In the first place, that would not constitute an argument that mutations cannot increase information, merely that the net effect of mutations since creation can't have been to increase information. In the second place, and more importantly, that would be a vast act of petitio principii. Their argument goes: "Evolution requires an increase of information. Mutations can't increase information. Therefore evolution is impossible". Now, if to justify the second premise in that line of reasoning, they have to add, as a premise, the claim that species were brought into existence by an act of fiat creation by God, then this is fallacious. If we grant them that as a premise, then what further do we have to discuss? It is obviously illegitimate to argue for creationism based on the premise that creation happened. I might as well argue: "Evolution requires an increase of information. We know that all species are a product of evolution. Therefore mutations can increase information." But there would be something rather dubious, in the context of this argument, about my introducing the second statement as a premise.
|
|||||||||||||||||||||||||||
AnswersInGenitals Member (Idle past 179 days) Posts: 673 Joined: |
There are 3 axioms which must be considered first:
1. Two identical pieces of DNA contain the same amount of information.2. The "null string" consisting of no DNA contains no information. 3. Some strings of DNA contain information. Let me revise this slightly: There are 3 axioms which must be considered first:
1. Two identical pieces of _________ contain the same amount of information.2. The "null string" consisting of no ________ contains no information. 3. Some ____________ contain information, where you can fill in the blanks with whatever you want. Just to be my usual obnoxious self, let me insert "shit" into the blanks. I suspect some will think I am just being argumentative by using something that obviously contains no information, but in fact feces contain a great deal of information - doctors can tell a lot about your bowel health by examining your stools (feces, not chairs) and paleontologists get very exited when they discover fossilized coprolites. My point is that anything you use to fill the blanks will work, and will work as well as anything else. Even things that do not actually exist such as flying blue unicorns. Their nonexistence may be considered to provide as much information as would be had if they actually existed. I suggest that any definitions or set of assumptions (which is just a type of definition) that applies to everything in existence (and perhaps to everything in non-existence as well) is pretty useless. I have yet to see any definition of information that, upon close examination, doesn't prove to be pretty useless. To put this in more rigorous terms, the set of all things that work when used to fill in the above blanks is the Universal set, or equivalently, the set of all things that don't work is the Null set. Since the above set of assumptions/definition does not partition the Universal set beyond its trivial subsets, it is fallow. So, where do we go from here? I have no friggin' idea.
|
|||||||||||||||||||||||||||
Dr Adequate Member (Idle past 312 days) Posts: 16113 Joined: |
where you can fill in the blanks with whatever you want. Just to be my usual obnoxious self, let me insert "shit" into the blanks. * sighs gently, drums fingers gently on desk * You may, if you wish to. But to what then would correspond to the fact that for every mutation there is an inverse mutation? If you wish to stipulate that for every grishnakhah, where grishnakhah is defined as a change of the information in shit, there is an inverse grishnakhah ... and if you wished to stipulate that you were not talking about "information", but about xyryxzlu, so long as you are willing to state that xyryxzlu was some quantity to which could be assigned a real number greater than or equal to zero ... and if you wished to stipulate that grishnakhahs could reduce the quantity of shit ... and if you agreed that the quantity of shit could never be less then zero ... and if you would agree that the null shit consisting of no shit had no xyryxzlu, and if you agreed that some shit had a strictly positive value of xyryxzlu ... then I guess that I could prove that some grishnakhahs increased the xyryxzlu of shit. That's why it's a logical analysis. It doesn't matter what we think the words mean, what matters is the logical relations between them. I really thought you were smarter than this, what is your point? Edited by Dr Adequate, : No reason given.
|
|||||||||||||||||||||||||||
Percy Member Posts: 22502 From: New Hampshire Joined: Member Rating: 4.9 |
Wounded King writes: In Percy's 'new allele => new information' example, see Evolving New Information, the number of alternative alleles and the frequency of the alleles can all change the amount of Shannon information conveyed by sampling one specific allele from the population. Using this approach the information content in a string can be changed by changing the environment rather than the string itself, i.e. in Percy's example the de novo mutation creating a new allele means that the pre-existing allele sequences can communicate more information. That's an interesting way of looking at it. To put it in concrete terms via a simple example, let's say a population has alleles A, B and C for a certain gene, so the gene has a message set of size 3. The information conveyed when one of these genes is "sampled" (presumably this means the organism has just become a proud parent) is 1.585 bits. But now one organism in the population experiences a mutation in this gene, call it allele D, so now there are 4 alleles, a message set of size 4, for this gene. This means that when any individual in this population reproduces that the allele it contributes for this gene now conveys more than 1.585 bits. But as we've discussed before, the likelihood of communicating this new allele to the next generation is very low because only one individual in the population possesses it at this point, so this means that the other alleles communicate only slightly more than 1.585 bits. As the new allele propagates through the population in subsequent generations the amount of information communicated by one of these alleles will gradually rise to 2 bits, assuming equal probability for all alleles. But that's more complicated than I like to make it. I've never succeeded in getting even my simplified example across. --Percy
|
|
|
Do Nothing Button
Copyright 2001-2023 by EvC Forum, All Rights Reserved
Version 4.2
Innovative software from Qwixotic © 2024