Apolipoprotein AI Mutations and Information

[ardipithecus]

Apolipoprotein AI Mutations and Information is a great new reasource.  It documents a clear case of a beneficial mutation in human populations.  It also debunks claims by Answers in Genesis that mutation is a loss of information and AiG's suggestion that the mutation is lethal when homozygous.  It also shows how AiG handles feedback....

 

 

[lucaspa]

Nice info, Ardipethecus. Notice that AiG plays the shell game when it comes to 'information'. They pick and choose what they will consider information and ignore anything else. Come to think about it, why are we surprised? That's the modus operendi for the professional creationists in general.

Becoming an antioxidant adds information because now there is information to perform a task that was absent before. Since AiG can't quantify information there is no way to tell whether the supposed loss of specificity involves more or less information than the gain of an entirely new activity.

 

[seebs]

What's really frustrating is that we have a good, solid, unambiguous definition of "information", which they won't use because mutations *obviously* add information. Thank Claude Shannon for this insight; it revolutionized several industries.

 

[Pete Harcoff]

I was always under the impression that Shannon information wasn't really applicable to genetics. Am I wrong?

 

[lucaspa]

Today at 01:42 PM seebs said this in Post #3

What's really frustrating is that we have a good, solid, unambiguous definition of "information", which they won't use because mutations *obviously* add information. Thank Claude Shannon for this insight; it revolutionized several industries.

But Shannon's definitions don't work in biology, seebs.  What "information" are we talking about? The sequence of DNA nucleotides? Sequence of amino acids in proteins? Or the tasks and functions the proteins do?

If the latter, which tasks?  Does an enzyme with low specificity and low rate for two catalytic reactions have more or less information than an enzyme with a specificity and high reaction rate for one reaction?

 

[seebs]

Well, that's the thing - with Shannon's definition, we find that mutations *almost always* add new information, but not necessarily new functionality.

The point is, this allows us to nail down the *actual* claim, which is "no new functionality", while dispelling the nonsense argument of "no new information". The only reason to try to call functionality "information" is to piggyback on a naive understanding of 2LoT, claiming that new information would be order from disorder... The entire idea of talking about "new information" is a red herring, and when we use a formal definition of information, this becomes clear.

 

[Jeptha]

“Nice info, Ardipethecus. Notice that AiG plays the shell game when it comes to 'information'. They pick and choose what they will consider information and ignore anything else. Come to think about it, why are we surprised? That's the modus operendi for the professional creationists in general.”

JEP: I resemble this. And, oh no. Another interesting topic. I’ve got so much going right now that I can’t get any work done in my studio, but I can’t resist this one because all of us in here seem to think so much alike, just from different perspectives.

“What's really frustrating is that we have a good, solid, unambiguous definition of "information", which they won't use because mutations *obviously* add information.”

JEP: Mutations do not add information, they change information. And what you guys are discussing as Shannon information could technically be a correct term. But I think a better terminology would be Shannon-Weaver entropy:

http://www.ex.ac.uk/Psychology/docs/courses/2012/lundy/shannon.html

Yes. This is another area of thermodynamics. Claude Shannon picked up on one of the father of thermodynamics statements, Ludwig Boltzmann of the 1890s, in that entropy is the opposite of information. Shannon then took this astute observation to new levels.

Shannon proposed mathematically that ‘loose’ information; that is, information that is not ‘fixed’ such as info on a CD or printed on a book or document, will degrade as it is passed around unless you add energy to the system to stabilize it. Information entropy will increase.

I used to have some fun with my kids when I taught this in high school. I would write down a paragraph on a piece of paper and let the first kid in the class read it. I then grabbed the note and made her whisper the message to the kid in back of her. This would go all around the class to the last kid who would then write the message on the board. The results were always hilarious because the final message would be so degraded as to not even resemble the original. A funny thing would also occur. The final message would be longer. This seems to be an attempt for the information to use more words in attempt to communicate the correct original.

Shannon called this phenomenon the addition of ‘noise.’

I discovered a perfect example of this on the net. It’s at altavista.com and is their translator. I’m going to steal SEEBS short post and pass this information around through a few different languages. If Shannon was correct, and since this is ‘loose’ information, this information should degrade rapidly as it is passed around increasing informational entropy. Here is the original post: “What's really frustrating is that we have a good, solid, unambiguous definition of "information", which they won't use because mutations *obviously* add information. Thank Claude Shannon for this insight; it revolutionized several industries.”

Here is English to Spanish, back to English, then English to French, French to German; and finally from the German back into English. Here is what we now have: “définition nous de bonne d'une d'avons de que d'est de frustre de qui de vraiment de le los?What, solidly, d'"information de clair ", mutation de les de que de parce de pas de n'emploieront de qu'ils * évidentes * l'information ajoutent. Est par de pénétration de cette de Claude Shannon reconnaissant; IL industries de plusieurs d'un révolutionné?”

Yes, this is actually English. But informational entropy has so increased that it is now just garbled non-sense. Informational entropy is at maximum as this no longer makes a lick of sense.

Now to genes. As this genetic information randomly mutates through the generations and genetic information is passed around, this information will tend to degrade. It finally so degrades that the genome is so disorganized it can no longer keep the organism alive. The organism then enters mutational melt down and goes into extinction. So much for evolution over time. Try devolution. Because that is what, in reality, actually happens. And we went from 33 words to 55 words. There’s your Shannon noise.

 

[Frumious Bandersnatch]

Kolmogorov-Chaitin information theory may provide a better measure of "information" relative to evolution. Shannon's theory of communication deals with the sending of messages. A message in the Shannon sense can only get noisy. There is no mechanism for duplicating the message and selecting for new traits among the duplicates based on the ability of the traits to further duplicate the modified message so I think it is problematic to apply Shannon entropy measure to evolution.

Mark Chu-Carroll had an interesting POM on this topic

http://www.talkorigins.org/origins/postmonth/feb01.html

Tom Schneider, on the other hand, takes the approach of using information as defined in Shannon's Channel Capacity theorum and adds duplication and selection to generate new information, so I don't think the issue of which type of information to use is really settled.

http://www.lecb.ncifcrf.gov/~toms/paper/ev/latex/index.html

The Frumious Bandersnatch