NIH: 100M Years to Change a Binding Site

[DavisBJ]

Your rules. Evolutionary rules. Under information theory, random changes are always bad for information. Now, since you seem to be "conversant with the exact definitions and rules applicable to information theory" enough to test me on the subject, perhaps you can answer the question. Can you tell us why information theory does not apply to the data from DNA.

It probably does apply. Just do it correctly. Part of the DNA message in a brown-fur boder is “Here’s what pigment goes in the fur to make it brown.” Part of the DNA message in a white-fur boder is “Here’s what pigment goes in the fur to make it white.” Like you say, “Brown to white and white to brown is irrelevant”. The DNA message in the white fur boder is every bit as meaningful, and perhaps as efficient, as the DNA message in the brown-fur boder. They are different messages, because they have different jobs (make white fur vs brown fur) and pertain to different situations (forest living vs snow living). That difference proves nothing about any “costs” involved, and it would not matter whether the white furred boders were the originals, or the brown-furred ones were.

 

[Stripe]

It probably does apply. Just do it correctly. Part of the DNA message in a brown-fur boder is “Here’s what pigment goes in the fur to make it brown.”

If you want to do it properly, you're going to have to look at the information. Fur colour is not the information.

The DNA message in the white fur boder is every bit as meaningful, and perhaps as efficient, as the DNA message in the brown-fur boder. They are different messages, because they have different jobs (make white fur vs brown fur) and pertain to different situations (forest living vs snow living).

And this you've just made up contrary to the rules governing information theory.

 

[DavisBJ]

If you want to do it properly, you're going to have to look at the information. Fur colour is not the information.

If you would read a bit more carefully, I said the message was how to make the fur color, not the fur itself.

Let me try to itemize step by step:

1 - If we start with brown-fur boders, and a random DNA change gives them white fur, you say that will necessarily cause a degradation in their genome. That is, the genome of white-fur boders will be degraded from that of the brown-fur boders they mutated from.

2 - On the example when Yorzhik chimed (post 247) in he said if the "white fur" DNA returned to “brown fur” DNA it would bring them back to where they started – no genome degradation. To that you said “Exactly” (post 250).

3 - Now consider a slightly modified scenario in which there had never been brown fur boders, only white ones, with DNA and a genome identical to the white ones mentioned in step 2 above. If they experience the DNA change that gives brown fur, their biological makeup will be identical to what the original brown-furred ones would have been.

4 – In this modified scenario, since boders started with white fur DNA, and there was a random change to it which resulted in brown fur, you say information theory dictates that there is a genome degradation.

5 – In both cases we end up with genetically identical brown fur boders, yet the second set of brown-fur boders have undergone genome degradation as a result of the random change to their original white-fur DNA. But as shown in step 1, white fur boders already have a degraded genome.

6 – How can the degradation of already degraded white fur boders (case 2) end up with the same genome integrity as boders which have no genome degradation (case 1)?

 

[Stripe]

If you would read a bit more carefully, I said the message was how to make the fur color, not the fur itself.

And you're ignoring the fact that changes to the information on how to make fur undoubtedly also make up the message on how to make other things.

1 - If we start with brown-fur boders, and a random DNA change gives them white fur, you say that will necessarily cause a degradation in their genome. That is, the genome of white-fur boders will be degraded from that of the brown-fur boders they mutated from.

Yip.

2 - On the example when Yorzhik chimed (post 247) in he said if the "white fur" DNA returned to “brown fur” DNA it would bring them back to where they started – no genome degradation. To that you said “Exactly” (post 250).

He said this granting you the assumption that there was no other cost. But there must always be a cost. And even without the cost your net gain was zero.

Changing your story from there will mean Y. has every right to now deny you that assumption.

3 - Now consider a slightly modified scenario in which there had never been brown fur boders, only white ones, with DNA and a genome identical to the white ones mentioned in step 2 above. If they experience the DNA change that gives brown fur, their biological makeup will be identical to what the original brown-furred ones would have been. 4 – In this modified scenario, since boders started with white fur DNA, and there was a random change to it which resulted in brown fur, you say information theory dictates that there is a genome degradation.

Yip.

5 – In both cases we end up with genetically identical brown fur boders, yet the second set of brown-fur boders have undergone genome degradation as a result of the random change to their original white-fur DNA. But as shown in step 1, white fur boders already have a degraded genome. 6 – How can the degradation of already degraded white fur boders (case 2) end up with the same genome integrity as boders which have no genome degradation (case 1)?

Because you're ignoring the rules governing information theory and assuming the truth of an evolutionary explanation.

 

[Yorzhik]

If you want to do it properly, you're going to have to look at the information. Fur colour is not the information.

If you would read a bit more carefully, I said the message was how to make the fur color, not the fur itself.

Your response doesn't seem to make sense. What does Stripe's quote say about the fur itself? Were you confused by the British spelling of "colour"?

Let me try to itemize step by step:

1 - If we start with brown-fur boders, and a random DNA change gives them white fur, you say that will necessarily cause a degradation in their genome. That is, the genome of white-fur boders will be degraded from that of the brown-fur boders they mutated from.

2 - On the example when Yorzhik chimed (post 247) in he said if the "white fur" DNA returned to “brown fur” DNA it would bring them back to where they started – no genome degradation. To that you said “Exactly” (post 250).

3 - Now consider a slightly modified scenario in which there had never been brown fur boders, only white ones, with DNA and a genome identical to the white ones mentioned in step 2 above. If they experience the DNA change [assuming "DNA change" means "random mutation"] that gives brown fur, their biological makeup will be identical to what the original brown-furred ones would have been.

4 – In this modified scenario, since boders started with white fur DNA, and there was a random change to it which resulted in brown fur, you say information theory dictates that there is a genome degradation.

5 – In both cases we end up with genetically identical brown fur boders, yet the second set of brown-fur boders have undergone genome degradation as a result of the random change to their original white-fur DNA. But as shown in step 1, white fur boders already have a degraded genome.

6 – How can the degradation of already degraded white fur boders (case 2) end up with the same genome integrity as boders which have no genome degradation (case 1)?

Because the information content isn't measured by the sender, it is, in effect, measured by the receiver. Thus, the receiver is expecting the original message, and any noise picked up in transmission is a degrading of the information. The conclusion of this should be clear to your example, but if you can't figure it out I'll try to explain further.

 

[Stripe]

the British spelling of "colour"?

New Zealand spelling! :IA:

 

[Yorzhik]

New Zealand spelling! :IA:

I'm a stupid American, how would I know? :idunno:

 

[Stripe]

I'm a stupid American, how would I know? :idunno:

:chuckle:

It's all in the pronounciation.

 

[Alate_One]

So, since Stripe and Yorz have this DNA as information stuff all figured out, maybe they can tell me which of these two segments of DNA have more information?

MWN

1 atgcccatgc aggagcccca gaggaggcta ctgggtcctt tcaactccac ccgcacaggc 61 gttcctcacc tcgagctatc tgccaaccag actggaccct ggtgcctgca cgtatccatc 121 ccagatggcc tcttcctcag cctggggctg gtgagcttgg tggaaaatgt gctggtggtg 181 atttccattg ccaagaacca aaacctgcat tcccccatgt actacttcat ctgctgcctg 241 gctttgtctg acctgcttgt gagtgtgagc attgtgctgg agaccactct catcttggtg 301 ctagaggcag gggccctggc cacccgggtg actgtggtac agcagctgga caatgtcatc 361 gacgtgctca tctgtggctc catggtctca agtctgtgct tcctcggagc catcgctgtg 421 gaccggtaca tctccatctt ctatgcactg cgctatcaca gtattgtgac actgccccgg 481 gctcggtggg ccatcgtggc catctgggta gccagcatct cttccagcac tctttttgtt 541 gcctactaca accacacagc ggtcctgctt tgtctcgtca ccttttttct agccacgctg 601 gcactcatgg tagttctgta tgtgcacatg cttgcacggg cacaccagca tgctcaggcc 661 attgctcagc tccacaagag acagcacctt gtccaccaag gtttccgact caaaggcgcg 721 gccaccctca ctatcctctt gggcattttc ttcctgtgct ggggcccctt cttcctgtac 781 ctcactctca ttgtcctctg cccgaagcac cctacctgtg gctgtttctt caagaacctc 841 aatctcttcc ttgccctcat catcttcaac tccattgttg accccctcat ctatgccttc 901 cgaagtcagg agctccgcat gacgctcaag gaggtgctgc tgtgctcctg gtga

HEH589

1 atgcccatgc aggagcccca gaggaggcta ctgggtcctt tcaactccac ccgcacaggc 61 gttcctcacc tcgagctatc tgccaaccag actggaccct ggtgcctgca cgtatccatc 121 ccagatggcc tcttcctcag cctggggctg gtgagcttgg tggaaaatgt gctggtggtg 181 atttccattg ccaagaacca aaacytgcat tcccccatgt actacttcat ctgctgcctg 241 gctttgtctg acctgcttgt gagtgtgagc attgtgctgg agaccactct catcttggtg 301 ctagaggcag gggccctggc cacccgggtg actgtggtac agcagctgga caatgtcatc 361 gacgtgctca tmtgtggctc catggtctca agtctgtgct tcctcggagc catcgctgtg 421 gaccggtaca tctccatctt ctatgcactg cgctatcaca gtattgtgac actgccccgg 481 gctcggtggg ccatcgtggc catctgggta gccagcatct cttccagcac tctttttgtt 541 gcctactaca accacacagc ggtcctgctt tgtctcgtca ccttttttct agccacgctg 601 gcactcatgg tagttctgta tgtgcacatg cttgcacggg cacacmagca tgctcaggcc 661 attgctcagc tccacaagag acagcacctt gtccaccaag gtttccgact caaaggcgct 721 gccaccctca ctatcctctt gggcattttc ttcctgtgct ggggcccctt cttcctgtac 781 ctcactctca tcgtcctctg cccgaagcac cctacctgcg gctgtttctt caagaacctc 841 aatctcttcc ttgccctcat catcttcaac tccattgttg accccctcat ctatgccttc 901 cgaagtcagg agctccgcat gacgctcaag gaggtgctgc tgtgctcctg gtga

It should be really easy, right?

 

[Stripe]

So, since Stripe and Yorz have this DNA as information stuff all figured out, maybe they can tell me which of these two segments of DNA have more information? It should be really easy, right?

Why do you think it should be easy? It's easy to understand the principle. I've presented a couple of easy to understand cases that have been pointedly ignored. But to do the maths takes a bit more effort. And this process I have also explained.

I'm guessing that these two strings of data code for the exact same protien. This fact is something you're going to use to proclaim them identical. But as I've also explained before, it is a waste of time to look at a trait produced by the information. Whether it is fur colour or protien type you are wasting your time.

If these two DNA strings are different in their ATCG form, then they will indeed have different information values.

Now that the facts have been explained to you numerous times, Alate, perhaps you'd like to take another shot at answering the question. Why is it that information theory cannot be applied to data from DNA?

 

[Alate_One]

Why do you think it should be easy? It's easy to understand the principle. I've presented a couple of easy to understand cases that have been pointedly ignored. But to do the maths takes a lot of effort. And this process I have also explained.

So do the math then. You always ask everyone else to do all the work. How about you do some yourself?

I'm guessing that these two strings of data code for the exact same protien. This fact is something you're going to use to proclaim them identical. But as I've also explained before, it is a waste of time to look at a trait produced by the information. Whether it is fur colour or protien type you are wasting your time.

They're not identical and it relates back to the fur color discussion.

If these two DNA strings are different in their ATCG form, then they will indeed have different information values.

So if they have different information content, how do you tell which one has more? Does a few differences mean more difference in information content?

Now that the facts have been explained to you numerous times, Alate, perhaps you'd like to take another shot at answering the question. Why is it that information theory cannot be applied to data from DNA?

If you can't tell me which of these two very short sequences have more information, why should I think you have anything useful to say about DNA and information theory?

 

[Stripe]

So do the math then.

I have not the inclination at this time.

They're not identical and it relates back to the fur color discussion.

Ok.

So if they have different information content, how do you tell which one has more?

Find out which one is generally more compressible.

Does a few differences mean more difference in information content?

Yes.

If you can't tell me which of these two very short sequences have more information, why should I think you have anything useful to say about DNA and information theory?

Well there's a stupid criteria upon which to assess a man. Why don't you answer the real question. Why can information theory not be applied to data from DNA? Or are you willing to allow that information theory can be applied, even though I am only prepared to explain how it can be applied rather than actually doing an example.

 

[Alate_One]

I have not the inclination at this time.

I have not the inclination to believe your story then.

Find out which one is generally more compressible.

Compressible?

Yes.

So a piece of DNA that had 100 changes should have a lot more or a lot less information than a piece of DNA with one change?

Well there's a stupid criteria upon which to assess a man. Why don't you answer the real question. Why can information theory not be applied to data from DNA? Or are you willing to allow that information theory can be applied, even though I am only prepared to explain how it can be applied rather than actually doing an example.

Stupid criteria is you showing that you can do what you're claiming you can do? :rotfl: Stripe "just so" stories don't count for anything in my book.

 

[Frayed Knot]

Find out which one is generally more compressible. Yes.

You know, the trouble with arguing dishonestly is that you have to keep track of everything you say very carefully, because you're liable to contradict a major point you were asserting earlier.

Back in this very thread, you were telling me that the term "information" cannot be used to describe the source of data itself, that this term only applies to reduction in uncertainty that the receiver of the information infers. You adamantly denied that the term could be applied to the source of data, and said that it's ridiculous to think that information is a measure of the entropy, the complexity, of the source data.

But here you're saying to Alate_One that information content of DNA can be determined by its compressibility, which is determined by its complexity, its entropy.

So, you thought that idea was ridiculous?

Stripe, ridicule thyself.

 

[Stripe]

I have not the inclination to believe your story then.

Suit yourself. :idunno:

You know, the trouble with arguing dishonestly is that you have to keep track of everything you say very carefully, because you're liable to contradict a major point you were asserting earlier.

Lucky I'm not the one being dishonest then, eh?

Back in this very thread, you were telling me that the term "information" cannot be used to describe the source of data itself, that this term only applies to reduction in uncertainty that the receiver of the information infers. You adamantly denied that the term could be applied to the source of data, and said that it's ridiculous to think that information is a measure of the entropy, the complexity, of the source data.

Yep.

But here you're saying to Alate_One that information content of DNA can be determined by its compressibility, which is determined by its complexity, its entropy.

Not at all! She asked me to calculate which of the two strings had more information. In order to do that I have to send the data and measure the reduction in uncertainty at the other end. In order to determine the difference we could just send them both as-is which would mean the longer one has greater reduction and more information. A better analysis would be to determine some effective compression system and measure the two after sending the compressed data*.

Information is always a measure of the reduction in uncertainty at the receiver.

And I thought you weren't talking to me. :idunno:

*DNA is probably not very compressible.

 

[DavisBJ]

Because the information content isn't measured by the sender, it is, in effect, measured by the receiver. Thus, the receiver is expecting the original message, and any noise picked up in transmission is a degrading of the information. The conclusion of this should be clear to your example, but if you can't figure it out I'll try to explain further.

I am OK with the receiver being the one measuring the information content. If that message was supposed to say “brown fur” but was corrupted to say “white fur”, is this corruption detrimental or beneficial to the snow-bound boder?

 

[Dr.Watson]

You know, the trouble with arguing dishonestly is that you have to keep track of everything you say very carefully, because you're liable to contradict a major point you were asserting earlier.

And he does so quite often. The thing is, when you catch him at it, he insists the error is with everyone else's comprehension and not what he plainly said. He's a joke that is better off ignored.

 

[Frayed Knot]

And he does so quite often. The thing is, when you catch him at it, he insists the error is with everyone else's comprehension and not what he plainly said. He's a joke that is better off ignored.

I did say a few days ago that I wasn't going to waste my time addressing him, and I didn't stick to that, and he caught me. What do they say about the spirit being willing and all that? The temptation was just too great to point out how he directly contradicted himself on fundamental points within this very thread.

He said that you can't talk about information content with respect to a set of data, that information only applies to a measure of how well you can transmit data. Then he tried unsuccessfully to substitute "layman's information" instead of any formally-defined idea of information, but was unable to be specific about what he meant. Then he ridiculed the idea that increasing the complexity of a set of data (its entropy) means that its information content is higher.

Now he says that compressibility is a way to measure the information content in a set of data - something less compressible would have more information. But we already know that the more entropy a data set has, the less compressible it is! That's one of the ways of measuring entropy - the size (in bits) of a maximally-compressed representation of the data set.

Oh noes! Stripe was caught contradicting himself!

 

[Frayed Knot]

I have no idea if information theory can be applied to DNA data. Unlike you, Pastor Bob etc. I recognize when I am beyond my technical expertise.

Information theory can indeed be applied to DNA code. And a proper understanding of it produces no problems w/ respect to the Theory of Evolution. The thing is, Stripe is DOIN' IT RONG.

 

[Stripe]

Information theory can indeed be applied to DNA code.

Great! An answer to the question. :thumb:

I wonder if you can explain to BJ and Alate how a random change can realise a reduction of uncertainty at the receiver because every example I've posted always increases the uncertainty.