Single-step mutation - Increase of information

A very good program could still be outperformed by chance mutation.

Category:

Science & Technology

Tags:

• information
• increase
• random
• mutations
• evolution
• design
• criticism
• program
• example

[schrodcat]

"Quantity IS defined by the number of bits."

Oh well, then that's your definition of info. I still consider the 1940's to have led us a bit further than that, but I've got no personal problem with it. I take it that the bits representing 1/3, which are infinite, for you contain the same information of the bits of pi, equally infinite. And an amoeba has enormous info more than a mammal.

"point mutations are not one of those mechanisms."

Since our definitions differ, that'd go without saying.

[zebraone100]

"Information isn't just merely the number of bits. For example there can be redundant sub-sequences or poor (low-efficiency) encoding."

An increase in information is about the quantity of info, not the quality. Quantity IS defined by the number of bits.

"Gene duplication isn't the only mechanism by which information can be increased."

True, however, point mutations are not one of those mechanisms.

[schrodcat]

Information isn't just merely the number of bits. For example there can be redundant sub-sequences or poor (low-efficiency) encoding.

The amoeba has more than 220 times the base pairs of a human, yet it's hardly more complex an organism, mainly because of interplay between genes and between genes and alleles.

Gene duplication isn't the only mechanism by which information can be increased.

[zebraone100]

"...the purported impossibility that a mutation can be beneficial..."

Not even creationist claim this. Your proving something that NOBODY disagrees with.

Their argument is that single point mutations only allow for a finite limit on complexity as they can only CHANGE the info that exists. ex. if you have 8 bits, there are only 256 possible configurations.

They are wrong in that they ignored mutations that increase information, ie gene duplication, 8 bits->16 bits (65536 possibilities)

[schrodcat]

"if the mutation went from X to +, that would actually be a loss, which is the basis of a really ridiculous creationist claim"

Time for videos is limited. The core claim I was thinking about was the purported impossibility that a mutation can be beneficial, if the organism is already fully functional. Of course, mutation anywhere else (not only + from X) can be detrimental, more often than beneficial. But they wouldn't be selected.

I was concentrating on the false claim of "impossibility".

[zebraone100]

"However, an ALREADY complex algorithm can become better in 1 step."

If that is your goal, then the real confusion comes from the title of your video. It should have been something like "Single point mutation improves system function." or something like that.

Of course, if the mutation went from X to +, that would actually be a loss, which is the basis of a really ridiculous creationist claim. I would take the extra time to show why the negative mutation would probably disappear in time.

[schrodcat]

"It would be best to show that."

I'll think about it. I'm still unsure that'd be the key. Presentation IS obscure though. All I had in mind was literally a computer program, with a monkey accidentally typing X where there's a + (with language granted). The "ornaments" came later, and might have been confusing.

"going from nothing (random) to something in biology in one step just doesn't happen"

That was never questioned. However, an ALREADY complex algorithm can become better in 1 step. - ->+

[zebraone100]

I chose addition and subtraction because it isn't a stretch to see a minor modification changing the sign of the operation. Also, each step in my process represents a future generation, and "ab", "ac", etc represent large sequences of DNA. b and c are obviously closely related.

going from nothing (random) to something in biology in one step just doesn't happen. How I laid it out, gene duplication, then single point mutation is how it actually happens. It would be best to show that.

[schrodcat]

"(no random)"

Another sign that I misfired. Random chances where only meant as a reference model for probabilities, e.g. valid for an organism devoid of the algorithm altogether.

Re: your 4 point process. Actually the only difference I see is in the number of steps. Conceptually is the same. You just need to allow the output to be the next new bit of the chain. In fact, I kind of disproved one step isn't enough (not literally true, multiplication is not one algorithmic step away from addition)

[zebraone100]

For your rewrite, I suggest the same model but allow the mutations to be only addition or subtraction. (no random)

a= the rest of the model

b= subtraction

c= addition

1. start with "ab" and show its inefficiency.

2. point mutate to "ac", showing improvement

3. gene duplicate "acc", showing no change

4. point mutate to "abc", showing 100% efficiency, because now it is evaluating both the sum and difference.

Its not one step, but the real process isn't one step either.

??s, PM me.