Posted on 03/05/2002 9:45:44 PM PST by Southack
This is part two of the famous "Million Monkeys Typing On Keyboards for a Million Years Could Produce The Works of Shakespeare" - Debunked Mathematically.
For the Thread that inadvertently kicked started these mathematical discussions, Click Here
For the Original math thread, Click Here
How is it that functionally equivalent variations of a viable genome would not be viable?
In the same way that not all variations of re-ordering bits from one viable computer program would result in new working programs...
No. The right-handed double-stranded helix is the most common. There's a left-handed helix (called Z-DNA), fairly common. There are also triple helices (called H-DNA). And then there's large-scale (tertiary structure) folding which takes all sorts of shapes!
In the end, what you will discover is that only a tiny subset (comparably) of DNA sequences are viable, and the odds of getting one of those subsets either randomly or naturally (i.e., without intelligent intervention), are not good (as the math for the article for this thread demonstrates).
You are nitpicking. Granted, I understand your desperation to find something, anything in fact, that I say that is "wrong", but semantics aren't going to help rectify your blatant logical errors (such as those in all of your dishonest responses to Post #557). Obviously there are going to be technical discrepancies between what I say at a layman's level from what really needs to be said at a dilletante's and especially an expert's level.
In this particular instance, it isn't even important that there may be variations of double-helix shapes. What is important is that we differentiate between various DNA strands based not upon "shapes" but rather upon which life form they can construct, and of course the way that we tell what they can create is by examining the data stored in the DNA strand's sequence of bases.
Ah, then you agree I've produced what you asked for; now you move the goalposts.
In the end, what you will discover is that only a tiny subset (comparably) of DNA sequences are viable...
You can't possibly know this. I've already admitted I don't - too many unknowns and complexities. How does one classify a genome as valid? Valid WRT what environment? Genomes of other living things are part of the environment. And so on.
However, I did relate to you in a previous post that frame-shift mutation which is a reason why I think your intuition is likely to be wrong. You didn't respond.
Actually (and I'm sure N will correct me if I'm in error), the shape is often what counts in terms of activity at least for proteins.
Absolutely! I've pointed this out to Southack a number of times. For instance, in #606 "One very obvious way in [computer programming] isn't analogous is the secondary, 3^0, or 4^0 structures of DNA and the function inherent in those structures. This function is context dependent. For instance, if you drop a protein or DNA in an acid or a base, its function changes because of conformational changes, even though the sequence is still the same. Likewise, some proteins are dependent on the presence of certain heavy metal ions which induce the proper conformational structure.
"Shape" is not even "context," much less "data."
To differentiate between the DNA of various life forms, we examine the sequential data stored by the bases in each DNA strand. Neither the "shape" nor the "context" can tell us which life form is coded in any particular DNA, but the data can tell us that answer.
Correct. Shape is affected by context and is important for FUNCTION.
To differentiate between the DNA of various life forms, we examine the sequential data ...
Naturally. And to compare compare dentition between fossils we examine features of teeth.
There are many ways to compare different organisms. We happen to be able to manipulate DNA sequence pretty well. It's the easiest place to start. Unlike software programs, however, given DNA sequence we still can't predict function. Sure, some of the sequences have been figured out: ie this type of sequence gives the following structure, like an alpha-helix. From there, we still don't know what it does.
No. Data is something that you can print out on a page. Try putting the data for A, C, G and T in a computer and print out a life form, then you will have data that reproduces.
"Data" remains as an analogy for what is going on, not what is really going on.
To be that incorrect about something so basic as the nature of data, you must truly be out of your league in this discussion...
Really? How many statewide math competitions have *you* achieved first place in? And I'll bet you $10,000 that your SAT math score is not higher than mine. Now that we've gotten that out of the way...
In the context of the discussion, Nebullis STIPULATED that we were talking about NONFUNCTIONAL mutations, as to which no natural selection could occur. So, there is no basis for assuming that the mutation proliferates. It could, by a random drift process, but just as likely the mutation would be snuffed out randomly.
Of course.
Hence, until natural selection occurs with respect to the mutation, the a prior probability that the mutation exists in Generation X is still the original probablity it occurs at all. So it doesn't matter when the mutations 1 to N occur. The probability is still the product of the individual probabilities PI.
Okay, I'll concede that since you didn't define your probability values, "P(i)", I had presumed that you were speaking of the "single instance" probability, and not the "long-term probability" of the mutation occurring. This is a natural mistake, since creationists have long been in the habit of making that error and I've lost count of the number of times I've caught them at it.
However, if that's what you actually meant, your original conclusion needs more development before you can actually state with any confidence that the combination of mutations would be, in your words, "next to impossible".
First, you throw around "N" as if it's some unspecified large number. Certainly, for a sufficiently large "N" a given combination of mutation becomes highly unlikely. But you fail to demonstrate that the number "N" in the case of several neutral mutations leading to a useful combination is necessarily a large one. "Lucky" combinations of 2-10 mutations are hardly astronomically unlikely, even if combinations of 20-100 of them are.
Similarly, I think you overstimate how unlikely a given mutation is if you're truly using the "long-term probability" value for your calculation. That probability includes the chances of the mutation happening in any individual of a lord-knows-how-large species population, multiplied by lord-knows-how-many generations over time.
You have yet to establish that N is sufficiently large, or P sufficiently small, to support your "next to impossible" assertion.
Finally, your P1*P2*P3...*PN calculation (as well as most other invocations of it in Creationist literature) makes the classic fallacy of discounting how many *other* combinations of mutations (at *any* location in the genome) might have been equally useful. It's easy to look at any particular sequence of mutations after the fact and say, "wow, the odds of *that* particular combination of mutations occurring is extremely low!" And so it is -- if you turned back the clock and "re-ran" the evolution experiment for that species, *that* exact sequence of (eventually) beneficial mutations would be highly unlikely to happen again in exactly the same way.
*However*, what this overlooks is the fact that there are nearly countless possible mutations which would have conferred some advantage on the species had they occurred. The odds of evolution hitting on *some* combination of mutations which gave an improvement (over a large chunk of time) are a damned sight better than P1*P2*P3...*PN. You can't argue that evolutionary improvement is mathematically unlikely by only looking at one *particular* improvement and showing the odds against it to be large, you have to look at the odds of getting *any* of the *large* number of mutations (or combinations) which would have resulted in increased fitness.
Let's say you're playing poker with three of your friends. Shuffle your deck of cards. Now write down the sequence of the cards in the deck. Congratulations, the odds against you getting that exact sequence are 80,658,175,170,943,878,571,660,636,856,404,000,000,000,000,000,000,000,000,000,000,000,000 to one. Did a miracle just occur for something so insanely unlikely to occur? No, since *some* combination was going to happen, and they're *all* that individually unlikely. If you deal those cards and win the pot, was that a 80,658,175[etc.etc.] to 1 miracle occurrence? No, since *many* different deck orderings could have let you win as well -- the actual likelihood of you winning a hand is actually far greater than the "P1*P2*P3" math of "how likely is that *one* deck arrangement" would lead you to believe -- in fact, it's one in four. But to use standard creationist math, you'd get beaten up by your friends for "cheating", because that *one* particular deck arrangement was *insanely* unlikely to occur...
You can't look at one particular evolutionary outcome and calculate the odds against that *exact* thing happening, because it artificially ignores the odds of *any* beneficial mutation happening. More concretely, one in a trillion odds for a particular beneficial mutation/combination aren't ridiculously low if there are five trillion possible beneficial mutations, out of 100 trillion possible mutations. It just looks that way if someone focusses on the "particular" odds and not the "overall" odds.
Because detrimental mutations like having two heads strongly tend to be weeded out (by interfering with propagation), whereas neutral mutations (the sort being discussed) don't.
It's apples and oranges. Different forces are at work on each, and thus it's no surprise that the results will differ.
I even pointed that out in the part of my post which you quoted in your post, try actually reading it next time.
"Ok, I'll take a stab at "truthfully addressing post #557": It's a giant load of nonsense. It, again, performs the classic Creationist fallacy of challenging someone to explain a hypothetical "single step" process which, sadly, bears no real resemblance to any evolutionary claim or process." - Dan DAySigh. Actually, Post 557 is a post that uses a cryptography analogy to illustrate why an earlier claim that DOS had multiple paths to Windows was in error.
Yes it was. And it was fatally flawed, on several counts. One of the biggest was that you were challenging someone to come up with a single operator (cryptographic key) which could alter DOS code into Windows code in a single step (i.e., a single cryptographic encoding).
They were positing the possibility of a stepwise evolutionary path between DOS and Windows (i.e., a path that leads from one to the other through many workable intermediate forms via small incremental changes), you were inexplicably changing the subject entirely to the feasibility of randomly finding one gigantic leap from the first form to the final form, which *disallows* any "stepping stone" modifications like the other person was suggesting.
I'm amazed you even considered it a valid comparison.
"Because detrimental mutations like having two heads strongly tend to be weeded out (by interfering with propagation), whereas neutral mutations (the sort being discussed) don't." - Dan Day
Your point above merely confirms your original error. Mutations can not have a 100% expected probability of being passed on to offspring if it can be demonstrated that any mutations fail to be inherited, such as the example you cite above.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.