Punctuated Equilibria theories

[Biochemist]

Originally Posted by Buffy

Bio, you are arguing again that changes must be "preferred" in order to survive. As Bumab points out and as I think I was saying too, changes *do not* have to be selected for immediately, rather as I understand it the mechanisms for self correction *remove* the dangerous changes but *leave* the good or *indifferent* changes. This process does in fact make the probability computations that you and Dembski use completely irrelevant

Oh come on. Given that you are now wandering into wild postulation, exactly how do the "mechanisms for self correction" identify which changes are going to be "dangerous" when there is no possibly way to know what the phenotypic expression will be?

Wouldn't that make the VAST majority of the code non-functional? Aren't you saying that the "mechanism for self correction" knows about danger and value in advance? Based on what? If you say that the genetic propensity was prespecified, you are now agreeing with me.

Tell me how the Dembski math does not apply here? What number of base sequences would be found to be "not dangerous"? If you contend that the number of potentially-useful proto-genes is anything more that one millionth of one percent of the code in higher phyla, this is again the triumph of hope over reason.

It certainly does not explain the fact that even in humans 2% of the code actually transcribes into proteins, and more of the non-coding portions are found to be required every year.

Buff- This position is nonsensical.

... And its the *indifferent* changes, driven by mutation, that cause morphological changes over time. ...

Then suggest why non functional DNA not 99.99999999 % of our genome. It is less than 98%. Not counting the ever increasing intron/functional but non-coding DNA load. OK make it 95%. Let's be gracious. You are saying that the "mechanisms for self correction" can get the accurate retention of proto-genetic DNA down to odds of less than one in twenty?? Those "self correction" mechanisms must be awfully accurate. Maybe that's because they (and their target genetic specification) were part of the genome.

SUGGEST SOME OTHER MECHANISM FOR THIS!!!!

Why you don't see this is lost on me, and in my book if you understand that you've put the cart before the horse on the issue in the first two sentances of this post, then you can't help but come around to the notion that random mutations can and do move things forward...

Last time. There is not a shred of evidence that this is true. And there is a cascade of evidence that genetic change come through genetically defined mechanisms.

[Buffy]

Originally Posted by Biochemist

Oh come on. Given that you are now wandering into wild postulation, exactly how do the "mechanisms for self correction" identify which changes are going to be "dangerous" when there is no possibly way to know what the phenotypic expression will be?

Now you're saying that these mechanisms don't exist? You said they do above. There's lots of documentation that the mechanisms for removal of mutations exist. If you're going to argue, at least don't contradict yourself!

Originally Posted by Biochemist

Wouldn't that make the VAST majority of the code non-functional? Aren't you saying that the "mechanism for self correction" knows about danger and value in advance?

Actually, the vast majority of code *is* non-functional! But you're contradicting yourself, and putting words in my mouth (and your's too, because you agreed that there are self-correction mechanisms earlier! really! you did!). This self correction does not have to "know" about danger, in fact it may not be dangerous at all, its just that over time the RNA and proteins have been selected that will zap a certain sequence.

Again, the main problem with your argument is that you insist that this selection process is inclusive rather than exclusive: if you simply see that yes its unlikely that this self correction process doesn't protect "good" but rather just gets rid of "selected bad", all of a sudden lots of mutations escape being excised, giving themselves opportunity to be expressed.

Originally Posted by Biochemist

Tell me how the Dembski math does not apply here? What number of base sequences would be found to be "not dangerous"? If you contend that the number of potentially-useful proto-genes is anything more that one millionth of one percent of the code in higher phyla, this is again the triumph of hope over reason.

First its not that they are "found not dangerous", they are not *known* to be dangerous: the self correction mechanisms have been selected over time because they *do* get rid of proven bad stuff (proven by selection!). Secondly, this whole quote proves you have no idea of why any mechanism that selects has a dramatic impact on probabilities. It turns unbelieveably remote possibilities in to near-certainties. You really need to read up on this. Further, you are stuck on this notion that "whole sequences are constructed randomly" when you should know that that is not what's in those introns: they are in most cases not just random sequences of neucleotides, they are in most cases copies of sequences that at one time were exons themselves, and a single change to one neucleotide pair in that sequence could make it useless or useful and ready to be expressed. If its useful, the organism that has it will survive, otherwise it won't.

Originally Posted by Biochemist

Then suggest why non functional DNA not 99.99999999 % of our genome. It is less than 98%. .... Those "self correction" mechanisms must be awfully accurate. Maybe that's because they (and their target genetic specification) were part of the genome.

Well, again, your math here is meaningless. All that has to happen is that the mechanisms get rid of a lot of stuff, but either of those numbers indicates there's a huge playground to work in. So what? And these correction mechanisms have evolved over time because they're selected for by protecting organisms and allowing them to succeed, which I guess to you is much more logical in saying they were programmed in from the very beginning....You wanna try to figure the odds that all this stuff was *bam* programmed perfectly in the first prokaryote?

Originally Posted by Biochemist

Last time. There is not a shred of evidence that this is true. And there is a cascade of evidence that genetic change come through genetically defined mechanisms.

Again you keep tripping over this absolutist position that any sort of one at a time change must be in conflict with PE. Again we agree that there are genetically defined mechanisms that control change, and the only difference here is that you insist that these changes were all programmed in in the beginning and that mutational changes have had no effect ever, and you insist that there's a large body of evidence that proves it which unfortunately I can't find... hmmmm...

Cheers,
Buffy

[Biochemist]

Buff-

I apologize for my last post above. I re-read it, and it sounded a little harsh. Apologies again.

[Buffy]

Originally Posted by Biochemist

I apologize for my last post above. I re-read it, and it sounded a little harsh. Apologies again.

No problem! I didn't get to be the Slayer without having a pretty tough hide!

Cheers,
Buffy

[Biochemist]

Originally Posted by Buffy

There's lots of documentation that the mechanisms for removal of mutations exist. If you're going to argue, at least don't contradict yourself!

Sorry for my lack of clarity. I certainly agree that genes are self correcting and self policing.

Actually, the vast majority of code *is* non-functional!

Not nearly enough. And every year we find out that more of the stuff that we though was "junk" isn't junk at all. But even if 95% is non-functional, there is no explanation for the incredible efficiency of removal of code from the primary DNA sequence.

...This self correction does not have to "know" about danger, in fact it may not be dangerous at all, its just that over time the RNA and proteins have been selected that will zap a certain sequence.

Really. Which sequences? Where is that information stored?

...if you simply see that yes its unlikely that this self correction process doesn't protect "good" but rather just gets rid of "selected bad", all of a sudden lots of mutations escape being excised, giving themselves opportunity to be expressed.

Again, that "selected bad" is an awfully large information file to get the efficieney of retention up to being useful 5% of the time. Where is that information stored?

...this whole quote proves you have no idea of why any mechanism that selects has a dramatic impact on probabilities. It turns unbelieveably remote possibilities in to near-certainties. You really need to read up on this.

This is mathematical sleight of hand. If you are suggesting that all previously recognized "bad sequences" are stored, remembered and then deselected, where is that information stored?

Further, you are stuck on this notion that "whole sequences are constructed randomly" when you should know that that is not what's in those introns: they are in most cases not just random sequences of neucleotides, they are in most cases copies of sequences that at one time were exons themselves, and a single change to one neucleotide pair in that sequence could make it useless or useful and ready to be expressed.

Well, are you saying with that "bad" sequences (in all of phylogeny?) are stored, recognized and rejected? Or that unexpressed codon sequences can be recognized before their phenotypic expression for rejection? I don't think it is my math that is bad here.

If its useful, the organism that has it will survive, otherwise it won't.

You are still dodging the math. Either we stored all previous errors in the phylogeny in the higher phylum code (which would take up some significant portion of the 95% of "junk") or we can predict bad stuff berfore it is expressed phenotypically. How? The quantity or previously rejected "errors" would have to dwarf the existing funcitional code by at least a factor of 10^1,000 (conservatively). That is a lot of recognition. Where is that stored?

Well, again, your math here is meaningless. All that has to happen is that the mechanisms get rid of a lot of stuff, but either of those numbers indicates there's a huge playground to work in.

Buff- This math is real. It seems to me that you wave your wand and say it is not a problem.

...these correction mechanisms have evolved over time because they're selected for by protecting organisms and allowing them to succeed, which I guess to you is much more logical in saying they were programmed in from the very beginning...

This is now a circular argument. You are saying that the selection mechanisms themselves were selected, but they require a learning capability to store previous adverse selections to make them functional. That is a pretty high bar for a system to be selected. When did this biochemical service arrive on the scene? The first prokaryote?

....You wanna try to figure the odds that all this stuff was *bam* programmed perfectly in the first prokaryote?

Buff, I was not trying to answer that problem. It is what it is. The issue I am addressing is the problem of a viable mechanism for PE. If the "simplest" solution for PE makes abiogenesis more complicated, so be it. But the evidence for mutation-based speciation is absent. The evidence for programmed speciation is inferential but strong, and it gets stronger every year.

...you insist that these changes were all programmed in in the beginning...

Technically no. I am suggesting that the specific propensity for the characteristics of the daughter species is contained in the genome of the parent species. The unfortunate logical corollary is that all phyla were coded a long way back. If you can think of some other mechanism, let me know.

... and that mutational changes have had no effect ever, and you insist that there's a large body of evidence that proves it which unfortunately I can't find...

Now you are being absolutist. What I said was that mutation had an insigificant effect on speciation. I don't really care if it EVER happened, because it is not the primary mechanism. You cannot prove a negative in the scientific method. No one can prove it never happened. But I noticed that you have not produced evidence that it did.

And there are numerous demonstrable mechanisms for viable genetic alteration that are not based on mutation.

[bumab]

Ha! The conversation's gettin' interesting!

Buffy's right- the vast majority of DNA is non-coding, functional or not. This allows mutations to accumulate RANDOMLY and still be expressed en mass after some stimuli. If the only argument for your position is "there's not enough proof of mutation induced change" then it's not far from... well, you know what

That might be over the line

Regardless, the fossil record shows spurts of rapid change to new lifeforms. I don't see evidence that those lifeforms were pre-programmed- one would expect to see the same morphologies show up time and again. Rather, we see all sorts of unique morphologies. RANDOM morphologies.

Second, one would assume we could find specific (say liver cell protiens) genes from jellyfish in our genes, and ours in theirs. We don't. Their coding genes are very different. Thanks to the whole genome being known in a few species, we can test this further, but I don't see a reason to think we'll find otherwise.

Finally, a mechanism for control. If organisms somehow "choose" a new morphology from their genetic database, how do they choose? Why doesn't every species on the brink of extinction switch into a new body?

Genes can change under natural selection, and that change is passed to new organisms. My parents goldfish are now all black- the orange ones got eaten by racoons. The babies are black. 95% are black. It's a superficial change, but it contradicts your theory. If you are just talking about speciation, there's lots of examples of that, too. A mechanism that simply increases the rapidity of mutations (they are not truely random, there are mutation correcting protiens in the cell, Buffy did a good explanation here) would go a long ways towards explaining PE.

[bumab]

Originally Posted by Biochemist

You are still dodging the math. Either we stored all previous errors in the phylogeny in the higher phylum code (which would take up some significant portion of the 95% of "junk") or we can predict bad stuff berfore it is expressed phenotypically.

Non-coding DNA is not just a random collection of bases. You can't assume that in the math- that's part of the problem. A lot of non-coding DNA USED to be genetic code, used for protiens. Occasionally, these genes are copied (it happens often). Those genes can mutate, forming a new coding sequence. If it's beneficial, good. If not, there's another coding sequence still up and running. So the mutated gene is no longer used, and that's that. It sticks around. It's got a useful code, it's just a little out of date. It's possible that gene can mutate further, and perhaps become useful again. "all it needs is a chance," or a TATA box, whichever comes first.

[Biochemist]

Originally Posted by bumab

the vast majority of DNA is non-coding, functional or not.

95% non-coding is not "vast". It would have to be well over 99.999999% for the mutation model to hold true

This allows mutations to accumulate RANDOMLY and still be expressed en mass after some stimuli.

Only if the VAST majority of non-functional solutions were weeded out. You have to get past the random odds of 1 in 10^1000 (conservatively) for a new enzyme system. That is a lot of code to store for an occasional win.

Regardless, the fossil record shows spurts of rapid change to new lifeforms. I don't see evidence that those lifeforms were pre-programmed- one would expect to see the same morphologies show up time and again. Rather, we see all sorts of unique morphologies. RANDOM morphologies.

How in the world could you make this statement? This is the problem that my solution addressed and mutation does not.

... one would assume we could find specific (say liver cell protiens) genes from jellyfish in our genes, and ours in theirs. We don't. Their coding genes are very different. Thanks to the whole genome being known in a few species, we can test this further, but I don't see a reason to think we'll find otherwise.

Isn't this making my point? We do NOT see serial increases in complexity. We see dramatic differences in what might appear to be "sequential" phyla. This is the argument for programming, not mutation.

...If organisms somehow "choose" a new morphology from their genetic database, how do they choose? Why doesn't every species on the brink of extinction switch into a new body?

Huh? I don't thnk they "chose" anything.

...My parents goldfish are now all black- the orange ones got eaten by racoons. The babies are black. 95% are black. It's a superficial change, but it contradicts your theory.

How? Species go extinct all of the time. 95%+ of them are already. What did your example show about speciation?

If you are just talking about speciation, there's lots of examples of that, too.

Is there ONE example of speciation by mutation?

A mechanism that simply increases the rapidity of mutations (they are not truely random, there are mutation correcting protiens in the cell, Buffy did a good explanation here) would go a long ways towards explaining PE.

I still think you are both dodging the math. The tools that improve the odds of success (after mutation) either store previous error history, or they select for functional characteristics that are predicted in the code. In the former case, there STILL is not enough non-functional DNA to act as fodder for random creaton of even moderately complex enzyme systems.

That emperor is getting might cold without his clothes.

[Buffy]

I must say, I'm getting dizzy. First you say:

Originally Posted by Biochemist

Sorry for my lack of clarity. I certainly agree that genes are self correcting and self policing.

...and then you spend most of the rest of the post saying that its not possible because it wouldn't be possible to remember all the bad sequences, so its not possible for them to police it. So, I'm really confused about what you're arguing here. The replication mechanism is what keeps most of it straight, and the fact that the correcting mechanism is *not* 100% effective is exactly why its possible to get the mutated genes to persist! Moreover, the mechanism for policing doesn't work like a computer, comparing sequences to a "bad sequence database": its an effect of what gets copied and what doesn't, and there's proof (the SciAm article I mention a few posts above this one) that introns *are* copied, as well as some being skipped (the bad ones), and the chemical process itself has the skipping of the bad stuff programmed into it. So what's your point?

Again, you insist that if we introduce *any* randomness, it must therefore *all* be random, so you think that your math is right. Its not. As you say:

Originally Posted by Biochemist

... every year we find out that more of the stuff that we though was "junk" isn't junk at all.

Yes, "junk" is a misnomer, that's exactly what I've been arguing: its useful stuff and when a mutation occurs its somewhat akin to this analogy: DNA is a car. There are a bunch of redunant introns that look like carburetor, one gets hit and mutates from a sequence that codes for a two-barrell carb into a three barrell, in another car it gets hit with a different mutation and becomes a 4-barrell. Because of the configuration of the intake manifold, when CO2 levels rise due to a meteorite impact, the 4-barrell proves more efficient when some of the cars with the 4-barrell gene have it expressed, whereas the 3-barrell cars die off, and the few that also have an enlarged intake manifold due to the mutation of another gene, beat everyone out because of the severe conditions and the fact that only those with both those variations will survive.

Now ask me the difference inprobabilities given these two situations:

• A bunch of pieces of metal, screws, bolts sitting in a pile with whizzing things all come together at once to form a carburetor.
• A carburetor is sitting there and gets another bore drilled into it, and an intake assembly plops into it and gets connected to the same linkage that operates the others.

According to your brand of math, these are equally likely. They really aren't. Honest. Think about it.

Originally Posted by Biochemist

This is now a circular argument. You are saying that the selection mechanisms themselves were selected, but they require a learning capability to store previous adverse selections to make them functional. That is a pretty high bar for a system to be selected.

Well, again, we're arguing the same mechanisms exist, so I don't follow why you're arguing against mechanisms that are required for your own theory, but moreover, except in your theory where the selection mechanisms are unaffected by any changes whatsoever over time, its not in the least bit illogical to argue that each of the mechanisms were selected over time. In fact, that explains why they *don't* need to work like a database: they're programs, and only the programs that are successful in excising bad genes survive!

Originally Posted by Biochemist

I don't really care if it EVER happened, because it is not the primary mechanism. You cannot prove a negative in the scientific method. No one can prove it never happened. But I noticed that you have not produced evidence that it did....And there are numerous demonstrable mechanisms for viable genetic alteration that are not based on mutation.

All I have seen you say is that you agree with the rest of us that there are mechanisms--call them programming, I'll agree--that cause and control the changes to species over time, but that its not the *primary* mechanism, for which you have shown no proof either, in fact all you say is that its inferentially strong. Big deal. Your argument against mutation sounds like the old creationist harping about "we can't actually observe major species changes because they take too long and if we can't see them happen that's proof that they *don't* happen." Since you don't offer any citations to back up what you're referring to as proof, I don't know where to even start understanding what you're trying to say. I'm pointing to literature along the way here that shows various mechanisms that show how mutations can be propagated at least. I don't mind if you belittle me and tell me I'm stupid or dogmatic, but you're not doing a very good job of being convincing! Sorry!

Still wanna know if we can examine our DNA and build a possible "nth-generation human" of what we will evolve into 10 million years from now: according to your theory, its all pre programmed in there! I bet PS2Huang would love to find out how we skip the intervening steps in it so he can be 8 ft tall tomorrow!

Cheers,
Buffy

[Biochemist]

Damn. I think we are completely wrapped around the axel again. I will reattempt to post a proposition from scratch in the morning.