Annoying creationists

[kleinman]

Annoying Creationists

What is the value for valuation[p] for a match in the binding site region versus a mismatch in the non-binding site region? If valuation[p] in the binding site region is much greater than valuation[p] in the non-binding site region, you have reduced the weight of errors in the non-binding site region. This is equivalent to ignoring harmful mutations in the non-binding site region. And as we all know, most mutations are harmful.

What do you mean by "binding site region"? There are binding sites and then there are all the other positions on the chromosome.

The binding site region is the portion of the genome where a match to the weight matrix is considered to be correct. In Dr Schneider’s original selection scheme, a failure to match in this region causes an error, where as matches in the region not assigned to the binding sites do cause errors. Now Unnamed is using the value summed from the weight matrix minus the threshold to determine the selection. The problem with this is his variable sv is biased to the matches in the binding site region. The value obtained from a good match of the weight matrix to a particular position in the binding site region gives a large value to valuation[p] which will dominate the selection process against poor matches against sites in the non-binding site region of the genome. Once you have a good match to a binding site in the binding site region, the large value for valuation[p] for that binding site will prevent a mutation which causes a match (albeit poor) in the non-binding site from affecting the selection process because valuation[p] will be a smaller in magnitude for that erroneous match. Erroneous matches in the non-binding site region no longer are exerting as much influence on the selection process.

This in effect is ignoring potentially harmful mutations in the non-binding site region of the genome. Once you have several good binding sites in the binding site region of the genome, it is very unlikely that these creatures will ever be selected against because of the high values of valuation[p] for these sites which will nullify whatever happens in their nonbinding site regions.

For a good discussion of how the weight matrix is used to compute a match, see the paper An Evaluation of “Ev” by I.G.D Strachan which is available on the internet. http://www.iscid.org/papers/Strachan_EvEvaluation_062803.pdf

 

[PixyMisa]

If y'all find mice and dogs meet your criteria as a demonstration of anything other than the fact that Mendel was basically right, your choice.

Mendel was right about what?

We have observed speciation - albeit not, despite all our efforts, in dogs.

Oops. I guess we agree, there. One species; check.

Yes, all dogs belong to a single species. This has been pointed out to you before.

Nah, I think something in the lab a bit more malleable than mammals might be a place to start ... oh, iirc, you evolutionarians have been trying in the labs since the early 50's?

Yes, and researchers have succeeded on a number of occasions, as you have seen in the documents you've been directed to.

Even the first one?

As I said, that's not evolution. You're talking about abiogenesis, and no-one knows how that happened, because there is no physical evidence left anywhere. There are a number of hypotheses as to how it may have happened, but none have been worked out in sufficient detail or backed by sufficient experimental evidence to be fully convincing.

If you are talking about evolution, though, then genes are never created de novo. New genes are always formed through mutations of existing genes. (And as a side note, recombination never creates new genes.)

PS. Don't you have a corrupt database somewhere that needs fixing?

No.

 

[Unnamed]

Unnamed, just checking that you are using these parameters:

population 32
binding sites 8
weight width 9
site width 10
1 mutation per 512 bases

Ok, I'm attaching the dialog again.

Current data:
1024, 231, 260, 186
2048, 264, 257, 266
4096, 263, 263, 253
8192, 273, 347, 333
16384, 345, 304, 315
32768, 440, 422, 474, 288
65536, 432, 443
132072, 584, 563
262144, (running 146 generations, 15 mistakes to fix)

Unnamed, could you run the 32K case and tell me the number of births and deaths?

32768, 288 generations, 4624 births, 4592 deaths, 2465 by selection

We could test this by increasing sv by a fixed amount for spurious bindings. A good value would be, say, threshold/2.

I can't get it to work well. First, I had to use threshold/200, because sv was saturating (becoming larger than MAX_INT). Now, it seems that the procedure is selecting for a small absolute value of threshold rather than small number of mistakes. That's curious, but it makes sense: a few mutations can lower threshold, and therefore (spuriousHits*threshold), much faster than correcting each mistake. I can still try a fixed value, but linking it to threshold is not working.

In addition, there is no selection process for a partially completed gene.

I don't think that's true, but if it were, then my change would be meaningless.

I think if you keep track of the value of valuation[p] in the following equation,

sv += Math.abs(valuation[p] - threshold);

you will see that this number will be much larger in the binding site region due to the good match of the weight matrix. In the non-binding site region this value will remain small. You are effectively giving a higher weight to good mutations in the binding site region than to harmful mutations in the non-binding site region.

That makes more sense than your previous post. I won't dispute that, but Paul's suggestion should provide a test for that. But remember that there are many more non-binding sites than binding sites, so one could argue that I am giving more weight to non-binding sites.

You are still selecting for good and harmful mutations in the binding site region. Here is a simple computational experiment you can do. Set the values for gene=1 and nongene=0 and sort on mistakes rather than sv. Then compare the number of generation for convergence for you series that you have run with sv. I believe you will obtain almost the same number of generations for convergence for the two approaches.

When I set nongene=0 I get a "perfect" creature that still has binding mistakes. It converges much slower than my sv code (I gave up waiting after 3x the number of generations, because the number of mistakes stopped decreasing).

This is not necessarily an unrealistic approach to the selection process. Months ago I proposed that instead of starting with a random genome, start with a real genome and then define your selection process such that if there is a fatal mutation in the non-binding site region that this creature be selected out no matter what is happening in the binding site region.

But then it's not de novo evolution of information, defeating the purpose of the experiment.

What is the value for valuation[p] for a match in the binding site region versus a mismatch in the non-binding site region? If valuation[p] in the binding site region is much greater than valuation[p] in the non-binding site region, you have reduced the weight of errors in the non-binding site region. This is equivalent to ignoring harmful mutations in the non-binding site region. And as we all know, most mutations are harmful.

I don't know what the values are, and I don't think you can say that because you don't know either.

This is frightening.

Assuming you have no bugs, this shows us that the particular selection criterion used by Ev is critical to the speed of evolution, and that we certainly cannot draw any sweeping conclusions about the real world.

I'd say the same happens for real-world evolution. Whatever can't be selected efficiently will evolve slowly, while showstoppers like fatal mutations and novel genes will just jump out of the process.

Once you have several good binding sites in the binding site region of the genome, it is very unlikely that these creatures will ever be selected against because of the high values of valuation[p] for these sites which will nullify whatever happens in their nonbinding site regions.

Once the binding site is good, it ceases to contribute to sv. Only errors contribute to sv.

 

[kjkent1]

In addition, there is no selection process for a partially completed gene.

Your comment here seems to suggest that you have some a priori knowledge of what a completed gene will be before it evolves. For instance, you keep talking about evolving hemoglobin de novo. Who says that the gene must begin as a hemoglobin gene? Why can't it start out providing some other selective advantage, then proceed to some other selective advantage, so that at present it happens to be a hemoglobin gene because we're here to observe it (anthropic principle), and sometime 10 million years in the future, it may be some other sort of gene that provides some selective advantage, or, perhaps no gene at all?

I understand that their is probably a bare minimum of bases necessary to provide genetic utility, but do you know precisely where that threshold is reached?

Do you know when a particular coding suddenly stops being junk DNA and becomes a selective advantage (or a disadvantage)?

If you do, then can you please tell me what the closing price of the NASDAQ 100 will be tomorrow? Baby needs new shoes!

This is equivalent to ignoring harmful mutations in the non-binding site region. And as we all know, most mutations are harmful.

If the region is non-binding, then how can the mutation be either harmful or beneficial? According to Schneider's "EV" paper, the mistakes in the non-binding region are simply "wasteful binding of the recognizer." I assume this means "junk DNA." Schneider continues: "For simplicity these mistakes are counted as equivalent, since other schemes should give similar final results."

Apparently, "unnamed" has just demonstrated that while "other schemes" may give similar "final" results, the final results are obtained in several orders of magnitude less generations.

So, the new algorithm isn't really ignoring harmful mutations -- it's just ignoring the junk.

 

[RecoveringYuppy]

For instance, you keep talking about evolving hemoglobin de novo. Who says that the gene must begin as a hemoglobin gene? Why can't it start out providing some other selective advantage, then proceed to some other selective advantage, so that at present it happens to be a hemoglobin gene because we're here to observe it (anthropic principle), and sometime 10 million years in the future, it may be some other sort of gene that provides some selective advantage, or, perhaps no gene at all?

I mentioned this before but it got lost in the noise. The point you are making is well illustrated by hemoglobin. In the first place, there is no single hemoglobin gene. Hemoglobin is the product of of multiple reactions catalyzed by several enzymes and precursor proteins. And many of the individual components have other uses or are closely related to other enzymes and proteins. And there are an extremely large number of hemoglobins.

 

[PixyMisa]

PS. Don't you have a corrupt database somewhere that needs fixing?

No.

Though come to think of it, I do have a 1.6TB RAID array with a failed disk...

 

[Unnamed]

The 256k case converged:
262144, 463 generations

 

[kjkent1]

I mentioned this before but it got lost in the noise. The point you are making is well illustrated by hemoglobin. In the first place, there is no single hemoglobin gene. Hemoglobin is the product of of multiple reactions catalyzed by several enzymes and precursor proteins. And many of the individual components have other uses or are closely related to other enzymes and proteins. And there are an extremely large number of hemoglobins.

I love this thread. People actually know what they're talking about (except for me -- I'm barely hanging on by my white knuckles).

 

[Kotatsu]

I can see how proponents of Modern Evolutionary Theory could feel the results bolster their position. Other arguments could be made from opponents, with the literature full of loose ends waiting to be tugged sfaics.

Notice that these articles are ten years old. I have no idea if any of these ends have been tugged at since, but I know that the results of Song et al. inspired at least one other article (Soltis & Soltis, which I believe I referenced above); possibly either group has continued the work, but I don't really know where I should look for it.

Perhaps we could agree on something more along the lines of a new, clearly different, phenotype substantiating genotype changes?

It's a pity that Song et al. don't elaborate on exactly what the phenotypic differences between AB/BA, AC/CA and the parent plants are, apart from the fertilities. However, they do point out that the differences are there. And as the natural polyploid hybrids are sufficiently different --- phenotypically --- to have warranted separate species status long before the coming of molecular studies, I'd hazard a guess that there are substantial phenotypical differences between, for instance, Brassica rapa (A genome) and B. juncea (AB/BA genime).

However, as I've said I am no botanist, and don't know how much of a difference is needed for someone to describe a new species. And, furthermore, I am not you, why elaboration on your part is needed before "clearly different" can be agreed upon. Supposedly --- and to my knowledge we have only Song's et al. brief sentence to go on --- there is some variation between the F5 plants and the parent plants in terms of leaf shape, leaf colour, branching patterns, and number of side shoots. If these are characters traditionally used to assign species status to other plants within the same genus (or higher taxonomical level), and the devaitaions in the F5 plants from the parent plants are large enough, surely this speaks in favour of this being at least an initial example of the kind of speciation you are after?

Couldn't the geologic timespans nature seems to require to provide change be much compressed by conscious human selection and technology?

Do you mean, "Is experiments on artificially produced plants really comparable to what goes on in nature?"

Soltis & Soltis raises this question as well, but with a different slant (I think). They notice that while the artificially produced F2 plants are homozygous (due to the protocol and methods used by Song et al., I believe), the natural polyploid hybrids are likely heterozygous (due to the processes with which allopolyploidy occurs). Soltis and Soltis draw the conclusion that they are, based on about ten other studies which apparently show this. I haven't looked at any of these studies, though, so for the moment, I'll have to take their word for it until I have downloaded the referenced articles.

However, I do assert that allopolyploidization could be a much quicker speciation process than many others --- polyploidization, whether allo- or auto-, is virtually instantaneous, and if the progeny is deemed sufficiently different to form a new species (which often seems to be the case due to reproductive incompatibility between progeny and parents), then you have a kind of Insta-Species mechanism.

Many of these polyploids are ephemeral, though, unless they in some way can get around the incompatibility and reproduce by some other mode. Many oligochaetes, for instance, couple sexual reproduction to asexual reproduction in a seasonal pattern, which means that a new polyploid individual may produce many polyploid individuals by, for example, paratomy or architomy, with which they can then reproduce when the season for sexual mating arrives.

Anyway: I think the central finding of Song et al. is the possibility of rapid genome change through allopolyploidization. Whether or not this actually occurs in nature is not answered in their article, nor --- satisfactorily --- in Soltis and Soltis'. However, when I have time today, I shall browse through the referenced material and see what I can find. Soltis and Soltis, in particular, has a lot of seemingly useful references, although I am uncertain to what extent my university has access to all the botanical journals, as I rarely have reason to look for them.

 

[Kotatsu]

Nah, I think something in the lab a bit more malleable than mammals might be a place to start ... oh, iirc, you evolutionarians have been trying in the labs since the early 50's?

This is where oligochaetes are very handy --- often easy to breed, generally short generation times, and they show more or less every kind of reproductive system known. In the worm I've worked with the most, Lumbriculus variegatus this is especially true, as it is architomic (1), and regenerates very quickly. For the purposes of my studies (I am a systematicist), I have never had occasion to let them grow very much, but they generally regenerate about half a centimetre per 3-4 weeks without any added nutrients, just spring water. I believe this rate is higher when it's fed properly (I have some articles on it, but can't remember the numbers).

To my knowledge, no studies using this or any other suitable oligochaete to study the same kind of processes as reported by Song et al. have been done, but I think there's a group somewhere in central US (Duluth?) that has recently started with something similar, though I base this on a vague recollection from a speech at a conference half a year ago.

The problem is then transferred from lack of malleability to lack of suitable characters on which to detect phenotypic change, as several of the more common worms (especially those which have short generation times and are easy to use) seem to be polyphyletic or show genotypic or phenotypic differences to begin with.

---
(1) That is, it can reproduce by having either end pinched off, whereupun the pinched-off part undergo morphological and physiological changes to produce a new worm, after a given pattern. Not to be confused with paratomy, where the differentiation of the tissues which will form the new worm occurs before the actual split.

 

[Cuddles]

Yes, all dogs belong to a single species. This has been pointed out to you before.

I actually disagree with this. As a little gedanken, imagine that almost all dogs were wiped out leaving only great danes and chihuahuas. How many species do we have? Given the general definition of two groups of organism that are incapable of interbreeding, we have two. Now put all the other dogs back. What has changed to suddenly make those two species back into one?

The problem is exactly the same as that with classifying fossils. There are very rarely sudden changed that form new species, it is usually a gradual process in which the transitional forms can be viewed as either belonging to the original species or the new one, depending on how you look at it. This is why the phylogenic tree can only be drawn for one point in time, since transitional forms blur the lines between species and would prevent it being drawn at all.

The problem with dogs is that all (or at least a lot of) the transitional forms are present at the same time, which means that it is not really possible to say either that there is one species or that there are more, since there is no way to draw a line that demarcates species or to group all of them into one interbreeding species. All we can really say is that our definition of species is not perfect and dogs are one area where it simply doesn't work. Of course, this is one of the major problems creationarians seem to have - they simply don't seem able to accept that species, genes and even life are not fixed things, both they and their definitions are fluid and there will always be exceptions no matter what definition we use, the best we can manage is a good working definition that serves in most situations.

Given this, it makes no sense at all to speak of things like "macroevolution". Since we can't rigidly define a species, it is pointless to try to distinguish between changes within a species and changes to a new species since we can't know when a new species is created.

 

[cyborg]

Cyborg, I have never said I was educated in everything.

Are you educated in ANYTHING?

You have thus far demonstrated:

1) Inadequate logic
2) Inadequate mathematics
3) Inadequate computer science
4) Inadequate theology

Just what the hell is it you are educated in? I suppose it must be to be annoying then.

I’m only claiming I know something about ev and what this computation is saying about random point mutations and natural selection.

No, you're claiming that Quake should be able to tell us about relativistic physics because despite ev clearly not being anywhere near a comprehensive model of evolution it should be able to show that events you cannot define cannot occur.

I’m content to say that the theory of evolution is without a mathematical basis and that ev shows that random point mutations and natural selection is so profoundly slow that nothing can evolve by this mechanism.

You cannot simultaneously say evolution is without mathematical basis and then say ev is the mathematical basis upon which you have disproved evolution.

Are you profoundly slow or what?

Do you understand why Quake is an insufficient model of physics or not?

Do you understand why ev is an insufficient model of evolution or not?

Maybe I am going too fast...

English? Understand? Understand English???

Does to, does to.

Apparently not.

5) Inadequate English

You can stomp your foot and say it is my task, but it is your theory to prove.

No. You are trying to disprove evolution mathematically. It is therefore your task.

I doubt any evolutionarians who have knowledge of mathematics will try to include other mechanisms of mutations because all these mechanisms will still face the same 4^G effect of the increasing search space as the genome length is increased.

Proof.

Show your working. Algebra please. Assertion is not mathematics.

When an evolutionarian retracts their belief in the theory of evolution. That is a macroevolutionary event.

And when does this occur in ev?

Annoying an evolutionarian, that’s a microevolutionary event.

And when does this occur in ev?

 

[Paul C. Anagnostopoulos]

The binding site region is the portion of the genome where a match to the weight matrix is considered to be correct. In Dr Schneider’s original selection scheme, a failure to match in this region causes an error, where as matches in the region not assigned to the binding sites do cause errors.

I think you misspoke here.

Now Unnamed is using the value summed from the weight matrix minus the threshold to determine the selection.

The absolute value thereof.

The value obtained from a good match of the weight matrix to a particular position in the binding site region gives a large value to valuation[p] which will dominate the selection process against poor matches against sites in the non-binding site region of the genome.

Binding sites that are matched do not contribute to the value of sv.

~~ Paul

 

[PixyMisa]

I actually disagree with this. As a little gedanken, imagine that almost all dogs were wiped out leaving only great danes and chihuahuas. How many species do we have?

Depends on your definition of species, of course.

Given the general definition of two groups of organism that are incapable of interbreeding, we have two.

I wouldn't count on that. They are genetically compatible, and they'll assuredly try to interbreed. They're dogs.

Now put all the other dogs back. What has changed to suddenly make those two species back into one?

Continuity of the spectrum of genetic variation within the species.

The problem is exactly the same as that with classifying fossils. There are very rarely sudden changed that form new species, it is usually a gradual process in which the transitional forms can be viewed as either belonging to the original species or the new one, depending on how you look at it. This is why the phylogenic tree can only be drawn for one point in time, since transitional forms blur the lines between species and would prevent it being drawn at all.

Absolutely.

The problem with dogs is that all (or at least a lot of) the transitional forms are present at the same time, which means that it is not really possible to say either that there is one species or that there are more, since there is no way to draw a line that demarcates species or to group all of them into one interbreeding species.

Agreed. But given a continuous spectrum and active interbreeding throughout that spectrum, it usually makes more sense to classify it as a single species.

All we can really say is that our definition of species is not perfect and dogs are one area where it simply doesn't work.

I wouldn't say that. What I would say is that the definition of species is somewhat arbitrary - and necessarily so.

Of course, this is one of the major problems creationarians seem to have - they simply don't seem able to accept that species, genes and even life are not fixed things, both they and their definitions are fluid and there will always be exceptions no matter what definition we use, the best we can manage is a good working definition that serves in most situations.

Yep. Hammy's "I know one when I see one" being an example of this.

Given this, it makes no sense at all to speak of things like "macroevolution". Since we can't rigidly define a species, it is pointless to try to distinguish between changes within a species and changes to a new species since we can't know when a new species is created.

Micro- and macro-evolution are relative terms, at best. The creationarians think that they are qualitatively different - which is of course untrue.

 

[Kotatsu]

I took a look online, and after following several citations felt like Alice after she went down the rabbit hole.

Regarding Tragopogon sp. and Brassica I have downloaded the article by Soltis et al. and some additional articles by Song's group, but it seems parts of the university server just broke down, so I couldn't get all the articles I wanted (I can't seem to access the university library). Anyway, I shall look through these tonight and see what they say.

 

[Paul C. Anagnostopoulos]

Unnamed:

Will you create a variant of your variant and run the same series of experiments? This time, rather than totaling the distances between the valuations and the threshold, just use the maximum one.

~~ Paul

 

[kleinman]

Annoying Creationists

In addition, there is no selection process for a partially completed gene.

Your comment here seems to suggest that you have some a priori knowledge of what a completed gene will be before it evolves. For instance, you keep talking about evolving hemoglobin de novo. Who says that the gene must begin as a hemoglobin gene? Why can't it start out providing some other selective advantage, then proceed to some other selective advantage, so that at present it happens to be a hemoglobin gene because we're here to observe it (anthropic principle), and sometime 10 million years in the future, it may be some other sort of gene that provides some selective advantage, or, perhaps no gene at all?

This is classic evolutionarian thinking. Speculate all you want, just don’t call this science.

This is equivalent to ignoring harmful mutations in the non-binding site region. And as we all know, most mutations are harmful.

If the region is non-binding, then how can the mutation be either harmful or beneficial? According to Schneider's "EV" paper, the mistakes in the non-binding region are simply "wasteful binding of the recognizer." I assume this means "junk DNA." Schneider continues: "For simplicity these mistakes are counted as equivalent, since other schemes should give similar final results."

Apparently, "unnamed" has just demonstrated that while "other schemes" may give similar "final" results, the final results are obtained in several orders of magnitude less generations.

Dr Schneider’s selection process does not have a sound scientific basis but I accept it because it demonstrates something about the mathematics of random point mutations and natural selection. Unnamed’s selection process is counter what is observed. You can not ignore mutations in other portions of a genome and just consider mutations at the location you want to evolve. Unnamed’s selection process is unscientific.

Cyborg, I have never said I was educated in everything.

Are you educated in ANYTHING?

I’m educated enough to annoy evolutionarians!

I’m content to say that the theory of evolution is without a mathematical basis and that ev shows that random point mutations and natural selection is so profoundly slow that nothing can evolve by this mechanism.

You cannot simultaneously say evolution is without mathematical basis and then say ev is the mathematical basis upon which you have disproved evolution.

I can, I did and its true.

When an evolutionarian retracts their belief in the theory of evolution. That is a macroevolutionary event.

And when does this occur in ev?

Remember, macroevolutionary event take some time.

Annoying an evolutionarian, that’s a microevolutionary event.

And when does this occur in ev?

That happens all the time.

The binding site region is the portion of the genome where a match to the weight matrix is considered to be correct. In Dr Schneider’s original selection scheme, a failure to match in this region causes an error, where as matches in the region not assigned to the binding sites do cause errors.

I think you misspoke here.

You are correct, the match to the weight matrix has to exceed the threshold.

Now Unnamed is using the value summed from the weight matrix minus the threshold to determine the selection.

The absolute value thereof.

Do you want to tell us why he has to use the absolute value?

The value obtained from a good match of the weight matrix to a particular position in the binding site region gives a large value to valuation[p] which will dominate the selection process against poor matches against sites in the non-binding site region of the genome.

Binding sites that are matched do not contribute to the value of sv.

Which is why Unnamed has to use very small values for threshold in order for his scheme to work, otherwise binding sites in the binding site region no longer contribute to the selection process.

 

[cyborg]

I’m educated enough to annoy evolutionarians!

A feat worthy of a five-year-old who has learnt to say, "why?"

Congratulations on this haughty intellectual achievement.

I can, I did and its true.

Either ev is the mathematical representation of evolution and hence evolution is mathematically grounded and you have some sort of justification for saying ev can disprove evolution or evolution doesn't have any mathematical grounding and hence ev cannot possibly have anything to do with evolution being that it is essentially an exercise in mathematics.

You cannot have it both ways. Choose.

Remember, macroevolutionary event take some time.

Again, I would like to be able to know how long this is exactly. I therefore require that you tell me how I can identify a macroevolutionary event. I do not care if your conclusion is that ev shows that in real life macroevolution cannot occur - I want you to show me how to identify macroevolution in ev.

That happens all the time.

Which still fails to answer the question.

6) Inadequate reading comprehension

 

[kleinman]

Annoying Creationists

I’m educated enough to annoy evolutionarians!

A feat worthy of a five-year-old who has learnt to say, "why?"

I thought it would take at least a seven or eight year old to annoy evolutionarians. So you better teach your theory to infants before they have learned to say “why?”

I can, I did and its true.

Either ev is the mathematical representation of evolution and hence evolution is mathematically grounded and you have some sort of justification for saying ev can disprove evolution or evolution doesn't have any mathematical grounding and hence ev cannot possibly have anything to do with evolution being that it is essentially an exercise in mathematics.

You cannot have it both ways. Choose.

Here is where you are missing the distinction. The theory of evolution started without a mathematical basis. So along comes Dr Schneider who writes a mathematical model of random point mutations and natural selection. I look at the model and see this as plausible representation of the phenomena. When you use realistic genome lengths and mutation rates in the model, the rate of information is so profoundly slow that it disproves macroevolution by this mechanism.

So I choose both.

Remember, macroevolutionary event take some time.

Again, I would like to be able to know how long this is exactly. I therefore require that you tell me how I can identify a macroevolutionary event. I do not care if your conclusion is that ev shows that in real life macroevolution cannot occur - I want you to show me how to identify macroevolution in ev.

I keep setting up the target of de novo evolution of a gene. It seems that PixyMisa thinks this is abiogenesis. Joobz won’t tell us how ribose can be formed nonenzymatically yet PixyMisa seems to think that entire genes can form in the primordial soup. Where is the science?

That happens all the time.

Which still fails to answer the question.

Sure it does, what is the name of this thread? Maybe we are in the equilibrium part of the punctuated equilibrium.

 

[cyborg]

I thought it would take at least a seven or eight year old to annoy evolutionarians. So you better teach your theory to infants before they have learned to say “why?”

Hmm, I thought that was the job of your camp. To stomp out questioning and to instil obedience to mysterious gods.

When you use realistic genome lengths and mutation rates in the model, the rate of information is so profoundly slow that it disproves macroevolution by this mechanism.

So again I must ask - WHAT is macroevolution?

If you cannot define it then you cannot say it is disproved. Screw what a realistic genome length and mutation rate is - they are irrelevant until you can say what macroevolution IS.

So I choose both.

Ah, cognitive dissonance at its finest.

I keep setting up the target of de novo evolution of a gene.

7) Inadequate memory

The words 'de novo evolution of a gene' did not pass your lips until fairly recently.