Annoying creationists

[Paul C. Anagnostopoulos]

Are you saying that it take more time for ev to satisfy all three selection conditions simultaneously than it takes to satisfy any single selection condition alone? If that is what you are saying, that is a bit of an understatement. It is about a million times slower on the 16k genome case.

Yes, I agree with that, for the specific selection pressures that Ev models. I believe "more time" covers a factor of a million.

Dr Richard, once you finish Sesame Street, we’ll start you on the mathematics of ev. You evolutionists think that huge populations markedly accelerate evolution. Ev is showing otherwise and if you realized that increasing populations affect the probabilities in less than an additive manner, the results from ev become understandable.

Let's run some population experiments, shall we? I'll use these parameters:

genome size 8K
binding sites 16
weight/site width 5/6
1 mutation per genome
pause on perfect creature

population, generations
64, 710152
256, 300000 (est.)

Stay tuned ...

~~ Paul

REQUEST FOR CYCLES: If anyone has a fast machine with lots of RAM, please run some large cases, for example, a population of 16,384 or 20,480. Remember, a cycle lost is gone forever.

 

[delphi_ote]

In a nutshell:

I can demonstrate that a process I believe to be theoretically impossible takes an implausible amount of time using a simulation I believe is totally inaccurate.​

The number of levels on which this argument eats itself is just spectacular.

 

[kleinman]

Annoying Creationists

Dr Richard, once you finish Sesame Street, we’ll start you on the mathematics of ev. You evolutionists think that huge populations markedly accelerate evolution. Ev is showing otherwise and if you realized that increasing populations affect the probabilities in less than an additive manner, the results from ev become understandable.

Let's run some population experiments, shall we? I'll use these parameters:

genome size 8K
binding sites 16
weight/site width 5/6
1 mutation per genome
pause on perfect creature

population, generations
64, 710152
256, 300000 (est.)

Stay tuned ...

This is going to be fun. One thing Paul, with your 1 gig of memory on your machine, you are going to be limited to about 256k population but we should see a rapid drop in the slope of the generations for convergence/population curve.

REQUEST FOR CYCLES: If anyone has a fast machine with lots of RAM, please run some large cases, for example, a population of 16,384 or 20,480. Remember, a cycle lost is gone forever.

Paul, there are also memory limitations to the Java version of ev. The Pascal version of ev can run larger population cases. You also have the capability of breaking up a case into a series of shorter runs that can be restarted from a given state. This allows running a case over multiple nights for example. If I remember correctly, the 1k genome length, 1meg population case took around 300meg of memory and 100 hours of computer time on a 2.7gigaHz machine.

I can demonstrate that a process I believe to be theoretically impossible takes an implausible amount of time using a simulation I believe is totally inaccurate.

Delphi, you have demonstrated nothing. What I have demonstrated with ev is that mutation and selection becomes profoundly slow as you lengthen the genome in the model. I have also shown that increasing population decreases the generations for convergence at a rapidly decreasing rate (Paul is going to demonstrate this again with the series he is now to running). I have also shown that reducing the selection conditions in ev to one condition markedly accelerates the rate of convergence of the model. I have also given a real example where multiple selection pressures slow the evolutionary process. So unless you are going to subscribe to the view that life evolved on tiny genome creatures that no longer exist that could sustain much higher mutation rates than living things do now, and were only subject to one selection condition at a time (whatever that selection condition could be), you are done in by the mathematics of mutation and selection.

The number of levels on which this argument eats itself is just spectacular.

This mathematics is not going to disappear. You have no selection process to evolve a gene from the beginning. You have no mathematical basis for your theory. You have no chemistry to explain abiogenesis. You do have speculations, extrapolations and denial as a basis for your pseudoscience.

Hey Delphi, would you tell us how the gene came to be that you use in your gene duplication story?

 

[kjkent1]

This mathematics is not going to disappear. You have no selection process to evolve a gene from the beginning.

False. You have proved this up yourself. With all selection conditions turned off, as they would have been before the first gene existed, a gene ab initio appears in the first generation. Must be a huge bummer to have your own proof used against you. Either your proof is irrational, or your entire theory is wrong. Good thing for you I'm giving you the benefit of the doubt, that you realize by now, that zero mistake weights in ev produce nothing but a bug. Otherwise, you would have to choose between being wrong and being wrong.

You have no mathematical basis for your theory. You have no chemistry to explain abiogenesis.

LOL! Riiight. See above.

You do have speculations, extrapolations and denial as a basis for your pseudoscience.

Let's see now. We have the theory of evolution wihch appears to answer the majority of questions about how life developed, using observations of the physical world.

Contrast that with the theory of God, which answers all questions about all subjects, using blind faith in what must, by definition, always remain unknown.

Given the two alternatives, it's no wonder that you would choose the latter.

You really must to remember to wear your aluminum hat.

 

[Paul C. Anagnostopoulos]

Paul, there are also memory limitations to the Java version of ev. The Pascal version of ev can run larger population cases. You also have the capability of breaking up a case into a series of shorter runs that can be restarted from a given state. This allows running a case over multiple nights for example. If I remember correctly, the 1k genome length, 1meg population case took around 300meg of memory and 100 hours of computer time on a 2.7gigaHz machine.

Great, thanks for volunteering to run the population 16,384 case.

~~ Paul

 

[kleinman]

Annoying Creationists

Paul, there are also memory limitations to the Java version of ev. The Pascal version of ev can run larger population cases. You also have the capability of breaking up a case into a series of shorter runs that can be restarted from a given state. This allows running a case over multiple nights for example. If I remember correctly, the 1k genome length, 1meg population case took around 300meg of memory and 100 hours of computer time on a 2.7gigaHz machine.

Great, thanks for volunteering to run the population 16,384 case.

There are a couple issues here Paul. The first is that a G=8192, population 16,384 case will not run in your online version of ev. Even if it could run in evjava, it would have to run continuously until converged and I can’t run my computer 24 hour a day. I’ll run this case on the Pascal version of ev but it will take me a while to get this case done.

And stop whining, all the largest population data we have I generated months ago. I sent you a compiled version of Pascal ev that would run a 2 meg population on a G=1000 and you have yet to run this case (this case requires too much memory to run on my machine).

I do commend you though because at least you are starting to look at the data from ev again.

One question for this post with respects to a comment you wrote a few posts back.

I agree that evolving a distinguishing binding site requires more time than evolving a nondistinguishing gene threshold.

Are you abandoning your Rcapacity concept?

 

[Paul C. Anagnostopoulos]

There are a couple issues here Paul. The first is that a G=8192, population 16,384 case will not run in your online version of ev. Even if it could run in evjava, it would have to run continuously until converged and I can’t run my computer 24 hour a day. I’ll run this case on the Pascal version of ev but it will take me a while to get this case done.

It'll run fine if you tweak your Java run-time parameters. Press the Pause button if you want to stop it for awhile. By all means, run it with the Pascal version.

I agree that evolving a distinguishing binding site requires more time than evolving a nondistinguishing gene threshold.

Are you abandoning your Rcapacity concept?

All I get from this is a non sequitur.

~~ Paul

 

[delphi_ote]

Hey Delphi, would you tell us how the gene came to be that you use in your gene duplication story?

Can put together an argument that doesn't derive inferences from the negation of the premise? (hint: If you want to convince people that evolution is impossible, you probably shouldn't use observations of evolution happening in nature and a simulation to make your case.)

 

[kleinman]

Annoying Creationists

I agree that evolving a distinguishing binding site requires more time than evolving a nondistinguishing gene threshold.

Are you abandoning your Rcapacity concept?

All I get from this is a non sequitur.

Let’s see if I can put this in sequence for you. You said the following concerning the series we had just been working on.

If you do run the 32K genome, I predict it will take about 18 million generations. Note that these experiments have both an increasing genome size and a decreasing rate of mutation.

What do you predict the number of generations for the perfect creature for the G=64k case? Binding site width is 6 for this case.

Hey Delphi, would you tell us how the gene came to be that you use in your gene duplication story?

Can put together an argument that doesn't derive inferences from the negation of the premise? (hint: If you want to convince people that evolution is impossible, you probably shouldn't use observations of evolution happening in nature and a simulation to make your case.)

I have always said that microevolution occurs. It is evolutionists who extrapolate these microevolutionary processes to a macroevolutionary theory. There are no selection processes that take a series of microevolutionary processes to accomplish a macroevolutionary change. If there was, you would describe the selection process that takes a duplicated gene and transforms this gene to some totally new function. Darwin did this when he observed variations in finch beaks which is a recombination and natural selection phenomena, which is a microevolutionary process. Stephen J. Gould did this with his concept of punctuated equilibrium which he inappropriately applied to mutation and selection but does appropriately fits recombination and natural selection.

If you think the evolution of drug resistant HIV viruses is a macroevolutionary change, then I’ll have to disagree with you. The slowing of the evolution of drug resistant HIV viruses by using three selection pressures is analogous to what ev show with it’s three selection conditions. Anything more than a single selection condition profoundly slows the mutation and selection process. That is what the mathematics of ev is showing and that is what happens in reality.

Delphi, your theory of evolution is mathematically impossible.

 

[delphi_ote]

I have always said that microevolution occurs... Anything more than a single selection condition profoundly slows the mutation and selection process.

So adaptation occurs, but is stopped as soon as it encounters too many selection pressures, which prevents speciation? Is this your hypothesis? If so, what is the maximum number of selection pressures to which an organism can adapt? How did you arrive at this number? Can you point me to biological literature that lists this number?

 

[kleinman]

Annoying Creationists

I have always said that microevolution occurs... Anything more than a single selection condition profoundly slows the mutation and selection process.

So adaptation occurs, but is stopped as soon as it encounters too many selection pressures, which prevents speciation? Is this your hypothesis? If so, what is the maximum number of selection pressures to which an organism can adapt? How did you arrive at this number? Can you point me to biological literature that lists this number?

The use of multiple antimicrobials is not limited to the treatment of HIV. Combination antibiotics are also often used in the treatment of bacterial infections. These are the most obvious use of multiple selective pressures to prevent the evolution of drug resistant microbes.

I don’t know of any evolutionist who ever considered that multiple selection pressures would interfere with the evolutionary process. This fact only became apparent when studying the behavior of ev. When I saw this type of mathematical behavior, I started looking for real examples of this phenomena.

Perhaps evolutionists will reconsider their interpretations of their observations in the light of this mathematical finding and the real examples of this finding.

Remember, these are examples of how multiple selective pressures prevent microevolution. Macroevolution requires huge numbers of mutation and selection events. These evolutionary events would be easily confounded by multiple selection conditions. If Paul modified ev to evolve two different sets of binding sites simultaneously with six selection conditions, I would be very surprised if this model would ever converge no matter what length genome was used.

 

[mijopaalmc]

Please forgive if this me if this has already asked this question:

Does any one know where I can find the primary literature on Ev?

 

[delphi_ote]

<Text which does not answer my questions.>

That's all very interesting, but I need three one word answers and one expository answer:

1. So adaptation occurs, but is stopped as soon as it encounters too many selection pressures, which prevents speciation? Is this your hypothesis?
2. If so, what is the maximum number of selection pressures to which an organism can adapt?
3. How did you arrive at this number?
4. Can you point me to biological literature that lists this number?

 

[kjkent1]

Please forgive if this me if this has already asked this question:

Does any one know where I can find the primary literature on Ev?

http://www.lecb.ncifcrf.gov/~toms/paper/ev/

 

[Taffer]

I see, as Dr. A has been saying for some time, that kleinman continues to spout what has shown to be false premises and conclusions. HIV antiretrovirals are not using to slow evolution, but to provide a broad spectrum antiviral agent which, if resistance to one or more drugs is developed, can be "hotswapped" for different antiretroviral agents without completely stopping treatment. Yet he continues to claim it is because it slows evolution. He, when asked, has failed to provide any citations nor evidence that slowing evolution is the reason for multiple antiretrovirals. He has failed to provide any citations or evidence that evolution slows down in reality. He has misinterpreted the reasons behind different results gained from ev. When this is pointed out to him, he then makes the strawman and false dichotemic argument that either ev perfectly models reality, or ev doesn't model reality at all. He then goes on to claim that anyone questioning ev is questioning its validity as a model for point mutation and natural selection in the evolution of a binding site. He continues to claim that evolution slows down, and/or stops entirely (depending on his mood), yet provides no evidence that this happens in reality. Paul has continued to bravely explain, over and over again, why kleinman's assumptions about ev are false, and why his conclusions are false. Kleinman continues to claim that ev shows that evolution is impossible, despite the fact that ev is not a model for the entire evolutionary process. He continues to ignore the entire field of population and evolutionary genetics, which clearly show why he is wrong.

In conclusion, I am forced to find that either kleinman is a liar and/or a troll, or too unable/unwilling to learn and understand his falsity to be worth interacting with.

Good day, kleinman. I may return when you have come up with some different lies.

 

[Dr Adequate]

Well, at least he was constructing grammatical sentences ...

... about "string cheese".

 

[Dr Adequate]

Lie #1.

Truth #1.

 

[Paul C. Anagnostopoulos]

What do you predict the number of generations for the perfect creature for the G=64k case? Binding site width is 6 for this case.

Now Rfrequency = Rcapacity, so I think it will take a very long time. No idea.

~~ Paul

 

[Paul C. Anagnostopoulos]

If Paul modified ev to evolve two different sets of binding sites simultaneously with six selection conditions, I would be very surprised if this model would ever converge no matter what length genome was used.

The problem with this idea is that we know precisely how such a creature would converge: Both transcription factors would evolve to recognize exactly the same binding site sequence. The model would have to be much more complex to force the transcription factors to be independent. In particular the binding sites would have to control genes that performed life-sustaining functions, so that when one transcription factor started interfering with the other's binding site, selection pressure would stop it from doing so. This would allow the two transcription factors to evolve independently.

If Kleinman is correct, then no transcription factors have ever evolved. They were all poofed into existence.

~~ Paul

 

[kleinman]

Annoying Creationists

Does any one know where I can find the primary literature on Ev?

No problem, the computer model was developed by Dr Tom Schneider, the head of computational molecular biology at the National Cancer Institute. The program can be accessed at:

http://www.ccrnp.ncifcrf.gov/~toms/paper/ev/evj/evjava/index.html

Dr Schneider has published this model and this paper is available on the following page:

http://www.ccrnp.ncifcrf.gov/~toms/paper/ev/ev.html

You can search around the site and find a blog page, glossary and a variety of other helpful information. The only thing lacking on these pages is any data from the model based on realistic parameters. This you can easily obtain by changing the input parameters to the model and you will find that the theory of evolution is mathematically impossible.

Paul C. Anagnostopoulos wrote the java version of ev for Dr Schneider. Paul and I have been having a friendly discussion about this model and the theory of evolution for about a year. Initially this discussion was done by email and included Dr Schneider in the discussion. I wanted to take the discussion public and started a thread on the Evolutionisdead forum http://www.evolutionisdead.com/forum/viewforum.php?f=1&sid=dab46a8a655e24347ad4f45a2efa181f . I chose this forum because Dr Schneider had used this forum previously to discuss his model publicly. As far as I know, other than a brief mention of me on his ev blog page, Dr Schneider has refused to talk about his model publicly since this thread was started.

Welcome to this discussion.

That's all very interesting, but I need three one word answers and one expository answer:

1. So adaptation occurs, but is stopped as soon as it encounters too many selection pressures, which prevents speciation? Is this your hypothesis?
2. If so, what is the maximum number of selection pressures to which an organism can adapt?
3. How did you arrive at this number?
4. Can you point me to biological literature that lists this number?

1. Yes, and not only do multiple selection pressures slow and ultimately stop evolution, natural selection actually impedes the transformation of genes from one to another because it prevents the gene from diverging too far from its functional form. You can counter this argument by telling us what the selection process is that would transform an existing gene from one form to another.
2. That depends on the severity of the selection pressures. A single profound selection pressure can cause extinction. When multiple non-fatal selection pressures are acting, it can vary on the severity of the selection pressures. The three drug therapy for HIV is very close to being sufficient.
3. Ev uses three selection conditions and the treatment of HIV uses three selection pressures. Both of these situations show how much only three selection conditions profoundly slow the evolutionary process.
4. I already posted the URL to the guidelines for treatment of HIV and posted a quote from these guidelines which describes why monotherapy is not used to treat this disease with the existing drugs available. That’s not to say that somebody won’t develop a single drug that puts sufficient selective pressure on the virus to make it go extinct.

In conclusion, I am forced to find that either kleinman is a liar and/or a troll, or too unable/unwilling to learn and understand his falsity to be worth interacting with.

Taffer, you are a graduate student in the theory of evolution and you don’t understand the fundamentals of your own theory. You demonstrated this when said this:

There are three antiretrovirals because the overall effect of all three is to more effectively inhibit the replication of the viron, not to slow evolution.

Fitness is measured by the ability of a creature to reproduce. You remember “survival of the fittest”? Ev and the example of treatment of HIV both demonstrate how multiple selection pressures slow evolution. I don’t mind discussing these issues with evolutionists who don’t understand the fundamentals of their own theory.

What do you predict the number of generations for the perfect creature for the G=64k case? Binding site width is 6 for this case.

Now Rfrequency = Rcapacity, so I think it will take a very long time. No idea.

Well, when you go to extrapolate the population data to genomes where Rfrequency >= Rcapacity, how do you propose to estimate the generations for convergence for these cases.

I propose that we do a series where we see if population can overcome the Rcapacity effect. The only possible cases we can run require relatively short genomes. Perhaps if we did a series with binding site width of 3 or 4 would give us short enough genomes that we can see whether larger populations will overcome the Rcapacity effect.

If Paul modified ev to evolve two different sets of binding sites simultaneously with six selection conditions, I would be very surprised if this model would ever converge no matter what length genome was used.

The problem with this idea is that we know precisely how such a creature would converge: Both transcription factors would evolve to recognize exactly the same binding site sequence. The model would have to be much more complex to force the transcription factors to be independent. In particular the binding sites would have to control genes that performed life-sustaining functions, so that when one transcription factor started interfering with the other's binding site, selection pressure would stop it from doing so. This would allow the two transcription factors to evolve independently.

That’s an interesting argument, Delphi argues the exact opposite. He says that selection pressures exist that transform identical genes into different genes. You are saying that your selection conditions will evolve two identical set of binding sites.

If Kleinman is correct, then no transcription factors have ever evolved. They were all poofed into existence.

If I am correct about what ev shows and the real life example of the treatment of HIV, what I have shown is that your theory is mathematically impossible. I have said nothing about how transcriptions factors came to be, other than I’m the annoying creationist.