Annoying creationists

[fishbob]

This is not my mathematics. This is Dr Schneider’s mathematics and it was peer reviewed and published in Nucleic Acids Research. I only did what Dr Schneider called for in his publication and the results show that your theory is mathematically impossible.

Weaselry incarnate.

 

[kleinman]

Annoying Creationists

Now isn’t that interesting. Paul’s online version does not have this feature. I guess Paul is too fatigued to put that version up on the net.

Lie #39. Look harder, dude.

I see you’ve added a new button since I last used your java version of ev. I’m going to enjoy co-opting this new feature to prove your Rcapacity problem has nothing to do with the weight matrix being able to locate binding sites and that the failure of the program to converge has everything to do with spurious binding sites controlling the selection process. Paul, how does it feel to have your own computer programming work used to disprove the theory of evolution? When are you going to simulate the evolution of two sets of different binding sites simultaneously? Or will that make the stylized ev computer model to realistic for the theory of evolution? You better take some vitamins or start taking power naps because the new features you have put in ev gives a new direction for parametric studies. I like number crunching with ev. Every time I run a case with ev I hear the theory of evolution go crunch. Ev is that gift that keeps on giving. Thank you, Paul.

This is not my mathematics. This is Dr Schneider’s mathematics and it was peer reviewed and published in Nucleic Acids Research. I only did what Dr Schneider called for in his publication and the results show that your theory is mathematically impossible.

Weaselry incarnate.

Stop being such a cry baby just because someone can take your mathematics and do a parametric study and show your silly theory is mathematically impossible.

 

[Paul C. Anagnostopoulos]

I see you’ve added a new button since I last used your java version of ev.

You haven't used it since November 2005?

I’m going to enjoy co-opting this new feature to prove your Rcapacity problem has nothing to do with the weight matrix being able to locate binding sites and that the failure of the program to converge has everything to do with spurious binding sites controlling the selection process.

Alan, if the binding sites are too narrow to allow a unique sequence logo, then spurious bindings are controlling the selection process. I agree with that. What you have to demonstrate is your claim that (a) at some point evolution abruptly stops; and (b) this stoppage is not due to Rcapacity. Simply run some experiments where evolution stops when Rcapacity is not an issue.

If you refuse to acknowledge that there is any Rcapacity problem at all, then you're just being silly. Obviously, for example, a binding site width of 1 base will not allow a unique sequence logo to emerge at binding sites. It's also gonna be awfully tough for 2-base sites to work. You need enough capacity in the sites to form a pattern that isn't present anywhere else.

Also, keep in mind that the Rcapacity problem is an artifact of the Ev model and is not a problem in nature.

~~ Paul

 

[Paul C. Anagnostopoulos]

These parameters show a cool result:

population 64
genome size 256
bindings sites 16
weight/site width 2/2
Rcapacity = Rfrequency = 4

After about 171,000 generations, the perfect creature evolves with a sequence logo of GG. There is not a single GG anywhere in the genome except at the binding sites.

Let's increase the genome size to 512 and see if we can conquer Rcapacity ...

No luck after 2,500,000 generations.

~~ Paul

 

[kleinman]

Annoying Creationists

I see you’ve added a new button since I last used your java version of ev.

You haven't used it since November 2005?

What can I say Paul, I’m not perfect. If I had noticed that feature, this discussion would have been much shorter. Anyway, it’s fun talking with you guys. It reminds me of my elementary school playground days.

I’m going to enjoy co-opting this new feature to prove your Rcapacity problem has nothing to do with the weight matrix being able to locate binding sites and that the failure of the program to converge has everything to do with spurious binding sites controlling the selection process.

Alan, if the binding sites are too narrow to allow a unique sequence logo, then spurious bindings are controlling the selection process. I agree with that. What you have to demonstrate is your claim that (a) at some point evolution abruptly stops; and (b) this stoppage is not due to Rcapacity. Simply run some experiments where evolution stops when Rcapacity is not an issue.

Paul, I’m going to write this as plainly as I can. It is the competing selection processes which halt the convergence of ev. The exact reason it occurs at genome lengths with your Rcapacity value, I don’t know but a more careful study of the model should give this answer. What I suspect is happening is as one selection condition is being satisfied, another selection condition is being overwhelmed and the system oscillates between the different selection conditions, never being able to satisfying all three at once. So since you pointed out that you could vary the weights for mistakes a chose a case which does not converge with all three selection conditions imposed.

Wow, am I fatigued. It took me 3 cases to demonstrate that it is competing selection conditions that slow ev’s convergence. The case I used consists of G=16384, binding sites=6, weight width=5, Rfreq=10, Paul’s infamous Rcapacity=2*site width=10. With the weights for each mistake, (missed site, spurious binding in gene, spurious binding outside gene), set to 1, this case fails to converge. But for varying weights you get the following results:
missed site/spurious binding-gene/spurious binding outside gene/gens to perfect creature
1/0/0/7
0/1/0/223
0/0/1/233
So, a case with three different selection conditions will not converge when all three conditions must be satisfied simultaneously. While if you select based on one condition at a time it takes only 7 generations to evolve the binding sites, 223 generations to eliminate spurious binding within the gene and only 233 generations to eliminate spurious binding outside the gene. These cases take so few generations to evolve, it is instructive to step through the evolution and watch the mistakes go to zero. Combining these three selection conditions is exactly analogous to using triple anti-HIV drugs to prevent the evolution of drug resistant strains. Paul’s this is the explanation for your Rcapacity problem. I appreciate your fine programming work.

If you refuse to acknowledge that there is any Rcapacity problem at all, then you're just being silly. Obviously, for example, a binding site width of 1 base will not allow a unique sequence logo to emerge at binding sites. It's also gonna be awfully tough for 2-base sites to work. You need enough capacity in the sites to form a pattern that isn't present anywhere else.

Paul, certainly there is an Rcapacity problem for the theory of evolution. What that Rcapacity problem is competing selection conditions.

Also, keep in mind that the Rcapacity problem is an artifact of the Ev model and is not a problem in nature.

Paul, you couldn’t be more wrong. Competing selection processes occur all the time in nature. The problem for you evolutionists is understanding that these competing selection processes stop evolution. Dr Schneider’s model has accurately captured a real physical phenomenon with his three different selection condition.

Perhaps if you consider the question from the opposite point of view, you may find a way of understanding this Rcapacity issue. Why do small genomes still evolve despite three selection conditions?

These parameters show a cool result:

population 64
genome size 256
bindings sites 16
weight/site width 2/2
Rcapacity = Rfrequency = 4

After about 171,000 generations, the perfect creature evolves with a sequence logo of GG. There is not a single GG anywhere in the genome except at the binding sites.

Let's increase the genome size to 512 and see if we can conquer Rcapacity ...

No luck after 2,500,000 generations.

Paul, run the case I have shown above setting 2 of the three selection conditions weights to zero and then step through all three cases and watch the mistakes disappear. Perhaps then you will understand why your theory of evolution is mathematically impossible. The mistakes do not disappear when you apply all three selection conditions simultaneously. Competing selection conditions stop evolution. Your model shows this and this is also seen in the real world.

 

[Paul C. Anagnostopoulos]

Wow, am I fatigued. It took me 3 cases to demonstrate that it is competing selection conditions that slow ev’s convergence.

Alan, please, stop spewing this claim over and over. I agree that a simpler selection requirement is likely to result in a faster convergence. Whoopie crap.

Your interesting claim is that evolution stops dead even when Rcapacity isn't an issue. Demonstrate it.

Paul’s this is the explanation for your Rcapacity problem.

Rcapacity is irrelevant in the experiments you ran, because the calculation of Rfrequency assumes all three mistake points greater than 0. Rfrequency, Rsequence, and Rcapacity are all irrelevant.

Paul, you couldn’t be more wrong. Competing selection processes occur all the time in nature.

But they have nothing to do with Rcapacity.

~~ Paul

 

[delphi_ote]

Oh really Delphi? Are you going to claim that the mathematics of mutation and selection is linear? Are you going to claim that selection processes can be superimposed and you obtain the same evolutionary events when the selection processes are applied independently? Perhaps you could explain to us when triple anti-HIV medications are used, it is much less likely to evolve drug resistant strains of HIV than when using the same drugs used in sequence, one at a time.

Wow. More irrelevant word salad. You can't arbitrarily use technical jargon from another field out of context to describe whatever you like and expect people to respect you. Once again, it doesn't make you sound intelligent. It's patently obvious you're trying to claim authority on a subject you don't understand by using terminology from your own field as a smoke screen.

I’m pretty sure I have more experience and training in solving non-linear mathematical problems than you.

Congratulations. You're an engineer. That's fantastic. Despite that elitist feeling that came packaged with your diploma, you are not omniscient. While you can probably snowball the other nontechnical people at your office with this behavior, we're not falling for it here.

Also, while I'm not going to get into a credentials pissing contest with you, you should be probably be careful saying things like that. Especially on this forum. You're talking to a lot of very educated people.

 

[Paul C. Anagnostopoulos]

In fairness to Kleinman, I ran one more experiment with his parameters, and here is the result:

missed site/spurious binding-gene/spurious binding outside gene/gens to perfect creature
0/0/0/1

Extraordinary! Exploring this in greater depth, I increased the genome size to 256,000 bases:

0/0/0/1

Absolutely amazing! We have just shown that if there are no selective pressures at all, a perfect creature evolves immediately, regardless of genome size and Rcapacity.

~~ Paul

 

[kjkent1]

Paul, I’m going to write this as plainly as I can. It is the competing selection processes which halt the convergence of ev. The exact reason it occurs at genome lengths with your Rcapacity value, I don’t know but a more careful study of the model should give this answer. What I suspect is happening is as one selection condition is being satisfied, another selection condition is being overwhelmed and the system oscillates between the different selection conditions, never being able to satisfying all three at once. So since you pointed out that you could vary the weights for mistakes a chose a case which does not converge with all three selection conditions imposed.

Wow, am I fatigued. It took me 3 cases to demonstrate that it is competing selection conditions that slow ev’s convergence. The case I used consists of G=16384, binding sites=6, weight width=5, Rfreq=10, Paul’s infamous Rcapacity=2*site width=10. With the weights for each mistake, (missed site, spurious binding in gene, spurious binding outside gene), set to 1, this case fails to converge. But for varying weights you get the following results:
missed site/spurious binding-gene/spurious binding outside gene/gens to perfect creature
1/0/0/7
0/1/0/223
0/0/1/233
So, a case with three different selection conditions will not converge when all three conditions must be satisfied simultaneously. While if you select based on one condition at a time it takes only 7 generations to evolve the binding sites, 223 generations to eliminate spurious binding within the gene and only 233 generations to eliminate spurious binding outside the gene. These cases take so few generations to evolve, it is instructive to step through the evolution and watch the mistakes go to zero. Combining these three selection conditions is exactly analogous to using triple anti-HIV drugs to prevent the evolution of drug resistant strains. Paul’s this is the explanation for your Rcapacity problem. I appreciate your fine programming work.

If you set any of the weights to zero, the program simply doesn't evaluate that mistake at all -- but it still produces the mistakes. So, when the program reports convergence or a perfect creature, the actual state of the genome is still random noise.

Any result obtained with a zero weight is meaningless. I suggest you try your experiments using 1 or 100 for a mistake weight.

 

[kleinman]

Annoying Creationists

Wow, am I fatigued. It took me 3 cases to demonstrate that it is competing selection conditions that slow ev’s convergence.

Alan, please, stop spewing this claim over and over. I agree that a simpler selection requirement is likely to result in a faster convergence. Whoopie crap.

These are not simpler selection requirements, there are three selection requirements and when imposed one at a time, evolution occurs rapidly. When all three are imposed simultaneously, no evolution occurs.

Your interesting claim is that evolution stops dead even when Rcapacity isn't an issue. Demonstrate it.

Paul, what you are calling Rcapacity is simply the failure of the model to converge due to three competing selection processes. When each process is applied individually, you eliminate each of the mistakes in turn. When you apply all three selection conditions simultaneously, the evolutionary process slows then ultimately stops as you lengthen the genome in the model. Paul, run the three cases I described in the previous post and single step through the simulation and watch what happens to mistakes when you impose only a single selection condition. The mistakes quickly disappear when you have only a single selection condition but when you have all three conditions, nothing evolves. If you want to satisfy yourself as to what is happening, store each of the three errors (missed binding sites, spurious sites in the gene and spurious sites outside the gene) at each step of the simulation and then plot that data. You can then determine precisely which mistakes are driving the selection process.

Paul’s this is the explanation for your Rcapacity problem.

Rcapacity is irrelevant in the experiments you ran, because the calculation of Rfrequency assumes all three mistakes points greater than 0. Rfrequency, Rsequence, and Rcapacity are all irrelevant.

Paul, Dr Schneider uses the following definition for Rfrequency:

Rfreqency = -log2(gamma/G)

Rfrequency is dependent only the number of binding sites and the length of the genome. The selection process is not in the definition of Rfrequency.

Your definition for Rcapacity is as follows:

Rcapacity = 2*binding site width

The selection process is not in your definition of Rcapacity.

What you are missing is it’s the multiple selection conditions that slows and ultimately stops evolution in ev as you lengthen the genome. This is a realistic feature of ev. This effect is seen in reality.

Paul, you couldn’t be more wrong. Competing selection processes occur all the time in nature.

But they have nothing to do with Rcapacity.

Paul, your Rcapacity concept has confused you and led you astray in understanding what your model is demonstrating. What slows and ultimately stops ev from evolving binding sites (or the lack of spurious binding sites) with larger genomes is the competing selection processes.

So, your model shows that you can still evolve binding sites on short genomes with 3 selection conditions or in order to evolve binding sites on larger genomes, you can have only one selection condition at a time. Your model certainly imposes some unusual conditions on the theory of evolution.

Also, while I'm not going to get into a credentials pissing contest with you, you should be probably be careful saying things like that. Especially on this forum. You're talking to a lot of very educated people.

I definitely don’t want to get into a pissing contest with someone who always needs a drink, I concede that honor to you.

In fairness to Kleinman, I ran one more experiment with his parameters, and here is the result:

missed site/spurious binding-gene/spurious binding outside gene/gens to perfect creature
0/0/0/1

Extraordinary! Exploring this in greater depth, I increased the genome size to 256,000 bases:

0/0/0/1

Absolutely amazing! We have just shown that if there are no selective pressures at all, a perfect creature evolves immediately, regardless of genome size and Rcapacity.

Paul, why don’t you run some experiments setting only 1 of the three selection conditions to 0? What do you think will happen?

If you set any of the weights to zero, the program simply doesn't evaluate that mistake at all -- but it still produces the mistakes. So, when the program reports convergence or a perfect creature, the actual state of the genome is still random noise.

That is not the case. If you set the selection weight to 1 for binding site mistakes and the other two weight conditions to 0, you evolve genomes with binding sites. If you set the selection weight to 1 for spurious bindings in the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding in the gene. If you set the selection weight to 1 for spurious binding outside the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding outside the gene. The rest of the genome does remain random.

Any result obtained with a zero weight is meaningless. I suggest you try your experiments using 1 or 100 for a mistake weight.

The case of setting a weight to zero is totally meaningful. What it means is that you have removed a selective pressure. This would be analogous to treating someone with HIV with only one anti-viral drug at a time rather than using 2 or 3 anti-viral drugs at a time. Three anti-viral drugs is equivalent to applying three selective pressures simultaneously, two anti-viral drugs is equivalent to applying two selective pressures simultaneously, and one anti-viral drug is equivalent to applying a single selective pressure at a time. The more anti-viral drugs used simultaneously, the less likely you are to evolve a drug resistant strain of the virus. Multiple selective pressure slow and ultimately stop evolution, ev shows this and it occurs in reality.

Did I forget to remind you, you have no selection process to evolve a gene from the beginning?

 

[Mercutio]

Did I forget to remind you, you have no selection process to evolve a gene from the beginning?

I really had hoped that you would have dropped this strawman. You have been reminded time and again that you are conflating natural selection with abiogenesis. You have been reminded that these are two separate processes, and that ev only models natural selection. It appears that dawn may be beginning to break, and that you are seeing that your complaints about ev are (maybe) being answered (time will tell); perhaps this retreat to "from the beginning" is a sign of this.

I did find another computer program that claims to model abiogenesis! Sadly, the guy who wrote it has moved on to other things, and it is no longer actively supported (he says he will provide the program to interested people; I have no computer expertise so I will not be one to ask for it). If you wish to model "from the beginning", perhaps you will have to try to disprove abiogenesis using an abiogenesis program, having failed to disprove evolution using an evolution program.

 

[kleinman]

Annoying Creationists

Did I forget to remind you, you have no selection process to evolve a gene from the beginning?

I really had hoped that you would have dropped this strawman. You have been reminded time and again that you are conflating natural selection with abiogenesis. You have been reminded that these are two separate processes, and that ev only models natural selection. It appears that dawn may be beginning to break, and that you are seeing that your complaints about ev are (maybe) being answered (time will tell); perhaps this retreat to "from the beginning" is a sign of this.

Let’s see if I can understand your logic. Genes arose in abiogenesis and given enough time these genes evolved into evolutionists. I will try not to confuse genes and evolutionists. In addition, I will try not to confuse abiogenesis with the theory of evolution. One is utter nonsense, the other is simply mathematically impossible. You can guess which one is which.

I did find another computer program that claims to model abiogenesis! Sadly, the guy who wrote it has moved on to other things, and it is no longer actively supported (he says he will provide the program to interested people; I have no computer expertise so I will not be one to ask for it). If you wish to model "from the beginning", perhaps you will have to try to disprove abiogenesis using an abiogenesis program, having failed to disprove evolution using an evolution program.

It seems like Dr Schneider has moved on to other things as well. I guess more than 20 years with ev is enough for him. Paul wants to move on to other things. Well, no, Paul has an Rcapacity fixation and can’t move on to other things. Me, I’m having too much fun with ev to move on right now. I actually think ev can be useful in modeling the evolution of drug resistance. Is it possible the ev is more than just a stylized mathematical model of random point mutation and natural selection? Will Paul overcome his fixation on Rcapacity? Tune in next week for the next episode of “As the Genome Evolves”.

 

[kjkent1]

That is not the case. If you set the selection weight to 1 for binding site mistakes and the other two weight conditions to 0, you evolve genomes with binding sites. If you set the selection weight to 1 for spurious bindings in the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding in the gene. If you set the selection weight to 1 for spurious binding outside the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding outside the gene. The rest of the genome does remain random.

No, you evolve crap in either case. With missing bindings set to 0, and "pause on perfect creature" enabled, the binding site region is filled with random noise at instantiation of the program, and then because there is no selection against missing binding sites, the binding site region remains random noise and immediately reports a perfect creature, even though when you observe the resulting genome, it's obviously random noise. Just look at the output. And, if you don't pause on a perfect creature, then the program makes mistakes, but never checks to see if there are missing binding sites, and the result is no convergence and the genome remains random noise.

Similarly, if you set missing bindings to 1 and spurious inside and out to 0, then at instantiation the program fills up with spurious bindings, but it doesn't recognize them, so once again, a perfect creature is reported regardless of any missing bindings and even though the genome is filled with spurious bindings.

If these meaningless results aren't obvious to you from observing the software in operation, then you are either not running the program, or you are not Alan Kleinman, Ph.D, because if you were, then you would instantly realize that your conclusion is totally incorrect.

The case of setting a weight to zero is totally meaningful. What it means is that you have removed a selective pressure. This would be analogous to treating someone with HIV with only one anti-viral drug at a time rather than using 2 or 3 anti-viral drugs at a time. Three anti-viral drugs is equivalent to applying three selective pressures simultaneously, two anti-viral drugs is equivalent to applying two selective pressures simultaneously, and one anti-viral drug is equivalent to applying a single selective pressure at a time. The more anti-viral drugs used simultaneously, the less likely you are to evolve a drug resistant strain of the virus. Multiple selective pressure slow and ultimately stop evolution, ev shows this and it occurs in reality.

The case of a setting of zero is totally meaningless -- you are introducing a bug into the program by preventing the program from selecting for a mistake that the program places in the genome.

The final result of setting any one of the mistake weights to zero is a genome filled with gibberish.

 

[kleinman]

Annoying Creationists

That is not the case. If you set the selection weight to 1 for binding site mistakes and the other two weight conditions to 0, you evolve genomes with binding sites. If you set the selection weight to 1 for spurious bindings in the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding in the gene. If you set the selection weight to 1 for spurious binding outside the gene and the other two weight conditions to 0, you evolve genomes with no spurious binding outside the gene. The rest of the genome does remain random.

No, you evolve crap in either case. With missing bindings set to 0, and "pause on perfect creature" enabled, the binding site region is filled with random noise at instantiation of the program, and then because there is no selection against missing binding sites, the binding site region remains random noise and immediately reports a perfect creature, even though when you observe the resulting genome, it's obviously random noise. Just look at the output. And, if you don't pause on a perfect creature, then the program makes mistakes, but never checks to see if there are missing binding sites, and the result is no convergence and the genome remains random noise.

With missing binding sites set to 0, spurious bindings inside the gene set to 1 and spurious binding outside the gene set to 0, and set “pause on perfect creature” does not yield a random binding site region. It yields a binding site region without spurious bindings. The non-binding site region remains random. Likewise, with missing binding sites set to 0, spurious binding inside the gene set to 0 and spurious bindings outside the gene set to 1, and set “pause on perfect creature” does not yield a random non-binding site region. It yields a non-binding site region without spurious bindings. The binding site region in this case does remain random.

Similarly, if you set missing bindings to 1 and spurious inside and out to 0, then at instantiation the program fills up with spurious bindings, but it doesn't recognize them, so once again, a perfect creature is reported regardless of any missing bindings and even though the genome is filled with spurious bindings.

However, the binding sites are recognized in the binding site region.

If these meaningless results aren't obvious to you from observing the software in operation, then you are either not running the program, or you are not Alan Kleinman, Ph.D, because if you were, then you would instantly realize that your conclusion is totally incorrect.

What you are having trouble recognizing is that each selection condition is leading to particular sequences of bases. In one case the selection condition yields sequences that give binding sites, in the other two selection conditions it leads to sequences without binding sites.

The case of setting a weight to zero is totally meaningful. What it means is that you have removed a selective pressure. This would be analogous to treating someone with HIV with only one anti-viral drug at a time rather than using 2 or 3 anti-viral drugs at a time. Three anti-viral drugs is equivalent to applying three selective pressures simultaneously, two anti-viral drugs is equivalent to applying two selective pressures simultaneously, and one anti-viral drug is equivalent to applying a single selective pressure at a time. The more anti-viral drugs used simultaneously, the less likely you are to evolve a drug resistant strain of the virus. Multiple selective pressure slow and ultimately stop evolution, ev shows this and it occurs in reality.

The case of a setting of zero is totally meaningless -- you are introducing a bug into the program by preventing the program from selecting for a mistake that the program places in the genome.

Nope, setting a weighting factor to zero is equivalent to removing a selective pressure. It is identical to removing one or two of the three antiviral drugs used in HIV treatment. You are removing selective pressures from the virus.

The final result of setting any one of the mistake weights to zero is a genome filled with gibberish.

Nope, if you set any one of the three selection conditions to a non-zero value, you end up with genomes that either have binding sites, or a portion of their genomes devoid of binding sites. In any case, the genomes are no longer completely random when you set any selection condition to a non-zero value and set “pause on perfect creature”.

This is fun, we need to talk about this some more next week.

 

[kjkent1]

With missing binding sites set to 0, spurious bindings inside the gene set to 1 and spurious binding outside the gene set to 0, and set “pause on perfect creature” does not yield a random binding site region. It yields a binding site region without spurious bindings. The non-binding site region remains random. Likewise, with missing binding sites set to 0, spurious binding inside the gene set to 0 and spurious bindings outside the gene set to 1, and set “pause on perfect creature” does not yield a random non-binding site region. It yields a non-binding site region without spurious bindings. The binding site region in this case does remain random.

However, the binding sites are recognized in the binding site region.

What you are having trouble recognizing is that each selection condition is leading to particular sequences of bases. In one case the selection condition yields sequences that give binding sites, in the other two selection conditions it leads to sequences without binding sites.

Nope, setting a weighting factor to zero is equivalent to removing a selective pressure. It is identical to removing one or two of the three antiviral drugs used in HIV treatment. You are removing selective pressures from the virus.

Nope, if you set any one of the three selection conditions to a non-zero value, you end up with genomes that either have binding sites, or a portion of their genomes devoid of binding sites. In any case, the genomes are no longer completely random when you set any selection condition to a non-zero value and set “pause on perfect creature”.

This is fun, we need to talk about this some more next week.

OK, meanwhile, perhaps Paul will provide his own opinion. If I'm wrong, that's fine, but it seems pretty obvious to me that setting any of the mistake weights to zero allows the program to introduce mistakes in the genome and then fail to select for those errors. This is nearly the same as turning selection off.

 

[delphi_ote]

I definitely don’t want to get into a pissing contest with someone who always needs a drink, I concede that honor to you.

After reading your same tired arguments couched in the same mathematical word salad in the same annoying formatting while you make the same lame joke about drinking for pages on end, even Muhammad would need a drink.

 

[fishbob]

Stop being such a cry baby just because someone can take your mathematics and do a parametric study and show your silly theory is mathematically impossible.

Even more weaselly than the last time.

 

[Taffer]

Do you want to explain how the overall fitness will evolve a gene from the beginning? How do you include what you have just said in a mathematical model of mutation and selection?

What ev reveals about competing selection processes is demonstrated with the use of multiple antimicrobials. Evolution is slowed if not halted in the HIV by using the three antimicrobials. The prevention of the changes in the genotype prevent the change in the phenotype.

Did I say anything about ev? Or abiogenesis? No. I was simply pointing out that you were wrong in your conception of natural selection.

 

[Taffer]

Let’s see if I can understand your logic. Genes arose in abiogenesis and given enough time these genes evolved into evolutionists. I will try not to confuse genes and evolutionists. In addition, I will try not to confuse abiogenesis with the theory of evolution. One is utter nonsense, the other is simply mathematically impossible. You can guess which one is which.

Abiogenesis and evolution are two completely different things. Now I think I am beginning to understand why you fail to grasp this. You are using a popular definition of a gene, not one that actual geneticists use.

 

[Paul C. Anagnostopoulos]

Paul, Dr Schneider uses the following definition for Rfrequency:

Rfreqency = -log2(gamma/G)

Rfrequency is dependent only the number of binding sites and the length of the genome. The selection process is not in the definition of Rfrequency.

Sigh. Rfrequency is a calculation of how many bits of information are required to uniquely identify the binding sites from among the rest of the genome. Implicit in this calculation is the idea that you are actually going to identify the binding sites from among the rest of the genome. If you are not, Rfrequency is irrelevant. Setting any of the three mistake counts to zero means you are not identifying the binding sites.

Your definition for Rcapacity is as follows:

Rcapacity = 2*binding site width

The selection process is not in your definition of Rcapacity.

Rcapacity is only interesting when compared to Rfrequency. Since Rfrequency is irrelevant in your scenarios, so is Rcapacity.

~~ Paul