Annoying creationists

[Dr Richard]

HIV Selection pressure database

Selection pressures in HIV:

Here

can be found a database of selection pressures acting on HIV.

The researchers have calculated the selection pressure for mutations at every amino acid codon for HIV protease and reverse transcriptase.

At every aa codon.

And studied interactions between individual point mutations.

That is a lot more that the magic "three" selection pressures acting in Kleinman fantasy land.

 

[Dr Adequate]

My view on this issue is that once you get beyond a single point mutation you are already starting to enter the realm of macroevolution.

Amino acid mutations at 90 displayed strong, unconditional positive selection, indicating that they directly cause drug resistance. In contrast, mutations at 10 are negatively selected in the absence of the 90 mutation, but become positively selected in the presence of the 90 mutation (Figure 3). These results closely match previous experimental studies showing that mutations at 90 cause drug resistance, while mutations at 10 have an accessory effect of compensating for the destabilizing effect of mutations at 90.

Reality shows that the mathematics does predict properly the behavior of mutation and selection.

It sure does.

 

[kleinman]

Annoying Creationists

is that plus 1 for god?

I think he's been mislead by the metaphor of a fitness "landscape" into thinking of the output of the fitness function as the "height" of the "landscape", hence another dimension.

Adebz, when you got your Phd in amathematics, didn’t they tell you that the dependent variable maps out as another dimension, you silly amathematician.

No, if you understood the mathematics, you would understand that the extra dimension is for the mapping of the value of the fitness function. Don’t alchemical engineers have to study mathematics?

I was making fun of you...

You were making fun of me by putting both of your feet in your mouth. You certainly are a twofooted humorist.

But your insistance on the defining demensions of the fitness function has presented a new insight into your math. We can ever only use 1 set of dimensions when describing a system. It is impossible to represent a feature using different axes.

Joobequate, you don’t understand the mathematics of mutation and selection when using real coordinates. How do you expect to understand the mathematics in some abstract, transformed coordinate system?

Why joobequate, you evolutionary mathematicians are brilliant. But you are not so brilliant to catch an error that both Thabiguy and I have made with respects to the fitness landscape. I had initially thought the fitness landscape has the number of dimensions equal to the number of selection pressures and Thabiguy assigned the number of dimensions to this space as equal to the number of traits that a genome has. Actually, the fitness space has a number of dimensions equal to the genome length plus one. It can actually be greater than this if you allow for insertions or gene duplications. Let’s see if your evolutionist mathematically challenged mind can grasp this?

The fitness landscape actually cannot be meaningfully plotted for a function of the genome itself, as genomes are discrete entities. That's why I preferred to call the fitness landscape a "multi-dimensional graph of fitness function with respect to small changes in the genome that correspond to these individual traits (assuming they are independent)".

The fitness landscape only is meaningful when you understand that the independent variables in the fitness function are the individual loci in the genome. It doesn’t matter that genomes are discrete entities. Each genetic sequence for a given genome maps out on the fitness landscape as a point on this surface.

That's because changes in the genome that affect traits that are not selected for/against do not change the fitness function and are therefore not relevant for the graph of that particular fitness function (it will be constant in those additional dimensions). I have no problem if you want to consider these additional dimensions as well. Add as many as you like.

You are wrong on this point. Neutral mutations represent possible steps on the fitness landscape surface for finding an optimum. Every mutation to a genome represents a movement on the fitness landscape.

The number of dimensions in the fitness landscape is determined by the number of loci in the genome. These loci represent the independent variables in the fitness function. The fitness function is the dependent variable. The selection conditions define the shape of the fitness function. Specify the genome sequence, and the selection pressures and you have fixed the value of the fitness function.

When selection pressures are fixed and you have an optimum genetic sequence for the virus, natural selection has a stabilizing effect on the genetic sequence. However, when you introduce medical treatment of HIV (a new selection pressure), natural selection now becomes directional. If you have only a single selection pressure (monotherapy) it is much easier for natural selection to find a new optimum on the fitness landscape. The search only requires a single direction. However, when you introduce combination therapy (multiple selection pressures), it is much more difficult for natural selection to find new optimums for each of these selection pressures simultaneously. Several directions must be searched simultaneously and this confounds the search.

Incorrect. A mutation that occurs in a dimension that is not being selected for has no effect on the fitness. What matters are those mutations that occur in the dimensions that are being selected for. There are as many of these dimensions as there are traits considered in the fitness function - the number of selection pressures.

It is you who is wrong on this point; you can not ignore neutral (or silent) mutation. Let me explain to you why you are wrong by example. Consider a codon UUA which codes for the amino acid leucine. You get a mutation at the third position which changes the codon to UUG which still codes for leucine. This would be a silent mutation and there would be no selection for or against since the polypeptides produced would be identical for both codons. Now you get a second mutation to position 2 of the codon which puts a G in the sequence. UGA gives a stop codon, UGG codes for tryptophan. The first mutation is silent; the second mutation leads to different polypeptide sequences. You can not ignore silent mutations on the fitness landscape.

Mutations in a population "search" through all dimensions simultaneously. A dimension is "searched" whenever a mutation occurs in it. If a trait is being selected for, then when the mutation occurs that affects that trait, a "search" is performed in that direction. The probability that this will occur depends on the population size and the mutation rate, but not on the overall number of dimensions. In other words, the length of the genome does not generally affect the probability that a mutation occurs in one particular gene.

What you still don’t understand is that when you expand the number of dimensions (by increasing the genome length) you are expanding the search space which profoundly slows the search. Try out Dr Schneider’s ev program and it gives you a very nice demonstration of this affect. Only when you have a single selection condition does the length of the genome have little effect on the rate of evolution. As you increase the number of selection pressures, then the length of the genome has a profound effect on rate of evolution. I have given numerous real examples of this as well as the results of the ev simulation of mutation and selection.

If there are multiple dimensions selected for, then a mutation in either of it will affect the fitness function. How soon this will happen depends on the mutation rate; the resulting selection effect depends on the change in the fitness function. None of this depends on the number of dimensions, either total or those selected for.

The mutation rate has approximately a linear affect on the rate of evolution, it is the genome length and the number of selection pressures which dominate the rate of evolution. Try the ev computer simulation of mutation and selection, it demonstrates this mathematics very nicely. Once you do this, you will gain some understanding of the mathematics of mutation and selection.

You must realize that "searching" in this context does not mean going in a direction, exploring the landscape and finding the best value there. Perhaps you are confused by the picture in the Wikipedia article and are interpreting the "search" to mean climbing up the slope and "success" to mean that a local extreme is found. That's not the case; the "search" means searching in the local neighborhood for a direction in which there is any slope. Every time that a slope is found (a mutation occurs that changes the fitness function), the "search" succeeds. Every little arrow on that picture represents a successful "search". And because, as I tried to explain, the dimensions are "searched" simultaneously (the probability of a mutation affecting a trait does not depend on the length of the genome), the speed of that happening is in no way "confounded" by the number of dimensions; it doesn't depend on it.

Selection imposes direction on the search process.

But the good news is that your understanding of the fitness landscape, while still not accurate, has improved.

Are you sure it is not you who is starting to understand this?

Post the quote(s) from your peer-reviewed science that you think supports your point.

The point of a peer-reviewed science article is that it reports a hypothesis and the results of an experiment intended to validate or reject that hypothesis (you must have been absent from class the days that your professors explained how scientific research is conducted). If I post an article that shows evidence of HGT, then that’s what it shows. There are only three possible interpretations of the existence of such evidence: (1) two lineages obtained identical genetic sequences by random chance; (2) two lineages obtained identical genetic sequences via some natural process (e.g., retroviral insertion, crossbreeding, etc.); (3) two lineages obtained identical genetic sequences by supernatural/magical intervention.

I guess you can’t post the quotes. Anybody care for another serving of red herring, string cheese and whine? At least Dr Richard posts the quotes.

Let’s consider a real example of this phenomenon, sickle cell anemia where the heterozygote is more fit in a particular environment. The selection pressure is due to the malaria parasite. I believe that is a single selection pressure.

1. "The gene conferring the Duffy blood group is the most striking example; people of this blood type are completely immune to Plasmodium vivax blood stage infection, as they lack the relevant receptor on RBC membranes."

I never said that there couldn’t be more than a single response to a selection pressure.

2. "Inducible nitric oxide synthase 2 (iNOS2) is an enzyme that modulates nitric oxide (NO) production, ultimately affecting malarial immunity"

So, recombination and selection to malaria selects people who are better producers of this enzyme.

3. "A polymorphism restricting the affinity of Fc gamma RII to IgG1 and IgG3 has been investigated in Western Kenyans, and correlates with P. falciparum immunity"

And this shows what?

4. "Heterozygosity for an allelic variant of CD36 is associated with protection from severe malaria in Africans. Importantly, MHC Class I B53 (MHC-B53), and II DQB1FNx010501 and DRB1FNx011302 alleles are associated with protective clinical responses in African populations."

I didn’t ignore this. This was my explanation to Taffer’s example of selection for heterozygotes.

5. "Natural immunity to malaria takes years to acquire, at least partly due to a very effective immune evasion strategy mediated by naturally occurring variants of the same antigenic epitopes, capable of inhibiting memory T cells [Figure - 2]. This so-called ′altered peptide ligand′ (APL) antagonism affects specific cell lysis and lymphocyte proliferation, as well as cytokine production."

Does it take years to acquire be heterozygotic with hemoglobin S allele?

6. "It is reasonable to assume that DCs also play a critical role in initiating immune responses to malaria"

The point is that creatures with complex immune systems can have a variety of responses to an infection like malaria. Some of these responses are modified by recombination and selection and some by mutation and selection.

The selection pressure is due to the malaria parasite. I believe that is a single selection pressure.

The "malaria parasite" is exerting a selection pressure. It is not single, unless you sum selection pressures.

I’m going to admit you are correct. Malaria affects multiple genetic systems. So what is the selection pressure that sums up to evolving reptiles into birds?

Thabiguy, why don’t you explain to us why combination therapy of HIV doesn’t evolve resistance to all medications as rapidly as monotherapy?

Because 1) viral load is dramatically reduced, thus the probability that any given mutation will occur in the population, 2) selection pressure for any individual resistance is lower in combination therapy than in monotherapy, because it no longer offers such a significant survival advantage. Only developping resistance to all medications is strongly selected for, and that is less likely. But it will eventually occur.

Monotherapy can and does dramatically reduce viral loads yet still leads to rapid emergence of resistance. What makes you think that the intensity of selection pressures for the individual drugs in combination therapy are lowered? Resistance may still evolve in combination therapy but at a much slower rate than with monotherapy.

That is a lot more that the magic "three" selection pressures acting in Kleinman fantasy land.

I didn’t notice the quote from your HIV database where they recommend returning to monotherapy.

You also still haven’t explained why chimpanzees and humans produce different preproinsulin yet produce identical insulin. This all despite you claim we descended from a common ancestor.

 

[Dr Adequate]

Adebz, when you got your Phd in amathematics, didn’t they tell you that the dependent variable maps out as another dimension, you silly amathematician.

The people who taught me math did not attempt to lie to me about the dimensionality of optimisation problems, because they were mathematicians, not puffed-up creationist windbags whining out lies about subjects of which they know nothing.

The question is, why are you trying to lie to me on this subject? I have a PhD in math, whereas, if this thread has taught you nothing else, it must surely have taught you that you know sweet damn all about math.

Therefore, your chances of successfully lying to me on this subject are fairly small, don't you think?

 

[kleinman]

Annoying Creationists

Adebz, when you got your Phd in amathematics, didn’t they tell you that the dependent variable maps out as another dimension, you silly amathematician.

The people who taught me math did not attempt to lie to me about the dimensionalisty of optimisation problems, because they were mathematicians, not puffed-up creationist windbags whining out lies about subjects of which they know nothing.

Are these the same people who taught you that the more conditions you are trying to optimize, the faster the process goes, you silly amathematician Adebz?

 

[joobz]

Joobequate, you don’t understand the mathematics of mutation and selection when using real coordinates. How do you expect to understand the mathematics in some abstract, transformed coordinate system?

Why do you keep on trying to make these lies? Do you honestly think anyone here is buying into your arguments?

What do you mean real coordinates vs abstract coordinates? The whole concept is an abstraction meant for ease of computation. You have made so many arbitrary distinctions (micro vs. macro evolution, number of selection pressures, which mutations are allowable for evolution, selection of coordinate systems, speciation, etc.), you seem to have lost touch with what is real and what has been your fabrication.

The amusing thing is, each of these distinctions were shown to not matter. Even when you have been allowed to set the stage, evolution was shown to be true. How does it feel that even when we allow to create the game, you still lose?

 

[delphi_ote]

longer genome lengths slow down the search

What the holy Hell are you talking about? By your logic, the more copies of a gene an organism has, the slower it will adapt to environmental stresses?!

huge populations do not markedly accelerate the search.

Doubling the size of the population doubles the chances of a given mutation in the population.

 

[Dr Adequate]

Are these the same people who taught you that the more conditions you are trying to optimize, the faster the process goes ...

No-one has taught me that.

What do you mean by this bizarre lie?

 

[Dr Adequate]

Doubling the size of the population doubles the chances of a given mutation in the population.

That could have been better phrased.

If you'd said "doubles the expected number of a given mutation per generation", then the halfwit would be less likely to misunderstand you.

Given his stupid blunders about the most trivial aspects of probability theory (see the FAQ) you can't be too careful.

 

[joobz]

No-one has taught me that.

What do you mean by this bizarre lie?

I believe he is propogating the same lie that:
Time to computationally model a system = (or is proportional to) the real world kinetics of the system.

 

[kleinman]

Annoying Creationists

Joobequate, you don’t understand the mathematics of mutation and selection when using real coordinates. How do you expect to understand the mathematics in some abstract, transformed coordinate system?

What do you mean real coordinates vs abstract coordinates? The whole concept is an abstraction meant for ease of computation. You have made so many arbitrary distinctions (micro vs. macro evolution, number of selection pressures, which mutations are allowable for evolution, selection of coordinate systems, speciation, etc.), you seem to have lost touch with what is real and what has been your fabrication.

It seems that when you studied alchemical engineering they neglected to teach you anything about the mathematics of coordinate transformations. For example, there are problems that are defined by space and time coordinates (real coordinates) that can be solve by mathematically combining space and time into a single mathematical coordinate (abstract coordinate). These are called similarity solutions. Don’t mistake your ignorance as my lying to you. You have plenty of ignorance of mathematics, especially the mathematics of mutation and selection.

longer genome lengths slow down the search

What the holy Hell are you talking about? By your logic, the more copies of a gene an organism has, the slower it will adapt to environmental stresses?!

The adaptation of organisms to environmental stresses is primarily done by recombination and natural selection, not mutation and natural selection. Recombination and natural selection can give rapid morphological changes, mutation and selection is a profoundly slow process. Ev shows this and reality shows this.

huge populations do not markedly accelerate the search.

Doubling the size of the population doubles the chances of a given mutation in the population.

Careful, you are making the same error the joobequate periodically reminds me of. That probability you speak of can only be approximated as additive with small populations. Think of it this way, if you keep doubling the population, then at some point you will have a probability greater than one. Myriad showed the proper way of computing this probability on the Evolutionisdead forum and population has less than an additive affect. This is also demonstrated in ev.

Are these the same people who taught you that the more conditions you are trying to optimize, the faster the process goes ...

No-one has taught me that.

Then you should understand that multiple selection pressures slow evolution, silly amathematician.

Doubling the size of the population doubles the chances of a given mutation in the population.

That could have been better phrased.

If you'd said "doubles the expected number of a given mutation per generation", then the halfwit would be less unlikely to misunderstand you.

So you think that doubling the “expected number of a given mutation per generation” accelerates evolution markedly with huge population, the perhaps you can explain ev does not show this you silly amathematician?

So tell us silly amathmetician, how do multiple selection pressures accelerate evolution?

 

[delphi_ote]

That could have been better phrased.

You're absolutely correct. In my defense, my brain was still reeling from the absolute stupidity of his comments.

If you'd said "doubles the expected number of a given mutation per generation", then the halfwit would be less unlikely to misunderstand you.

Under the assumption that your more precise phrasing would roughly double his chances of understanding, I ran some calculations. Unfortunately, a double isn't precise enough to hold the value.

 

[delphi_ote]

The adaptation of organisms to environmental stresses is primarily done by recombination and natural selection, not mutation and natural selection. Recombination and natural selection can give rapid morphological changes, mutation and selection is a profoundly slow process. Ev shows this and reality shows this.

Viruses and bacteria no longer evolve quickly. We're all saved!

 

[joobz]

It seems that when you studied alchemical engineering they neglected to teach you anything about the mathematics of coordinate transformations. For example, there are problems that are defined by space and time coordinates (real coordinates) that can be solve by mathematically combining space and time into a single mathematical coordinate (abstract coordinate). These are called similarity solutions. Don’t mistake your ignorance as my lying to you. You have plenty of ignorance of mathematics, especially the mathematics of mutation and selection.

And what does this have to do with what I asked you? I'm glad you provided a googled explanation of these concepts, but I was asking you how they relate to the question at hand.

The fitness landscape is an abstract concept. You are free to select any variable space you wish that makes the computions easier. I was highlighting again your inability to distinguish between the relavent and irrelavent. Your arbitrary space selection has no bearing on the reality that is modeled. Life isn't beholdened to your definitions.

 

[Dr Richard]

I’m going to admit you are correct. Malaria affects multiple genetic systems.

I am glad to see you admit your mistake about "malaria" being a single selection pressure.

So what is the selection pressure that sums up to evolving reptiles into birds?

I know what you mean, even if the words are wrong.

I didn’t notice the quote from your HIV database where they recommend returning to monotherapy.

A lovely Kleinmanism, thank you. But irrelevant. As with the malaria link above, this paper provides evidence that there are greater than 500 selection pressures acting on the HIV genome alone. And yet it still evolves incredibly fast.

If your theory is true, how can this be?

If multiple selection pressures slow evolution, how can there be a measurable selection pressure on each aa codon of the HIV genome?

Previously, you claimed that from your ev modeling, 3 selection pressures would render evolution impossible. What do you know believe to be the correct number of selection pressures? 500? 5000?

As you have a detailed mathematical model, answers to within an order of magnitude would be sufficient.

You also still haven’t explained why chimpanzees and humans produce different preproinsulin yet produce identical insulin. This all despite you claim we descended from a common ancestor.

And you never defined macroevolution, so you first. If one base pair mutation is a microevolutionary event, how many base pair mutations to make a macroevolutionary event? 2? 10? 100?

As you have a detailed mathematical model, answers to within an order of magnitude would be sufficient.

 

[Dr Adequate]

Then you should understand that multiple selection pressures slow evolution, silly amathematician.

"Should"? I am under no obligation to believe your halfwitted fantasies.

So you think that doubling the “expected number of a given mutation per generation” accelerates evolution markedly with huge population.

No. That's why I said no such thing, you pathetic little liar.

So tell us silly amathmetician, how do multiple selection pressures accelerate evolution?

An unordered set of chance events will come up quicker than the same set in some specified order.

Hence we should expect simultaneous selection pressures to fix a given set of adaptive mutations faster then the same pressures acting in sequence, so long as the population sizes are the same, i.e. the selective pressures on the individual are not demographic pressures on the population.

---

Is "amathematician" your new magic word?

I predict that it will neither make my mathematical abilities disappear, nor magically summon into existence any justification for your bloated delusions of competence.

None of your magic words change reality, kleinman.

 

[Dr Adequate]

And you never defined macroevolution, so you first. If one base pair mutation is a microevolutionary event, how many base pair mutations to make a macroevolutionary event? 2? 10? 100?

2, apparently.

My view on this issue is that once you get beyond a single point mutation you are already starting to enter the realm of macroevolution.

 

[kjkent1]

I guess you can’t post the quotes. Anybody care for another serving of red herring, string cheese and whine? At least Dr Richard posts the quotes.

I have posted several peer-reviewed articles which demonstrate that HGT has occured between various primate lineages -- among them, homo sapiens.

Your response is a classic and innane appeal to incredulity. Because the evidence suggests something which completely forecloses your hypothesis that evolution is mathematically impossible, you cannot deal with it, except to demand that I "post the quotes." There is no supernatural "event horizon" between species which absolutely prevents HGT from occurring. So either modify your theory to accomodate the evidence, or admit that your theory is falsified.

Or, have another Captain Morgan and continue to delude yourself that you're the only scientist on planet Earth who knows the "truth" about the massive left-wing conspiracy to replace Jesus with Charles Darwin.

 

[Thabiguy]

It is you who is wrong on this point; you can not ignore neutral (or silent) mutation.

It's not a question of ignoring them, it's a question of how they're represented. They can equivalently be considered component parts of the mutation that does affect a trait. However, that is a moot point; I already told you that you may consider as many dimensions as you like, including all mutations that do not affect the fitness, hence including all silent mutations.

What you still don’t understand is that when you expand the number of dimensions (by increasing the genome length) you are expanding the search space which profoundly slows the search.

What you still don't understand is that the "search" space is not the genome space, it is the local neighborhood of the population genomes. Until you realize that, your interpretation of the fitness landscape will remain distorted and your conclusions incorrect.

Selection imposes direction on the search process.

It does not and could not possibly, as selection comes into play only after the "search" is complete (after a mutation occurs that does change the fitness).

 

[Ichneumonwasp]

I know I'm going to hate myself in the morning for stepping back into this steaming pile (and by the way, Thabiguy and Taffer outstanding posts, and Dr. A, Joobz, Dr. Richard, delphi and kjkent, same to you), but here goes.......

You also still haven’t explained why chimpanzees and humans produce different preproinsulin yet produce identical insulin.

Um, because, we have different diets and there is big difference in translation efficiency between the two varieties of preproinsulin. Humans need more of it. Look here

Or is the question why the protein splice site is maintained? By the way, this is not intended as a knock-down argument, only that such alternative mRNA splice sites are important in the variations between species (and even between individuals sometimes).

I was just thinking. This thread would make an excellent text for anyone wanting to know more about evolution. Thanks, Kleinman, for granting us this opportunity to rehash the important aspects of evolution for a wider audience. I, personally, am learning a ton from all the posters, well except for you, Kleinman.