|
Register | Sign In |
|
QuickSearch
Thread ▼ Details |
|
Thread Info
|
|
|
Author | Topic: Rebuttal To Creationists - "Since We Can't Directly Observe Evolution..." | |||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1
|
Kleinman writes: And the selection condition doesn't change the math, it only changes the target gene(s) as demonstrated by the different selection conditions use in the Kishony and Lenski experiments. Sexual selection does change your math. All of your math is based on one beneficial mutation reaching fixation at a time. That's ridiculous. That is not how it works in sexually reproducing populations.
There are not many selection conditions on a savanna, but a few, are starvation, dehydration, thermal stress (both excessively high and excessively low temperatures), disease (bacterial, fungal, parasitic, viral), toxins, and predation, to list a few. It would be sad for some member of the population to get an adaptive mutation that would give a step toward standing upright and end up dying of tetanus or starvation or any of the myriad of other selection conditions that member would face. There wouldn't be much of an improvement in fitness from that first mutation. The earliest populations wouldn't have to be fully adapted to the open savanna. They could live at the edges of their arboreal range and only venture out a bit into the savanna. As they acquired more an more adaptations to the savanna they could move further out into the savanna which would remove them from competition with other arboreal apes. This is nothing like antibiotic selection.
If both parents are homozygous for the resistance allele, then you are doing 2 random trials for the next adaptive mutation in that replication.
Again, there is more than one gene in a genome. Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent.
Drug-resistant bacteria appear as a matter of course simply by neutral evolution. The same for phage resistance, as shown in both the Lederberg plate replica experiment and the Luria and Delbruck fluctuation assay. In those experiments, they start with a single bacterium and grow a population. They then expose that population to antibiotics or phage. They find resistant bacteria that already had the resistance mutation before being exposed to either challenge. However, as discussed in the Lederberg paper, antibiotic resistance appeared 1,000 times less often than phage resistance. How do you explain this difference in the appearance of resistance in both cases?
Doesn't relative fitness differences of different variants determine which variants increase in frequency and which decrease? That is only for alleles of the same gene. This is not true for different genes in a sexually reproducing population. Again, a genome has more than one gene. If there are two beneficial alleles for two different genes then they will both increase in frequency unless there is an interaction between the genes that lowers fitness. For example, the two mutations in different genes that I spoke about earlier, one for lighter skin and one for lactase persistence. They are both beneficial on their own, and they don't compete with one another because the mutations are on different genes. Eukaryotes have more than one chromosome. Genes on separate chromosomes are not linked in any way. They independently disperse through the population.Even then, meiosis will switch alleles across the paired chromosomes so you will have different linkages between alleles of different genes on the same chromosome through time. There is about 1 cross over event per chromosome per offspring, at least in humans. The only difference in the math is that for a clonal replicator, each genome replication is one adaptive allele replication. There is more than one gene in the genome.
This is the paper that explains how to compute the probability of adaptive mutations occurring at two or more genetic loci. Those are asexual populations under extremely stringent selection conditions, both of which are not true for human evolution.
And you still don't get it, competition and fixation slow adaptation. This calculation is all about determining the probability of getting an A1A2 variant as a function of the entire population size. YES, IN AN ASEXUAL POPULATION. What are you not getting here? Just to repeat . . . Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:You are missing the point Taq. I've never claimed that you can't get an increase in frequency of adaptive alleles in a population. My point is the alleles have to exist in the first place. Put this into the context of a real example, Darwin's Finches. You start with a diverse population of Finches with a diverse distribution of alleles that cause all different sizes and shapes of beaks. You put that population into an environment with a food source that requires a particular beak size and shape to feed, for example, insects in crevices. Those variants with wrong-shaped beaks die from starvation leaving a population with a high frequency of beaks that are long and narrow. Those remaining variants breed among themselves and any alleles that improve beak size for that food source have an increased probability of recombining into future offspring. That same process occurs for those variants with short, stout beaks that are suitable for eating nuts in some other environment. Ultimately, this process reduces the diversity of the population. All these alleles must first exist in the population for this process to work which means you need to start with billions of replications of the replicator (Finches in this case) before you even have a chance for this type of selection process to work. Then, when this process occurs, you reduce the diversity of the lineage that remains. If this reduced diversity population is to have a chance to again go through this kind of selection process again, this population must recover size, and do billions of replications to again become a more diverse population. When you try to apply this principle to human evolution, you only have about a billion replications over whatever number of millions of years you want for this to work. Selection like this bottlenecks and reduces the genetic diversity of the population. If you are going to assume that humans and chimps arose from a common ancestor, humans and chimps start with the same alleles. You simply do not have sufficient population size to explain the evolution of humans and chimps from a common ancestor to what we see today using this selection process. You do have sufficient population size to see populations with blond hair and light-colored skin, other populations with curly hair and dark-colored skins, tall populations, and short populations,... Humans have something which enables us to reach a population of over 7 billion while chimps only have achieved 300,000. What genetic differences allow this?
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:You still haven't learned that the competition and fixation process slows the DNA adaptive evolutionary process. And you haven't thought through what is required for recombination to operate. Think about what happens with Darwin's Finches and what would happen to this population if it was put through another selection process like it was with the changing food sources. Kleinman:You are trying to simulate a savanna-like Kishony's experiment but you don't even realize it. Do you think that starvation, dehydration, thermal stress (both excessively high and excessively low temperatures), disease (bacterial, fungal, parasitic, viral), toxins, and predation don't exist on the edge of the savanna? Kleinman:Are you ready to do the probability mathematics of random recombination? Here's another hint for you, it is a trinomial distribution. Kleinman:First, you explain how the phage got the resistance allele. Kleinman:Doesn't that same process occur with chimps as well? Why don't they have a population of over 7 billion today? Kleinman:Would you compute the probability of two beneficial mutations occurring in two different genetic loci in a single replication for us? Kleinman:You still don't get or don't want to get that the only difference in the math between asexual replicators and sexually reproducing replicators is the former you can use genome replications while in the latter, use allele replications. And this math only considers the number of selection conditions, not the intensity of selection. Do you understand why the intensity of selection doesn't make a difference? Kleinman:You are doing a sloppy and superficial analysis. You are not taking into account the population sizes necessary and population recovery necessary for this kind of selection to occur more than once. And you appear to be claiming that this kind of selection only occurred with humans and not chimpanzees. Think about what this kind of selection did to the Finch populations in Darwin's Finch case.
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1 |
Kleinman writes: All these alleles must first exist in the population for this process to work which means you need to start with billions of replications of the replicator (Finches in this case) before you even have a chance for this type of selection process to work. Then, when this process occurs, you reduce the diversity of the lineage that remains. Billions of replications would put beneficial mutations in many different genes. You keep asserting that those mutations on separate genes stay separate from one another. They don't.
You start with a diverse population of Finches with a diverse distribution of alleles that cause all different sizes and shapes of beaks. You put that population into an environment with a food source that requires a particular beak size and shape to feed, for example, insects in crevices. Those variants with wrong-shaped beaks die from starvation leaving a population with a high frequency of beaks that are long and narrow. Those remaining variants breed among themselves and any alleles that improve beak size for that food source have an increased probability of recombining into future offspring. That same process occurs for those variants with short, stout beaks that are suitable for eating nuts in some other environment. Ultimately, this process reduces the diversity of the population. Actually, hybridization was a big part of the process:
quote: This was an increase in genetic variation. It was the mixture of mutations from different lineages that resulted in adaptation, something you can't seem to get your head wrapped around.
You simply do not have sufficient population size to explain the evolution of humans and chimps from a common ancestor to what we see today using this selection process. The only reason you are saying this is because you only allow one beneficial mutation to move towards fixation at a time, which is ridiculous.
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1 |
Kleinman writes: You still haven't learned that the competition and fixation process slows the DNA adaptive evolutionary process. And you haven't thought through what is required for recombination to operate. Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent. Please show me how this is wrong.
You are trying to simulate a savanna-like Kishony's experiment but you don't even realize it. Do you think that starvation, dehydration, thermal stress (both excessively high and excessively low temperatures), disease (bacterial, fungal, parasitic, viral), toxins, and predation don't exist on the edge of the savanna? Human ancestors weren't plucked from their arboreal environment and placed smack dab in the middle of the savanna. That's not what happened. That is what happens in the analogous Kishony experiment. The Kishony experiment requires adaptation in almost a single generation, something our human ancestors would not have had to do.
Are you ready to do the probability mathematics of random recombination? Here's another hint for you, it is a trinomial distribution. Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent. Please show me how this is wrong.
First, you explain how the phage got the resistance allele. The same way streptomycin gets antibiotic resistance. Do cancer drugs get drug resistance? It is the bacteria that are phage resistant. The phage aren't phage resistant. That makes no sense. The bacteria had already evolved phage resistance before they came into contact with phage, so it wasn't the phage that gave the bacteria a resistance allele. So I will ask again. As discussed in the Lederberg paper, antibiotic resistance appeared 1,000 times less often than phage resistance. How do you explain this difference in the appearance of resistance in both cases? Doesn't that same process occur with chimps as well? Why don't they have a population of over 7 billion today? Please address what I wrote: you: Doesn't relative fitness differences of different variants determine which variants increase in frequency and which decrease? me: That is only for alleles of the same gene. This is not true for different genes in a sexually reproducing population. Again, a genome has more than one gene. If there are two beneficial alleles for two different genes then they will both increase in frequency unless there is an interaction between the genes that lowers fitness. For example, the two mutations in different genes that I spoke about earlier, one for lighter skin and one for lactase persistence. They are both beneficial on their own, and they don't compete with one another because the mutations are on different genes. Eukaryotes have more than one chromosome. Genes on separate chromosomes are not linked in any way. They independently disperse through the population. Even then, meiosis will switch alleles across the paired chromosomes so you will have different linkages between alleles of different genes on the same chromosome through time. There is about 1 cross over event per chromosome per offspring, at least in humans. You still don't get or don't want to get that the only difference in the math between asexual replicators and sexually reproducing replicators is the former you can use genome replications while in the latter, use allele replications. There is more than one gene in the genome.
You are not taking into account the population sizes necessary and population recovery necessary for this kind of selection to occur more than once. You don't need population bottlenecks in order for beneficial mutations in different genes to all reach fixation in parallel.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:Taq, you have mastered Mendelian Genetics. Now you need to master the mathematics of selection. What must happen to a population in the wild in order for your recombination example to have a reasonable probability of occurring? Kleinman:I'm not aware of any real environments devoid of selection pressures. And the Kishony experiment requires about 30 generations of doubling for each adaptive mutational step. So which selection condition did humans adapt to in order to achieve a population of greater than 7 billion today? Kleinman:You just don't want to accept the fact that this kind of selection bottlenecks a population. Kleinman:Are you saying that phages are what give humans the reproductive advantage over chimps? I thought phages only infect bacteria. Kleinman:It appears you need some help with the meaning of reproductive fitness. Fitness - Wikipedia(biology) Wikipedia:And for the meaning of reproductive success: Reproductive success - Wikipedia Wikipedia: Kleinman:Sure, there are thousands of coding genes and most of the genome controls the expression of the genes. Every gene and the rest of the genome are potential targets for mutations. What's the probability of more than one adaptive mutation occurring in a genome in a single replication Kleinman:That's an interesting claim. Since you are in speculation mode, give us an example of this in human evolution and explain why it didn't happen with chimps.
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1
|
Kleinman writes: Taq, you have mastered Mendelian Genetics. Now you need to master the mathematics of selection. What must happen to a population in the wild in order for your recombination example to have a reasonable probability of occurring? 1. Each of the two beneficial mutations must be under positive selection.2. Free interbreeding within the population. 3. Positive selection lasts long enough so that both beneficial mutations reach a small percentage of the population. If 5% of the population is heterozygous for one of the two beneficial mutations then 2.5% of births will have parents where each one has one of the two beneficial mutations. If they are heterozygous, 12.5% of their children will have both mutations. Dual carriers will only increase as each mutation increases in frequency until 100% of the population has both mutations if the benefice of each mutation is strong enough.
I'm not aware of any real environments devoid of selection pressures. I'm not aware of very many environments where only a few individuals survive out of billions in a single generation.
And the Kishony experiment requires about 30 generations of doubling for each adaptive mutational step. Again, they are asexual organisms. That's why there are steps. If they are reproducing sexually then multiple steps could occur in one set of 30 generations and be combined into a single genome.
You just don't want to accept the fact that this kind of selection bottlenecks a population. Bottlenecks do occur, they just aren't necessary.
Are you saying that phages are what give humans the reproductive advantage over chimps? I thought phages only infect bacteria. So I will ask again. As discussed in the Lederberg paper, antibiotic resistance appeared 1,000 times less often than phage resistance. How do you explain this difference in the appearance of resistance in both cases? Sure, there are thousands of coding genes and most of the genome controls the expression of the genes. Every gene and the rest of the genome are potential targets for mutations. What's the probability of more than one adaptive mutation occurring in a genome in a single replication In a single replication the odds are low. Across hundreds of millions of replications in humans, the odds are extremely high that there will be different beneficial mutations in a whole host of different genes. You want to make the claim that only the fittest mutation out of all them will move towards fixation at the demise of all those other mutations in other genes. THIS ISN'T TRUE. All of those mutations will have their frequencies changed by comparison to the other alleles for that gene, not the frequency of alleles in other genes.
That's an interesting claim. Since you are in speculation mode, give us an example of this in human evolution and explain why it didn't happen with chimps.
Speculation mode?????? Are you once again denying this scenario? Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent. Please show me how this is wrong. This would have happened in the chimp lineage as well. The difference is that the chimp lineage and the human lineage adapted to different environments, so they had different mutations reach fixation. It's one of the basic concepts in evolution.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:Do you have any empirical examples that correlate with your calculation? Do you think your math correlates with the Darwin's Finch example that I describe in Message 302. You are doing what people that don't have any experience with mathematical and computer models do. You plug numbers into models and obtain a prediction and think that automatically correlates with reality. You need to present experimental and empirical data that correlates with your math. That's what I've done with my papers and that's why you agree that they are correct for asexually replicating populations. HIV doesn't do what you claim. HIV is diploid and does recombination. Why doesn't it evolve resistance to 3 drug therapy by the mechanism that you are describing? Aren't any adaptive alleles to one drug or another moving toward fixation? Kleinman:That happens all the time with antimicrobial therapy. All the drug-sensitive variants are killed and a few drug-resistant variants survive. The same kind of thing happens with pesticides and herbicides. And biologists talk about their 5 mass extinction events. Something must have survived, we are here. Kleinman:The same kind of DNA adaptive evolutionary steps happens with sexually replicating organisms. The resistance alleles must form by a sequence of adaptive mutations accumulating on an allele. The mathematical behavior of the formation of those alleles isn't computed by addition, you must compute this joint probability using the multiplication rule. How many adaptive alleles have been recombined in the human lineage in order to give the reproductive fitness advantage humans have over chimps? And why couldn't chimps do this as well since you believe that humans and chimps started from the same progenitors?
Kleinman:Are you claiming that you can increase the frequency of two alleles to fixation, yet other variants still exist in the population? I'm still trying to figure out how you can get two variants fixed in a population simultaneously. Could you give us empirical examples of your claim? Kleinman:I've never studied the experiment, but if you think it explains how humans and chimpanzees evolved from a common ancestor, please explain. Kleinman:Low is quite an understatement. For a mutation rate of 1e-8, that probability is on the order of 1e-16. I don't make any claims about the fixation of any beneficial alleles in humans. It is you that make all the claims about alleles moving toward fixation simultaneously. So, you have mutations across hundreds of millions of replications in humans. How does a beneficial mutation occurring in an Alaskan Eskimo get recombined with a beneficial mutation from an Australian Aborigine? Kleinman:Speculation mode!!!!! Yes With regards to your scenario, where's the beef? You pass your Mendelian Genetics exam. So humans and chimps that originally lived in the same environment because they came from the same progenitor and had the same alleles, had different alleles go to fixation allowing the human lineage to live on the savanna but the chimp alleles didn't go to fixation so they were stuck living in the forest. And hundreds of thousands of generations of chimps didn't give any members of the chimp population those mutations that would allow them to live in the savanna. Do I have your scenario correct?
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1
|
Kleinman writes: Do you have any empirical examples that correlate with your calculation? It's simply math. If you agree that selection increases the frequency of beneficial alleles then the outcome I described is inevitable. And I already showed you the real world example:
quote: HIV doesn't do what you claim. HIV is diploid and does recombination. If you put human ancestors in an environment where only a few out of billions survived in each generation then they would go extinct. However, that is not what happened in the evolution of humans. It is the extremely harsh environment that makes your HIV model irrelevant to human evolution. I fully agree that if the founding population of the human lineage was put in a similar extreme environment that we would not be here today.
That happens all the time with antimicrobial therapy. But it didn't happen in human evolution.
The same kind of DNA adaptive evolutionary steps happens with sexually replicating organisms. The resistance alleles must form by a sequence of adaptive mutations accumulating on an allele. Parent One is homozygous for a beneficial mutation in gene A. Parent Two is homozygous for a beneficial mutation in gene B. All offspring will be heterozygous for both beneficial mutations. There's no need for another random trial to get the 2nd mutation in each parent. Please show me how this is wrong. I'm still trying to figure out how you can get two variants fixed in a population simultaneously. Are you remembering that they are on different genes? I'm guessing you forgot that part. Are you also forgetting about separate chromosomes and meiosis?
I've never studied the experiment, but if you think it explains how humans and chimpanzees evolved from a common ancestor, please explain. I am asking you to explain. So I will ask again. As discussed in the Lederberg paper, antibiotic resistance appeared 1,000 times less often than phage resistance. How do you explain this difference in the appearance of resistance in both cases? For a mutation rate of 1e-8, that probability is on the order of 1e-16. Where are the units on these rates?
I don't make any claims about the fixation of any beneficial alleles in humans. Bullshit. From message 246: "The math is way beyond you. 20 million beneficial mutations * 300 generations/fixation = 6 billion generations" You have one beneficial moving to fixation at a time, and no other beneficial mutations are even allowed to occur or move towards fixation while one beneficial mutation is moving towards fixation. That's ridiculous. This is ALL about how different mutations in different genes move towards fixation. All you can seem to do is see one gene and one beneficial mutation at a time. You simply can't wrap your head around the idea that there can be more than one possible beneficial mutation in more than one gene. You also can't seem to understand that these mutations are spread over many chromosomes.
How does a beneficial mutation occurring in an Alaskan Eskimo get recombined with a beneficial mutation from an Australian Aborigine? It may take a long time. The most recent history of humans has seen isolated subpopulations. This doesn't seem to be the case for most of our history where we were found in Africa.
With regards to your scenario, where's the beef? You pass your Mendelian Genetics exam. It demonstrates the very thing you claim can't happen, the combination of different beneficial mutations into a single genome.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:That's a computer simulation. How about an experimental example, one with real living things such as the Kishony or Lenski experiment but with sexual replicators? Do you know that there are similar experiments to Lenski's experiment but performed with yeast? Kleinman:HIV hasn't gone extinct and it doesn't go extinct in someone treated with 3 drug therapy. And your effective population size is only 20,000. I'm surprised you think that the real world isn't an extreme environment since 99% of all species have gone extinct. That doesn't sound like an evolution-friendly environment. The reason why single drug-resistant alleles in HIV treatment don't increase in frequency is that these variants don't have reproductive fitness any better than the drug-sensitive alleles. The key portion of your statement above is "selection increases the frequency of beneficial alleles then the outcome I described is inevitable". In an environment with multiple low-intensity selection pressures, any mutation for one selection pressure or another will not give a significant improvement in reproductive fitness for that member.
Kleinman:Really? There haven't been famines, pandemics, droughts, wars,...? How about this: Population Bottlenecks and Pleistocene Human Evolution quote: Kleinman:So 100% of the population all have the adaptive allele at one genetic locus and 100% of the population have the other adaptive allele at the second genetic locus. Why do they need recombination for a descendant to get both adaptive alleles. Everyone in the population already has both adaptive alleles. Kleinman:mutations per base per replication Kleinman:Those aren't my claims, that's Haldane's math for selective substitutions. The values for neutral evolution are based on Kimura's and your simple neutral evolution equation. All my papers are on DNA adaptive evolution, a subject that biologists have bungled. I've never written a paper on biological evolutionary competition other than to use Haldane's model along with my DNA adaptive evolutionary model to simulate and predict the Lenski experiment which it does very nicely. The claim that you are making is that multiple potentially adaptive mutations will all give some improvement in reproductive fitness to those members with those mutations. Why wouldn't Haldane and Kimura say something like this in their papers if this really happens? The reason is the concept only exists in your imagination. If multiple alleles are going to increase in frequency simultaneously, it will happen as it does with Darwin's Finches example. You are blowing smoke. Here's a question for you about your cockamamy idea about fixation. You are talking about fixation of adaptive alleles. This would be similar to what Haldane's Cost of Selection model addresses. His model gives a fixation rate for a single adaptive allele of 300 generations. Do 2 adaptive alleles at two different genetic loci fix faster than a single adaptive allele at one locus? Do 3 fix faster than 2, 4 fix faster than 3,..., 19,999,999 fix faster than 19,999,998, 20,000,000 fix faster than 19,999,999? I look forward to hearing your response to that question.
Kleinman:How about a beneficial mutation occurring on someone from North Africa and a different beneficial mutation on someone from South Africa? That's only 4600 miles. Are there any long-distance relationships in your population of 20,000? Kleinman:Mendelian Genetics, the mechanism that explains the evolution of humans. You should publish it, in MAD magazine.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Taq:Taq needs a little help with his math. A probability of 0.025 does not equal 25%, it equals 2.5%. And according to his calculation, that's when 5% of his population has the A allele and another 5% have the B allele. That is 1000 members each in his 20,000-member population. How many generations does it take for these beneficial alleles to "fix" and this recombination process to occur? Start by recognizing the correct probability distribution. A good mathematical description of the trinomial distribution can be found here:
The Trinomial Distribution quote: So, let's take Taq's example of a population of 20,000. Let there be 3 variants in the population. The A variants have a beneficial allele at one genetic locus, a second B variant has a different beneficial allele at a different genetic locus, and a third variant (call them C) has neither allele A nor allele B at either genetic locus. For simplicity, assume the A and B variants are homozygous at their respective loci. Each mating and replication is a random trial. And assume the random trial occurs with replacement. If the total population size is large, (for example, 20,000), the probabilities computed will be accurate. Define the following variables:n – is the total population size. nA – is the number of members in the population with beneficial allele A. nB – is the number of members in the population with beneficial allele B. nC – is the number of members in the population that have neither beneficial allele A nor beneficial allele B. In addition, we have the following condition: nA + nB + nC = n. And the frequency of each of the variants are:f_A = nA/n f_B = nB/n f_C = nC/n The trinomial distribution for this example is: f (x, y) = 2!/(x!y!(2 − x − y)!) *(nA/n)^x *(nB/n)^y *(nC/n)^(2−x−y)where (x + y =⩽ 2) Note that the 2!/(x!y!(2 − x − y)!) takes into account the permutations of drawing A's, B's, and C's from the population. If you set x=y=1, you can obtain the probability equation for selecting an A and B variant from the population. f (x = 1, y = 1) = P(A, B) = 2*(nA/n)*(nB/n) One can draw several conclusions from this math. Neither variant A nor variant B can go to fixation because if A is fixed, nA=n and nB and nC must be zero. And likewise for variant B, it cannot go to fixation. The number of members with A and B alleles that give the maximum probability of an AB recombination event is nA=nB=n/2. Consider Taq's population of 20,000. Initially, it has no A or B variants. If we assume a mutation rate of 1e-8, it will take about 1/(1e-8) replications or 1e8 = 100,000,000 replications. Cut that number in half for diploid gives 50,000,000 births. It will take about 2500 generations for those A and B variants to initially appear. So, in generation 2501, nA=1, nB=1, and nC=19998. Then assume that the A and B variants have equal reproductive fitness, according to Haldane's model, it will take about 300 generations for those variants to go to "fixation" where nA=nB=n/2 and the probability of an AB recombination event occurring is: P(A, B) = 2*(10000/20000)*(10000/20000) = 1/2 Each one of these AB recombination events will take about 2800 generations according to Taq's model. An evolutionary process over 5 million years with 20 years per generation allows for 250,000 generations. At 2800 generation for each adaptation step (2 adaptive mutations/step) gives about 180 possible adaptive mutations in some lineage. That is a much, much smaller number than Taq's claim of 20,000,000 over the evolutionary interval possible. Perhaps Taq wants to try the example of more than 2 adaptive alleles? That would require using a multinomial distribution. You can read and study about that math here:Multinomial distribution - Wikipedia quote: In the case of three adaptive alleles, we have 4 different variants in the population, one variant has beneficial allele A, another variant has beneficial allele B, a third variant has beneficial allele C, and the fourth variant has none of the beneficial alleles, call that variant D. The random mutation process will still take about 2500 generations to evolve the A, B, and C variants. And if we assume that each of these variants gives identical improvement in fitness such that the sum of the frequencies at fixation equals 1 at a form of fixation then the frequency of each of the adaptive alleles is 1/3. But it is not as simple as that. In a single recombination event under this circumstance, you can't get an ABC variant, you can only get an AB, AC, or BC variant. Those variants must go to "fixation" to improve the probability of getting an ABC variant on recombination. That adds another 300 generations to the adaptive evolutionary process according to Haldane's math. Do you see a pattern forming here? And do you see the problem with your mathematically deficient argument? At the end of just 50,000 generations and a total of a billion replications, how do your 20,000,000 adaptive mutations accumulate in every member of the population at the end of such a recombination process?
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1
|
Kleinman writes: Taq needs a little help with his math. A probability of 0.025 does not equal 25%, it equals 2.5%. And according to his calculation, that's when 5% of his population has the A allele and another 5% have the B allele. That is 1000 members each in his 20,000-member population. How many generations does it take for these beneficial alleles to "fix" and this recombination process to occur? I don't need any help. We have two genes with two alleles, Aa and Bb. Parent A is Aa bb, Parent B is aa Bb. Half of the children will be Aa and half will be Bb. The odds of being Aa and Bb are the product of those probabilities, 0.25 or 25%.
Define the following variables: n – is the total population size. nA – is the number of members in the population with beneficial allele A. nB – is the number of members in the population with beneficial allele B. nC – is the number of members in the population that have neither beneficial allele A nor beneficial allele B. In addition, we have the following condition: nA + nB + nC = n. And the frequency of each of the variants are: f_A = nA/n f_B = nB/n f_C = nC/n You are calculating 3 alleles at the same gene, not 3 mutations in 3 different genes. The frequency of each mutation can be 90%, and that obviously does not add up to the total population size. In other words, you are modeling AA, Aa, and aa. You are not modeling the combination of Aa and Bb.
Perhaps Taq wants to try the example of more than 2 adaptive alleles? Perhaps you should try modeling mutations in different genes instead of alleles of a single gene. You are also stopping your modelling at one generation and not factoring in all of the incoming mutations that occur while those mutations are moving towards fixation.
|
|||||||||||||||||||||||||||||||||||||||
Taq Member Posts: 10084 Joined: Member Rating: 5.1
|
Kleinman writes: That's a computer simulation. How about an experimental example, one with real living things such as the Kishony or Lenski experiment but with sexual replicators? It's a computer simulation of a real population of fish, and the simulation matched the distribution of 3 alleles in 3 different genes associated with high salinity adaptations. Read the paper:
quote: HIV hasn't gone extinct and it doesn't go extinct in someone treated with 3 drug therapy. And your effective population size is only 20,000. I'm surprised you think that the real world isn't an extreme environment since 99% of all species have gone extinct. That doesn't sound like an evolution-friendly environment. More than 99% of bacteria die or fail to reproduce in concentrations of antibiotics above the minimal inhibitory concentration.
The reason why single drug-resistant alleles in HIV treatment don't increase in frequency is that these variants don't have reproductive fitness any better than the drug-sensitive alleles. If the drugs don't reduce the replication rate of HIV without the resistance mutation then what does the drug do and what is resistance?
Really? There haven't been famines, pandemics, droughts, wars,...? Which one of those wiped out more than 99% of humans? In what way were the post bottleneck environments stable enough to induce adaptation to them?
So 100% of the population all have the adaptive allele at one genetic locus and 100% of the population have the other adaptive allele at the second genetic locus. Why do they need recombination for a descendant to get both adaptive alleles. Everyone in the population already has both adaptive alleles. You tell me, you are the one who keeps talking about recombination events.
I've never written a paper on biological evolutionary competition other than to use Haldane's model along with my DNA adaptive evolutionary model to simulate and predict the Lenski experiment which it does very nicely. How does Haldane's model apply to different mutations on different genes in a diploid sexually replicating species?
This would be similar to what Haldane's Cost of Selection model addresses. His model gives a fixation rate for a single adaptive allele of 300 generations. Do 2 adaptive alleles at two different genetic loci fix faster than a single adaptive allele at one locus? Do 3 fix faster than 2, 4 fix faster than 3,..., 19,999,999 fix faster than 19,999,998, 20,000,000 fix faster than 19,999,999? I look forward to hearing your response to that question. They reach fixation independently of each other if they are not close to each other on the same chromosome, which is called linkage disequilibrium. If we take your 300 generation number, then in generation 1 there are 5 adaptation mutations and they reach fixation at generation 301. In generation 2 there are 5 beneficial mutations, and they reach fixation at generation 302. In generation 3 there are 5 beneficial mutations, and they reach fixation in generation 303. At no time does evolution just stop and wait for the previous beneficial mutations to reach fixation while ignoring all of the beneficial mutations that happen in the mean time.
How about a beneficial mutation occurring on someone from North Africa and a different beneficial mutation on someone from South Africa? That's only 4600 miles. Are there any long-distance relationships in your population of 20,000? It would depend on gene flow.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
The first thing I want to say is thank you to Taq. His admission that my DNA adaptive evolutionary mathematical model is correct for asexual replicators can't endear him to most of the posters to this forum. Now, the question becomes does sexual replication and recombination significantly alter this DNA adaptive evolutionary model.
Kleinman:Oh? I thought that we were assuming that each member of the population were homogeneous at these two loci. That would give an upper limit of the probability of the two adaptive alleles occurring in the offspring. Do you want to show us how to compute the joint probability of one Aa parent mating with a second Bb parent to give an AB offspring from your population of 20,000 as a function of the distribution of different frequencies of variants? That is a multinomial distribution calculation with lots of different possible outcomes. Start with AA, BB, CC, Ab, AC,... You have a few permutations to compute. Kleinman:Apparently, I've confused you. Adaptive allele A occurs at one genetic locus, adaptive allele B occurs at a different genetic locus, and the C alleles are the subset of all alleles that don't give an improvement in fitness at either genetic loci. If you want to distinguish the C alleles at each locus such as "a" and "b" alleles and include a Mendelian computation but that just lowers the probabilities. Try doing the math for the population and assume the adaptive alleles are homogeneous so that the mating of an A and B parents gives an AB offspring. Kleinman:I'm modeling two genetic loci in this case. You are having some difficulty correlating the variables with the physical problem. The "A" adaptive allele happens at one genetic locus and the "B" adaptive allele happens at a different genetic locus. Taq, have you thought about what would happen in the fixation process if either the A adaptive allele or B adaptive allele gave greater reproductive fitness than the other? In the biological evolutionary competition, wouldn't the variant with the adaptive allele that gives greater reproductive fitness go to a frequency of 1 and the variant that gives a lower degree of improvement in reproductive fitness go to a frequency of 0? Then your population of 20,000 would have to replicate for 2500 generations to get an AB variant by DNA evolution.
|
|||||||||||||||||||||||||||||||||||||||
Kleinman Member (Idle past 364 days) Posts: 2142 From: United States Joined: |
Kleinman:Let's say that every time you flap your arms, you generate a small amount of lift. I then take the laws of physics and formulate a computer simulation that determines the amount of lift as a function of the number of times you flap your arms. I plug in a number of times you flap your arms that is sufficient to generate a lift to get you off the ground. Does that mean you can actually fly? Kleinman:But that leaves 1% still able to reproduce because they have the alleles that enable them to do so. And in the case of HIV, this virus can do recombination. The problem for the HIV virus is that it cannot increase the frequency of any of the single drug-resistant variants to give a reasonable probability of an advantageous recombination event occurring. Darwin's Finches starvation bottleneck does increase the frequency of resistance alleles. Kleinman:All these drugs inhibit the reproduction of the virus, they don't kill the virus. Recombination in the 2 drug environment might well get a two-drug resistant variant but the addition of a third drug (third selection condition) affects the mathematics of DNA (RNA in this case) evolution and random recombination for the virus to adapt. Kleinman:The point you are missing is that these bottlenecks select for variants with some degree of resistance to these selection conditions. For example, the black plague killed between 75-200 million people when the world population was about 450 million at that time. Black Death - Wikipedia and How Many People Have Ever Lived on Earth? And if all humans lived in Africa at some time in the past, you would only need a continental drought and famine to cause mass death in the human population. Even without these considerations, you only have about a billion replications to do your accounting of common ancestor to the human lineage. That does not allow for the accumulation of many adaptive mutations in a single lineage (except in your imagination). Kleinman:You can't have different mutations fixing simultaneously unless you have hitchhikers. Haldane's math applies to the fixation of a single allele. You want to consider two different adaptive alleles at two different genetic loci fixing simultaneously. But that only allows for a frequency of 0.5 for each variant. That can only happen if both alleles give the same improvement in reproductive fitness to both variants. If one allele gives greater reproductive fitness than the other, that allele will fix and the less fit variant will add to the cost of natural selection. That's how it works in the Lenski experiment. Perhaps you think it works differently with sexual reproducers. How does that biological evolutionary competition work? Kleinman:So the frequency of the A variant is 1 in the population and the frequency of the B variant is 1? You need to check your math. Kleinman:I don't think they opened that pipeline yet.
|
|
|
Do Nothing Button
Copyright 2001-2023 by EvC Forum, All Rights Reserved
Version 4.2
Innovative software from Qwixotic © 2024