A synopsis of natural selection for Peter Nyikos

[Alan Kleinman MD PhD]

Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
.
“By the way, Alan, Harshman has been needling me to comment on 
how natural selection enters into those peer-reviewed articles of 
yours. Would you like to give me a synopsis so I can accurately 
judge it for myself?  ”
.
Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
.
However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
.
Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
.
This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur. Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region. This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
.
The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step. Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.

[John Harshman]

Note that the only actual mathemetics he mentions here is about the
probability of mutation and has nothing to do with selection.

[Alan Kleinman MD PhD]

John, you still don't get it. The last two sentences of the post address how natural selection enters into my model. I think you are stuck on the notion of differential survival as the key to natural selection where for rmns, reproductive success is actually the key. Perhaps more detail will help you understand.
.
The Kishony experiment has several different colonies of bacteria growing on his plate. Each of these colonies is creating different lineages by various mutations where each of these lineages is on their own particular evolutionary trajectory. Each of these trajectories is made up of sets of binomial probability problems linked by the multiplication rule. If on any step of a particular evolutionary trajectory, there is not reproductive success, that variant will have a low probability of improving fitness with another (beneficial) mutation and will not be able to grow in a higher concentration antibiotic region. Those variants that do have reproductive success at every step of their particular evolutionary trajectory will be able to grow in higher concentration antibiotic regions because they have enough members (due to reproductive success) at each evolutionary step to have a reasonable probability of the next beneficial mutation. This is the cycle of beneficial mutation/amplification of the beneficial mutation. The work of Haldane and Kimura is an attempt to do the mathematics of the amplification phase of this cycle.

[John Harshman]

If natural selection enters into your model, where does it? You have a
variable for population size, and you have a variable for number of
generations. But where's your variable representing change in population
size over time (what you call "amplification")?

> The Kishony experiment has several different colonies of bacteria
> growing on his plate. Each of these colonies is creating different
> lineages by various mutations where each of these lineages is on
> their own particular evolutionary trajectory. Each of these
> trajectories is made up of sets of binomial probability problems
> linked by the multiplication rule. If on any step of a particular
> evolutionary trajectory, there is not reproductive success, that
> variant will have a low probability of improving fitness with another
> (beneficial) mutation and will not be able to grow in a higher
> concentration antibiotic region. Those variants that do have
> reproductive success at every step of their particular evolutionary
> trajectory will be able to grow in higher concentration antibiotic
> regions because they have enough members (due to reproductive
> success) at each evolutionary step to have a reasonable probability
> of the next beneficial mutation. This is the cycle of beneficial
> mutation/amplification of the beneficial mutation. The work of
> Haldane and Kimura is an attempt to do the mathematics of the
> amplification phase of this cycle.

How do you treat "success" and "amplification" in your math?

[Alan Kleinman MD PhD]

John, debating with you is like debating with Ray. You don't understand the basics of probability theory and you are unwilling to learn these basic principles. For those who do understand the basic principles of probability theory, the term n*nG is simply the number of replications and the number of replications is the measure of reproductive fitness. It is the total number of replications which determines the probability of the next beneficial mutation occurring. My model does not address the rate at which these mutations occur, only the number of replications necessary for there to be a reasonable probability of a beneficial mutation occurring. For any readers of this thread, perhaps you can find a way to explain this to John but I doubt he will understand any explanation until he takes the time to learn introductory probability theory.

>
> > The Kishony experiment has several different colonies of bacteria
> > growing on his plate. Each of these colonies is creating different
> > lineages by various mutations where each of these lineages is on
> > their own particular evolutionary trajectory. Each of these
> > trajectories is made up of sets of binomial probability problems
> > linked by the multiplication rule. If on any step of a particular
> > evolutionary trajectory, there is not reproductive success, that
> > variant will have a low probability of improving fitness with another
> > (beneficial) mutation and will not be able to grow in a higher
> > concentration antibiotic region. Those variants that do have
> > reproductive success at every step of their particular evolutionary
> > trajectory will be able to grow in higher concentration antibiotic
> > regions because they have enough members (due to reproductive
> > success) at each evolutionary step to have a reasonable probability
> > of the next beneficial mutation. This is the cycle of beneficial
> > mutation/amplification of the beneficial mutation. The work of
> > Haldane and Kimura is an attempt to do the mathematics of the
> > amplification phase of this cycle.
>
> How do you treat "success" and "amplification" in your math?

By the increase in the number of replications. If a picture would help you, look at the figures in my publications.

[John Harshman]

No, n*nG is the number of replications given a constant number of
replications per generation, and it's the change in number of
replications per generation that's a measure of fitness. You don't model
that change at all.

> It is the total number of replications which
> determines the probability of the next beneficial mutation occurring.
> My model does not address the rate at which these mutations occur,
> only the number of replications necessary for there to be a
> reasonable probability of a beneficial mutation occurring.

How can you consider that probability without knowing the rate? Or
perhaps you are using your personal definition of "rate" here?

> For any
> readers of this thread, perhaps you can find a way to explain this to
> John but I doubt he will understand any explanation until he takes
> the time to learn introductory probability theory.

I'd be interesting in anyone explaining that to me.

>>> The Kishony experiment has several different colonies of bacteria
>>> growing on his plate. Each of these colonies is creating different
>>> lineages by various mutations where each of these lineages is on
>>> their own particular evolutionary trajectory. Each of these
>>> trajectories is made up of sets of binomial probability problems
>>> linked by the multiplication rule. If on any step of a particular
>>> evolutionary trajectory, there is not reproductive success, that
>>> variant will have a low probability of improving fitness with another
>>> (beneficial) mutation and will not be able to grow in a higher
>>> concentration antibiotic region. Those variants that do have
>>> reproductive success at every step of their particular evolutionary
>>> trajectory will be able to grow in higher concentration antibiotic
>>> regions because they have enough members (due to reproductive
>>> success) at each evolutionary step to have a reasonable probability
>>> of the next beneficial mutation. This is the cycle of beneficial
>>> mutation/amplification of the beneficial mutation. The work of
>>> Haldane and Kimura is an attempt to do the mathematics of the
>>> amplification phase of this cycle.
>>
>> How do you treat "success" and "amplification" in your math?

> By the increase in the number of replications. If a picture would
> help you, look at the figures in my publications.

But you don't deal with increase in replications in your math.

[Alan Kleinman MD PhD]

So a population of e6 can only replicate for 100 generations, and not 1000 generations?

>
> > It is the total number of replications which
> > determines the probability of the next beneficial mutation occurring.
> > My model does not address the rate at which these mutations occur,
> > only the number of replications necessary for there to be a
> > reasonable probability of a beneficial mutation occurring.
>
> How can you consider that probability without knowing the rate? Or
> perhaps you are using your personal definition of "rate" here?

John, what is the difference in the sample space if you roll 5 dice 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?

>
> > For any
> > readers of this thread, perhaps you can find a way to explain this to
> > John but I doubt he will understand any explanation until he takes
> > the time to learn introductory probability theory.
> I'd be interesting in anyone explaining that to me.

I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.

>
> >>> The Kishony experiment has several different colonies of bacteria
> >>> growing on his plate. Each of these colonies is creating different
> >>> lineages by various mutations where each of these lineages is on
> >>> their own particular evolutionary trajectory. Each of these
> >>> trajectories is made up of sets of binomial probability problems
> >>> linked by the multiplication rule. If on any step of a particular
> >>> evolutionary trajectory, there is not reproductive success, that
> >>> variant will have a low probability of improving fitness with another
> >>> (beneficial) mutation and will not be able to grow in a higher
> >>> concentration antibiotic region. Those variants that do have
> >>> reproductive success at every step of their particular evolutionary
> >>> trajectory will be able to grow in higher concentration antibiotic
> >>> regions because they have enough members (due to reproductive
> >>> success) at each evolutionary step to have a reasonable probability
> >>> of the next beneficial mutation. This is the cycle of beneficial
> >>> mutation/amplification of the beneficial mutation. The work of
> >>> Haldane and Kimura is an attempt to do the mathematics of the
> >>> amplification phase of this cycle.
> >>
> >> How do you treat "success" and "amplification" in your math?
>
> > By the increase in the number of replications. If a picture would
> > help you, look at the figures in my publications.
> But you don't deal with increase in replications in your math.

So what value of n and nG am I using throughout my publications?

[John Harshman]

Of course it could. But that doesn't model fitness.

>>> It is the total number of replications which
>>> determines the probability of the next beneficial mutation occurring.
>>> My model does not address the rate at which these mutations occur,
>>> only the number of replications necessary for there to be a
>>> reasonable probability of a beneficial mutation occurring.
>>
>> How can you consider that probability without knowing the rate? Or
>> perhaps you are using your personal definition of "rate" here?

> John, what is the difference in the sample space if you roll 5 dice
> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?

Ah, I see that you are using your personal definition of "rate". A
mutation rate is generally considered to be something like "number of
expected mutations per site per generation".

>>> For any
>>> readers of this thread, perhaps you can find a way to explain this to
>>> John but I doubt he will understand any explanation until he takes
>>> the time to learn introductory probability theory.
>> I'd be interesting in anyone explaining that to me.

> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.

Typical response: having nothing you say, you attempt to denigrate the
person you're talking to.

>>>>> The Kishony experiment has several different colonies of bacteria
>>>>> growing on his plate. Each of these colonies is creating different
>>>>> lineages by various mutations where each of these lineages is on
>>>>> their own particular evolutionary trajectory. Each of these
>>>>> trajectories is made up of sets of binomial probability problems
>>>>> linked by the multiplication rule. If on any step of a particular
>>>>> evolutionary trajectory, there is not reproductive success, that
>>>>> variant will have a low probability of improving fitness with another
>>>>> (beneficial) mutation and will not be able to grow in a higher
>>>>> concentration antibiotic region. Those variants that do have
>>>>> reproductive success at every step of their particular evolutionary
>>>>> trajectory will be able to grow in higher concentration antibiotic
>>>>> regions because they have enough members (due to reproductive
>>>>> success) at each evolutionary step to have a reasonable probability
>>>>> of the next beneficial mutation. This is the cycle of beneficial
>>>>> mutation/amplification of the beneficial mutation. The work of
>>>>> Haldane and Kimura is an attempt to do the mathematics of the
>>>>> amplification phase of this cycle.
>>>>
>>>> How do you treat "success" and "amplification" in your math?
>>
>>> By the increase in the number of replications. If a picture would
>>> help you, look at the figures in my publications.
>> But you don't deal with increase in replications in your math.

> So what value of n and nG am I using throughout my publications?

Why, you use various different values. But you don't deal with the way
in which n changes, and that would be the "ns" bit of "rmns".

[Oxyaena]

On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> .
> “By the way, Alan, Harshman has been needling me to comment on 
> how natural selection enters into those peer-reviewed articles of 
> yours. Would you like to give me a synopsis so I can accurately 
> judge it for myself?  ”
> .
> Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> .

And?

> However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)

CITE? This in no way conflicts with biological fitness, you're confusing sexual selection with natural selection. Sexual selection refers to the competition between potential mates as to who gets to spread their seed, in other words, a competition for the right to fuck with the opposite gender, or, in more technical terms, a competition to decide which potential mate gets to spread its genes unto the next generation. Natural selection comes into play with whether or not a specific allele is favored by natural selection and gets to successfully spread through the population at the expense of other, less successful alleles.

The two aren't mutually exclusive. You're not providing a synopsis of anything, you're just stating your own ignorance.

> .
> Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> .
> This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur. Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region. This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
> .
> The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step. Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.

How does any of this conflict with the "traditional" view of natural selection?

[Alan Kleinman MD PhD]

Really? So more generations of replication does not contribute more to the gene pool?

>
> >>> It is the total number of replications which
> >>> determines the probability of the next beneficial mutation occurring.
> >>> My model does not address the rate at which these mutations occur,
> >>> only the number of replications necessary for there to be a
> >>> reasonable probability of a beneficial mutation occurring.
> >>
> >> How can you consider that probability without knowing the rate? Or
> >> perhaps you are using your personal definition of "rate" here?
>
> > John, what is the difference in the sample space if you roll 5 dice
> > 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
> Ah, I see that you are using your personal definition of "rate". A
> mutation rate is generally considered to be something like "number of
> expected mutations per site per generation".

Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.

>
> >>> For any
> >>> readers of this thread, perhaps you can find a way to explain this to
> >>> John but I doubt he will understand any explanation until he takes
> >>> the time to learn introductory probability theory.
> >> I'd be interesting in anyone explaining that to me.
>
> > I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
>
> Typical response: having nothing you say, you attempt to denigrate the
> person you're talking to.

John, if you want me to stop responding to you like this, take an introductory course in probability theory.

I don't have to deal with the way in which n changes because it doesn't affect the sample space. The only thing which affects the probability of a beneficial mutation occurring is the number of replications the particular variant is able to do. But you won't understand what I'm saying because you haven't taken an introductory course in probability theory.

[Alan Kleinman MD PhD]

On Monday, September 11, 2017 at 12:10:04 PM UTC-7, Oxyaena wrote:
> On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > .
> > “By the way, Alan, Harshman has been needling me to comment on 
> > how natural selection enters into those peer-reviewed articles of 
> > yours. Would you like to give me a synopsis so I can accurately 
> > judge it for myself?  ”
> > .
> > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > .
>
>
> And?

And those models are inadequate for describing how rmns works.

>
>
> > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
>
>
> CITE? This in no way conflicts with biological fitness, you're confusing sexual selection with natural selection. Sexual selection refers to the competition between potential mates as to who gets to spread their seed, in other words, a competition for the right to fuck with the opposite gender, or, in more technical terms, a competition to decide which potential mate gets to spread its genes unto the next generation. Natural selection comes into play with whether or not a specific allele is favored by natural selection and gets to successfully spread through the population at the expense of other, less successful alleles.

From the same link: "With asexual reproduction, it is sufficient to assign fitnesses to genotypes." Now if you read my posts carefully, I'm saying that reproductive fitness is the correct measure for rmns, not the differential survival of variants. If you are talking about random recombination, differential survival of variants is the correct way to model natural selection. Then again, if you are talking about rmns in sexually reproducing replicators, then reproductive fitness is the correct measure for natural selection.

>
> The two aren't mutually exclusive. You're not providing a synopsis of anything, you're just stating your own ignorance.

And John whines that I'm name calling when I tell him to take an introductory course in probability theory.

>
>
>
> > .
> > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> > .
> > This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur. Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region. This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
> > .
> > The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step. Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.
>
>
> How does any of this conflict with the "traditional" view of natural selection?

When it comes to rmns, it is not the differential survival that determines whether a particular variant can take a step on an evolutionary trajectory, it is the variant's reproductive fitness. Haldane and Kimura model natural selection as differential survival, my model for rmns correctly models natural selection as reproductive fitness.

[John Harshman]

Really? So avocados aren't green and bumpy?

>>>>> It is the total number of replications which
>>>>> determines the probability of the next beneficial mutation occurring.
>>>>> My model does not address the rate at which these mutations occur,
>>>>> only the number of replications necessary for there to be a
>>>>> reasonable probability of a beneficial mutation occurring.
>>>>
>>>> How can you consider that probability without knowing the rate? Or
>>>> perhaps you are using your personal definition of "rate" here?
>>
>>> John, what is the difference in the sample space if you roll 5 dice
>>> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
>> Ah, I see that you are using your personal definition of "rate". A
>> mutation rate is generally considered to be something like "number of
>> expected mutations per site per generation".
> Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.

Have you ever taken an introductory course in evolutionary biology?

>>>>> For any
>>>>> readers of this thread, perhaps you can find a way to explain this to
>>>>> John but I doubt he will understand any explanation until he takes
>>>>> the time to learn introductory probability theory.
>>>> I'd be interesting in anyone explaining that to me.
>>
>>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
>>
>> Typical response: having nothing you say, you attempt to denigrate the
>> person you're talking to.
> John, if you want me to stop responding to you like this, take an introductory course in probability theory.

I don't mind. I just point out that it's a non-response that avoids
engaging with the subject.

Oh, I agree that the probability of a particular mutation occurring
depends only on the number of replications and the mutation rate per
repiclication. But of course that's mutation, not selection, and my
claim, if you recall, is that you don't deal with selection. The
probability of a mutation occurring in n opportunities is not a question
involving natural selection. Since that's all your math is about, your
math doesn't deal with natural selection. And bringing up the fact that
your math deals with mutation is hardly a refutation of my claim.

[Alan Kleinman MD PhD]

Some avocados are black, but more generations of replications do contribute more to the gene pool.

>
> >>>>> It is the total number of replications which
> >>>>> determines the probability of the next beneficial mutation occurring.
> >>>>> My model does not address the rate at which these mutations occur,
> >>>>> only the number of replications necessary for there to be a
> >>>>> reasonable probability of a beneficial mutation occurring.
> >>>>
> >>>> How can you consider that probability without knowing the rate? Or
> >>>> perhaps you are using your personal definition of "rate" here?
> >>
> >>> John, what is the difference in the sample space if you roll 5 dice
> >>> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
> >> Ah, I see that you are using your personal definition of "rate". A
> >> mutation rate is generally considered to be something like "number of
> >> expected mutations per site per generation".
> > Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.
>
> Have you ever taken an introductory course in evolutionary biology?

It wouldn't help in understanding how rmns works. If it would help, you wouldn't be so confused on the subject despite all the courses you've taken in evolutionary biology. An introductory course in probability theory would help you much more if you want to understand stochastic processes like rmns.

>
> >>>>> For any
> >>>>> readers of this thread, perhaps you can find a way to explain this to
> >>>>> John but I doubt he will understand any explanation until he takes
> >>>>> the time to learn introductory probability theory.
> >>>> I'd be interesting in anyone explaining that to me.
> >>
> >>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
> >>
> >> Typical response: having nothing you say, you attempt to denigrate the
> >> person you're talking to.
> > John, if you want me to stop responding to you like this, take an introductory course in probability theory.
>
> I don't mind. I just point out that it's a non-response that avoids
> engaging with the subject.

John, debating with you is like debating with Ray. The challenge I have with you is trying to find a way of explaining this phenomenon when you refuse to try to understand the mathematical principles which govern it. But I don't mind either. It makes me try to find other ways to explain it.

Good, at least we have the n*nG issue behind us. So let's try and understand what natural selection does. As a starting point, do you agree that variants can change in both the absolute number of members and relative frequency with respects to the entire population?

[John Harshman]

All of which -- avocados and generations -- is equally irrelevant to
modeling fitness and natural selection.

>>>>>>> It is the total number of replications which
>>>>>>> determines the probability of the next beneficial mutation occurring.
>>>>>>> My model does not address the rate at which these mutations occur,
>>>>>>> only the number of replications necessary for there to be a
>>>>>>> reasonable probability of a beneficial mutation occurring.
>>>>>>
>>>>>> How can you consider that probability without knowing the rate? Or
>>>>>> perhaps you are using your personal definition of "rate" here?
>>>>
>>>>> John, what is the difference in the sample space if you roll 5 dice
>>>>> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
>>>> Ah, I see that you are using your personal definition of "rate". A
>>>> mutation rate is generally considered to be something like "number of
>>>> expected mutations per site per generation".
>>> Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.
>>
>> Have you ever taken an introductory course in evolutionary biology?

> It wouldn't help in understanding how rmns works. If it would help,
> you wouldn't be so confused on the subject despite all the courses
> you've taken in evolutionary biology. An introductory course in
> probability theory would help you much more if you want to understand
> stochastic processes like rmns.

I'll take that as "no". I'm afraid you won't be able to understand any
of this unless you take an introductory course in evolutionary biology.

>>>>>>> For any
>>>>>>> readers of this thread, perhaps you can find a way to explain this to
>>>>>>> John but I doubt he will understand any explanation until he takes
>>>>>>> the time to learn introductory probability theory.
>>>>>> I'd be interesting in anyone explaining that to me.
>>>>
>>>>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
>>>>
>>>> Typical response: having nothing you say, you attempt to denigrate the
>>>> person you're talking to.
>>> John, if you want me to stop responding to you like this, take an introductory course in probability theory.
>>
>> I don't mind. I just point out that it's a non-response that avoids
>> engaging with the subject.

> John, debating with you is like debating with Ray. The challenge I
> have with you is trying to find a way of explaining this phenomenon
> when you refuse to try to understand the mathematical principles
> which govern it. But I don't mind either. It makes me try to find
> other ways to explain it.

For what it's worth, I do understand the only math you have ever
presented here, your precious multiplication rule of probabilities. And
you never try to find other ways to say anything. You just repeat what
you said before, generally with identical phrasing.

There never was any n*nG issue except for the fact that it doesn't deal
with natural selection.

> So let's try and
> understand what natural selection does. As a starting point, do you
> agree that variants can change in both the absolute number of members
> and relative frequency with respects to the entire population?

Yes. Go on.

[Alan Kleinman MD PhD]

Don't stop there, you have us on the edge of our seats.

>
> >>>>>>> It is the total number of replications which
> >>>>>>> determines the probability of the next beneficial mutation occurring.
> >>>>>>> My model does not address the rate at which these mutations occur,
> >>>>>>> only the number of replications necessary for there to be a
> >>>>>>> reasonable probability of a beneficial mutation occurring.
> >>>>>>
> >>>>>> How can you consider that probability without knowing the rate? Or
> >>>>>> perhaps you are using your personal definition of "rate" here?
> >>>>
> >>>>> John, what is the difference in the sample space if you roll 5 dice
> >>>>> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
> >>>> Ah, I see that you are using your personal definition of "rate". A
> >>>> mutation rate is generally considered to be something like "number of
> >>>> expected mutations per site per generation".
> >>> Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.
> >>
> >> Have you ever taken an introductory course in evolutionary biology?
>
> > It wouldn't help in understanding how rmns works. If it would help,
> > you wouldn't be so confused on the subject despite all the courses
> > you've taken in evolutionary biology. An introductory course in
> > probability theory would help you much more if you want to understand
> > stochastic processes like rmns.
> I'll take that as "no". I'm afraid you won't be able to understand any
> of this unless you take an introductory course in evolutionary biology.

I know enough about evolutionary biology to find an exact solution to Haldane's equation for which he only gives an approximate solution. Did they teach you how to do that in your evolutionary biology courses? You don't have to answer that, we already know that's a no. So teach us how reptiles grow feathers.

>
> >>>>>>> For any
> >>>>>>> readers of this thread, perhaps you can find a way to explain this to
> >>>>>>> John but I doubt he will understand any explanation until he takes
> >>>>>>> the time to learn introductory probability theory.
> >>>>>> I'd be interesting in anyone explaining that to me.
> >>>>
> >>>>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
> >>>>
> >>>> Typical response: having nothing you say, you attempt to denigrate the
> >>>> person you're talking to.
> >>> John, if you want me to stop responding to you like this, take an introductory course in probability theory.
> >>
> >> I don't mind. I just point out that it's a non-response that avoids
> >> engaging with the subject.
>
> > John, debating with you is like debating with Ray. The challenge I
> > have with you is trying to find a way of explaining this phenomenon
> > when you refuse to try to understand the mathematical principles
> > which govern it. But I don't mind either. It makes me try to find
> > other ways to explain it.
>
> For what it's worth, I do understand the only math you have ever
> presented here, your precious multiplication rule of probabilities. And
> you never try to find other ways to say anything. You just repeat what
> you said before, generally with identical phrasing.

John, I presented the key equation, the fundamental binomial probability equation. Every step on an evolutionary trajectory is of this form where each step is linked to the others by the multiplication rule of probabilities.

It does if you understand that fitness to reproduced is measured by the number of replications.

>
> > So let's try and
> > understand what natural selection does. As a starting point, do you
> > agree that variants can change in both the absolute number of members
> > and relative frequency with respects to the entire population?
>
> Yes. Go on.

So does the probability of a beneficial mutation occurring on some member of the variant depend on the absolute number of members of the variant (ie the number of replications) or the relative frequency of the variant with respects to the entire population (hint, n*nG)?

[John Harshman]

Us? I doubt anyone else is reading this. It's a low impact factor post.
And once again you make a lame joke to evade the point.

Again, us? Nobody knows just how "reptile grow feathers". But we do know
it happened. Again, have you looked up Caudipteryx?

>>>>>>>>> For any
>>>>>>>>> readers of this thread, perhaps you can find a way to explain this to
>>>>>>>>> John but I doubt he will understand any explanation until he takes
>>>>>>>>> the time to learn introductory probability theory.
>>>>>>>> I'd be interesting in anyone explaining that to me.
>>>>>>
>>>>>>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
>>>>>>
>>>>>> Typical response: having nothing you say, you attempt to denigrate the
>>>>>> person you're talking to.
>>>>> John, if you want me to stop responding to you like this, take an introductory course in probability theory.
>>>>
>>>> I don't mind. I just point out that it's a non-response that avoids
>>>> engaging with the subject.
>>
>>> John, debating with you is like debating with Ray. The challenge I
>>> have with you is trying to find a way of explaining this phenomenon
>>> when you refuse to try to understand the mathematical principles
>>> which govern it. But I don't mind either. It makes me try to find
>>> other ways to explain it.
>>
>> For what it's worth, I do understand the only math you have ever
>> presented here, your precious multiplication rule of probabilities. And
>> you never try to find other ways to say anything. You just repeat what
>> you said before, generally with identical phrasing.

> John, I presented the key equation, the fundamental binomial
> probability equation. Every step on an evolutionary trajectory is of
> this form where each step is linked to the others by the
> multiplication rule of probabilities.

Your equation is all about the probability of two particular mutations.
Nothing about selection.

No it isn't. It's measured by changes in the number or frequency of
particular genotypes.

>>> So let's try and
>>> understand what natural selection does. As a starting point, do you
>>> agree that variants can change in both the absolute number of members
>>> and relative frequency with respects to the entire population?
>>
>> Yes. Go on.

> So does the probability of a beneficial mutation occurring on some
> member of the variant depend on the absolute number of members of the
> variant (ie the number of replications) or the relative frequency of
> the variant with respects to the entire population (hint, n*nG)?

It depends on the absolute number. Of course if the population is of
constant size, as you implicitly assume, number and frequency are
equivalent measures. And did you note that you're asking about the
probability of a mutation? That's mutation, not selection.

[Alan Kleinman MD PhD]

Too bad, I thought you were going to tell us how generations don't have any effect on fitness. What if fitness is zero? How many generations for that variant?

And everyone knows that aliens landed at area 51.

Other than the probability remains low if the reproductive fitness is low, not enough replications.

John, the population size does not remain constant, reproductive fitness determines what that population size is. That's how natural selection works for rmns.

[John Harshman]

You're just exposing your ignorance of selection. What you're saying
there isn't even wrong, just nonsensical assembly of words.

Again, avoidance with a stupid joke. Do you even know what Caudipteryx is?

True, but you don't model changes in the number of replications. You
don't consider fitness in determining the number of replications.

Then why did you model a constant population size? And yes, natural
selection can happen in a population of constant size; why would you
think otherwise?

[Alan Kleinman MD PhD]

I get your argument, no generations, no replicators, no replicators, no selection. Your argument makes great guacamole.

Of course I do, the aliens have feathers.

The number of replications reflects the fitness.

Apparently, I did it to confuse you. And it has worked way beyond what I had hoped. Any chance I can get a pillow filled with Caudipteryx feathers?

[John Harshman]

Whenever you can't deal with an issue, you deflect, avoid, make a joke.

Deflect, avoid, make a joke.

It's true that the number of replications in a particular period has
some relation to fitness. But the actually way to deal with fitness is
to use some function to deal with changes in number of replications per
generation over time, and you don't do any of that.

Deflection, avoidance, joke.

[Peter Nyikos]

I think you could make a better case for yourself by talking
about your "three drug cocktail" scenario and showing why it
works so much better than "one drug at a time, until resistance
makes the last drug ineffective against a burgeoning population".

There is a lot of natural selection implicit in this scenario, and Harshman
could easily see it if he weren't so stubbornly lazy. He keeps insinuating
that your knowledge of natural selection is close to nonexistent,
and a concrete example like this could lay these insinuations to rest.

Don't expect anyone else to do it for you. I am too busy with Harshman
and Martinez's chicanery in other issues, and most of the others are
not going to lift a finger to help you: Martinez keeps trying to nail
you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
and almost everyone else is too interested in seeing another creationist
--yourself-- go down in flames to be objective about what is really going on.

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina
http://people.math.sc.edu/nyikos/

[John Harshman]

I already know why it works better than one drug at a time. This is not
at issue. One issue (though not the main one) is that Alan thinks his
drug cocktail scenario is analogous to every possible instance of
natural selection.

> There is a lot of natural selection implicit in this scenario, and Harshman
> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> that your knowledge of natural selection is close to nonexistent,
> and a concrete example like this could lay these insinuations to rest.

You mistake several things here. You mistake my main point, which is not
that Alan doesn't know anything about natural selection (though that
also seems true) but that his mathematical models do not consider it. He
thinks that "rmns" is a single thing, models "rm" and thinks he has
modeled "ns" too.

> Don't expect anyone else to do it for you. I am too busy with Harshman
> and Martinez's chicanery in other issues, and most of the others are
> not going to lift a finger to help you: Martinez keeps trying to nail
> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> and almost everyone else is too interested in seeing another creationist
> --yourself-- go down in flames to be objective about what is really going on.

What is really going on? I don't think you have a clue.

[r3p...@gmail.com]

Alan: Peter is wrong, I have not attempted to nail YOU as an Atheist.

For the Record:

What I do say about Theists who accept the concept of natural selection to exist in nature: The concept was produced by the assumptions of Naturalism, which are pro-Atheism. In this factual context I then observe that a contradiction exists: Theists accepting claims produced in service to Naturalism.

Ray

[Peter Nyikos]

On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> .
> “By the way, Alan, Harshman has been needling me to comment on 
> how natural selection enters into those peer-reviewed articles of 
> yours. Would you like to give me a synopsis so I can accurately 
> judge it for myself?  ”

What I meant was "a synopsis of those two papers." You obviously took
it differently.

> Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.

Why "leads to"? Some of the other variants might hang on to their
relative frequency or even improve it while the formerly populous ones might
undergo a severe decline.

> Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> .
> However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
>

You call this "conceptual conflict," I call it "attention to different
contributions to overall relative frequency."

This applies to your next paragraph also.

> Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> .

> This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur.

Of course. But if one variant has a big advantage over another [like in
your drug cocktail scenario, where the resistant variants pretty much
take over the huge environment of the host from the nonresistant ones]
then there will be replications galore.

In his typically hasty reply to my first post to this thread, Harshman
claimed "that Alan thinks his drug cocktail scenario is analogous to
every possible instance of natural selection." Yet here, you don't
seem to give it the weight it deserves!

> Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region.

Here too you are downplaying the advantages of the resistant variant.

> This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
> .

The two can be very strongly linked, see above.

> The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step.

But the reproductive success is intimately tied in with overall
frequency.

> Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.

No argument there. But you haven't given me a synopsis of those
two papers, which MIGHT show me how this, and hence natural
selection, is incorporated.

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer--
University of South Carolina

http://www.math.sc.edu/~nyikos/

[Peter Nyikos]

What you say next is the most significant way in which you keep attempting
to nail ME as an "Atheist".

> For the Record:
>
> What I do say about Theists who accept the concept of natural selection to exist in nature: The concept was produced by the assumptions of Naturalism, which are pro-Atheism.

And then you take the next step and accuse me of Atheism on these
guilt-by-association grounds.

>In this factual context I then observe that a contradiction exists: Theists accepting claims produced in service to Naturalism.

Yes, but you usually use the word "contradiction" in the standard
English sense rather than the Marxist-English sense in which you
are using it here. For instance, you go on to say that this acceptance of
natural selection falsifies the claim of Theism, or following Jesus, etc.

I think the reason for this difference is that Kleinman has not
accused you of violating the commandment of Jesus, "Do not bear
false witness" nor any other form of dishonesty and/or hypocrisy,
whereas I have nailed you on these things repeatedly.

And so you have a vested interest in the revenge you get by
calling me an Atheist on grounds like the above.

Peter Nyikos

[Alan Kleinman MD PhD]

Whenever you can't deal with an issue, you change the subject.

Stop trying to change the subject. Who wants to talk about reading fossil tea leaves anyway? Except maybe you.

I used a very simple function (n*nG) as an initial expediency in the derivation of the equations. If you want, I'll teach you what a sample space is and why this is a valid approximation (unlike your linear approxiation of doubling the population doubles the probability of a beneficial mutation).

Don't lose your sense of humor, you are going to need it.

[Alan Kleinman MD PhD]

You haven't followed the threads very carefully. I've used this example many times. The basic argument being made against this example is that nature doesn't target populations with three toxins simultaneously. Those using this argument don't realize that most "natural" selection pressures generally target multiple genetic loci simultaneously. Thermal stress affects the function of virtually all enzymes, starvation affects every metabolic pathway which requires energy to function, dehydration same thing.

>
> There is a lot of natural selection implicit in this scenario, and Harshman
> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> that your knowledge of natural selection is close to nonexistent,
> and a concrete example like this could lay these insinuations to rest.

I agree with you, the suppression of the evolution of HIV by three drug therapy is the quintessential example of why multiple selection pressures inhibit the rmns phenomenon. HIV has everything going for it when it comes to rmns. It has huge populations, high mutation rates, it does recombination but the multiplication rule of probabilities stops it.

>
> Don't expect anyone else to do it for you. I am too busy with Harshman
> and Martinez's chicanery in other issues, and most of the others are
> not going to lift a finger to help you: Martinez keeps trying to nail
> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> and almost everyone else is too interested in seeing another creationist
> --yourself-- go down in flames to be objective about what is really going on.

The only thing going down in flames here is the theory of evolution. In exchange, people will learn how to reduce or prevent multi-drug resistant microbes, multi-herbicide resistant weeds, and produce more durable cancer treatments. Not a bad exchange for a theory atheists use to argue God doesn't exist.

[Alan Kleinman MD PhD]

HIV's failure to evolve to three drug cocktail demonstrates exactly how natural selection works, the multiplication rule of probabilities in action. This is why this scenario is analogous to every possible instance of natural selection. You could disprove my claim by producing a real, measurable and repeatable example of rmns that works otherwise. But you won't because you can't.

>
> > There is a lot of natural selection implicit in this scenario, and Harshman
> > could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> > that your knowledge of natural selection is close to nonexistent,
> > and a concrete example like this could lay these insinuations to rest.
>
> You mistake several things here. You mistake my main point, which is not
> that Alan doesn't know anything about natural selection (though that
> also seems true) but that his mathematical models do not consider it. He
> thinks that "rmns" is a single thing, models "rm" and thinks he has
> modeled "ns" too.

John, what term in either Haldane's model or Kimura's model is the "natural selection" term?

>
> > Don't expect anyone else to do it for you. I am too busy with Harshman
> > and Martinez's chicanery in other issues, and most of the others are
> > not going to lift a finger to help you: Martinez keeps trying to nail
> > you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> > and almost everyone else is too interested in seeing another creationist
> > --yourself-- go down in flames to be objective about what is really going on.
>
> What is really going on? I don't think you have a clue.

Hmmm

[Alan Kleinman MD PhD]

I'm going to break my promise here not to respond to any of your posts until you watch the Kahn Academy lectures on probability theory on youtube and say that it is very easy to disprove atheism. Ask the avowed atheists if they know everything in the universe. The more logical approach to this subject is to be Agnostic.

[John Harshman]

Ah, I had forgotten the kindergarten response: oh yeah, you're another.
OK. I'll add it to the list.

Another example of deflect, avoid, make a joke.

It seems to have been a final expediency too. So how does your math deal
with selection?

Deflection, avoidance, joke.

[John Harshman]

Not exactly. The main argument is that nature doesn't usually target
populations with three selective agents that cause there to be no
benefit for a response to just one of them. Your non-standard usage of
"beneficial" helps to disguise that assumption, but not enough.

> Those using this argument don't realize that most
> "natural" selection pressures generally target multiple genetic loci
> simultaneously. Thermal stress affects the function of virtually all
> enzymes, starvation affects every metabolic pathway which requires
> energy to function, dehydration same thing.

To the extent that no mutation at any single locus is beneficial?

>> There is a lot of natural selection implicit in this scenario, and Harshman
>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
>> that your knowledge of natural selection is close to nonexistent,
>> and a concrete example like this could lay these insinuations to rest.

> I agree with you, the suppression of the evolution of HIV by three
> drug therapy is the quintessential example of why multiple selection
> pressures inhibit the rmns phenomenon. HIV has everything going for it
> when it comes to rmns. It has huge populations, high mutation rates, it
> does recombination but the multiplication rule of probabilities stops it.

This is a peripheral issue, but I have my doubts about the frequency of
recombination. Does HIV recombine genomes infesting different cells or
only the same cell? If the latter, then the effective population size
for recombination events is much smaller than you imply.

>> Don't expect anyone else to do it for you. I am too busy with Harshman
>> and Martinez's chicanery in other issues, and most of the others are
>> not going to lift a finger to help you: Martinez keeps trying to nail
>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
>> and almost everyone else is too interested in seeing another creationist
>> --yourself-- go down in flames to be objective about what is really going on.

> The only thing going down in flames here is the theory of evolution.
> In exchange, people will learn how to reduce or prevent multi-drug
> resistant microbes, multi-herbicide resistant weeds, and produce more
> durable cancer treatments. Not a bad exchange for a theory atheists
> use to argue God doesn't exist.

And so we see Alan's true motivation, only thinly disguised, if at all.
He's closer to Ray that either of them supposes.

[John Harshman]

Once again, the multiplication rule of probabilities has nothing to do
with selection, just mutation. You do not treat selection at all in your
math. Your example actually shows the absence of selection, since in the
environment of the cocktail no single mutation has a selective advantage
over the unmutated form. Selection can only occur after a form resistant
to all three drugs emerges by chance, and that's where the
multiplication rule comes leaves off.

As for examples, is it your claim that most species in the wild are
experiencing only a single selection pressure at any time?

>>> There is a lot of natural selection implicit in this scenario, and Harshman
>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
>>> that your knowledge of natural selection is close to nonexistent,
>>> and a concrete example like this could lay these insinuations to rest.
>>
>> You mistake several things here. You mistake my main point, which is not
>> that Alan doesn't know anything about natural selection (though that
>> also seems true) but that his mathematical models do not consider it. He
>> thinks that "rmns" is a single thing, models "rm" and thinks he has
>> modeled "ns" too.

> John, what term in either Haldane's model or Kimura's model is the "natural selection" term?

Fitness, the term that deals with expected change in frequency (or
numbers if you prefer) from one generation to the next.

>>> Don't expect anyone else to do it for you. I am too busy with Harshman
>>> and Martinez's chicanery in other issues, and most of the others are
>>> not going to lift a finger to help you: Martinez keeps trying to nail
>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
>>> and almost everyone else is too interested in seeing another creationist
>>> --yourself-- go down in flames to be objective about what is really going on.
>>
>> What is really going on? I don't think you have a clue.

> Hmmm

You would seem to agree, based on your comments to him above.

[r3p...@gmail.com]

You've said this twice now, using my name inaccurately. So I have the right to barge in and remind the group of what I have said against your claims:

Alan claims that he has seen RMNS in his medical laboratory. Yet I've pointed out that the accepted **conceptual** model of natural selection in the wild consists of multiple inferences which then result in the occurrence of micro-evolution over generational time. So Alan has not seen natural selection in his medical laboratory; rather, he has seen "natural selection."

I've also pointed out that unintelligent processes (RMNS) cannot be stumping our most brilliant medical minds and their computers concerning, for example, antibiotic resistance, mutating disease. As a matter of fact unintelligence cannot be stumping intelligence, simply impossible; therefore the causal agents must be described accordingly in teleological terms, not counter factually in non-teleological terms. The Bible most clearly states that God unleashes disease into the biosphere. Alan has never answered these points except to restate his claim that RMNS exists in his medical laboratory.

And the only thing I said about Alan's use of mathematics and probability theory is the fact that he has exalted the same into a place of preeminence, when in fact observation and logic occupy the place of preeminence. Alan has the proverbial cart before the horse because he really believes mathematics and probability theory are preeminent, when in fact the same are always supplementary to observation and logic.

Moreover, Alan has admitted that certain persons, presumably possessing Ph.Ds, who peer-reviewed some of his work, were unqualified to do so because they did not understand the mathematics! In this context I observed that Alan's claims have ZERO chance of enacting harm onto macro-evolutionary theory because only a handful can understand his claims. Yet we are told that science, unlike religion, is for everyone. But in Alan's world he and a few others are king. Sorry, Alan, observation and words (logic) are king.

I wish Alan all the luck in the world. I hope he obtains some degree of satisfying success. But I've studied evolutionary theory long enough to know that Darwinists will not abandon the fact of evolution, via nested hierarchies, just because the numbers don't add up.

Ray (Old Earth; species immutabilist)

[Alan Kleinman MD PhD]

On Tuesday, September 12, 2017 at 2:50:05 PM UTC-7, Peter Nyikos wrote:
> On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > .
> > “By the way, Alan, Harshman has been needling me to comment on 
> > how natural selection enters into those peer-reviewed articles of 
> > yours. Would you like to give me a synopsis so I can accurately 
> > judge it for myself?  ”
>
> What I meant was "a synopsis of those two papers." You obviously took
> it differently.

My first paper address the mathematics which govern rmns for a single selection pressure targeting a single gene. The derivation of the equations was based on empirical example published by Weinreich and others: "Darwinian evolution can follow only very few mutational paths to fitter proteins." My other paper on rmns addressed the mathematics of multiple simultaneous selection pressures and was based on the empirical example published by our friend Bill Rogers, "Failure of artesunate–mefloquine combination therapy for uncomplicated Plasmodium falciparum malaria in southern Cambodia"

>
>
> > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.
>
> Why "leads to"? Some of the other variants might hang on to their
> relative frequency or even improve it while the formerly populous ones might
> undergo a severe decline.

Agreed, that's why relative frequency is not the appropriate variable to model rmns. The absolute number of replications is the appropriate variable because the replication is the random trial which is key to determining the probability of a beneficial mutation occurring.

>
>
>
>
> > Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > .
> > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
> >
>
> You call this "conceptual conflict," I call it "attention to different
> contributions to overall relative frequency."

The relative frequency of different variants in a population has no bearing on the mathematics of rmns.

>
> This applies to your next paragraph also.
>
> > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> > .
>
> > This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur.
>
> Of course. But if one variant has a big advantage over another [like in
> your drug cocktail scenario, where the resistant variants pretty much
> take over the huge environment of the host from the nonresistant ones]
> then there will be replications galore.

The drug sensitive variants don't disappear from the world, they just can't grow in the environment with the drug. I can tell you from direct personal experience that there are still plenty of drug sensitive bacteria around.

>
> In his typically hasty reply to my first post to this thread, Harshman
> claimed "that Alan thinks his drug cocktail scenario is analogous to
> every possible instance of natural selection." Yet here, you don't
> seem to give it the weight it deserves!

If you read my publications, I explain that the reason HIV combination therapy works is because of the mathematics I presented. This quote is from my first publication: "Any disruption of this beneficial mutation/amplification of beneficial mutation cycle will stifle the
evolutionary process. Introducing a second selection pressure, which targets a different genetic locus than the first selection pressure along with the first selection pressure, forces the population to attempt to take two evolutionary trajectories simultaneously and will disrupt the evolutionary cycle. It is this principle, which has led to the success of combination therapy for the treatment of HIV [4]"

>
>
>
> > Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region.
>
> Here too you are downplaying the advantages of the resistant variant.

Au contraire, I'm explaining how resistant variants come about. Once you understand this, you can take the proper steps to prevent it.

>
>
> > This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
> > .
>
> The two can be very strongly linked, see above.

The linkage occurs when one lineage is competing for the same resources of the environment as other lineages. The empirical example of this is the Lenski experiment. His bacteria are not only evolving against starvation pressure but are competing for the limiting resource of the environment. But the mathematics of rmns is not changed in this circumstance, it only slows the ability of the slightly more fit variant to accumulate the replications necessary for the next beneficial mutation.

>
>
> > The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step.
>
> But the reproductive success is intimately tied in with overall
> frequency.

Only when the environment puts limits on resources as well as applying other stressors to the different lineages. The mathematics of rmns remains the same. The ability of a lineage to accumulate beneficial mutations against a selection pressure is dependent only on its reproductive success on each evolutionary step. Other competitors in the environment limit their reproductive success and slow the evolutionary process. This is why Lenski's experiment takes more than 1000 generations for each beneficial mutation.

>
>
> > Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.
>
> No argument there. But you haven't given me a synopsis of those
> two papers, which MIGHT show me how this, and hence natural
> selection, is incorporated.

See above.

[John Harshman]

On 9/12/17 3:54 PM, r3p...@gmail.com wrote:

> I've also pointed out that unintelligent processes (RMNS) cannot be
> stumping our most brilliant medical minds and their computers
> concerning, for example, antibiotic resistance, mutating disease. As
> a matter of fact unintelligence cannot be stumping intelligence,
> simply impossible; therefore the causal agents must be described
> accordingly in teleological terms, not counter factually in
> non-teleological terms. The Bible most clearly states that God
> unleashes disease into the biosphere. Alan has never answered these
> points except to restate his claim that RMNS exists in his medical
> laboratory.

God is trying to kill us?

[Alan Kleinman MD PhD]

That explains why farmers never have trouble growing their crops or animals. Weeds, insects, droughts, floods, thermal stress, birds (animals w/ feathers), predation, nutrient deficits, diseases...

>
> > Those using this argument don't realize that most
> > "natural" selection pressures generally target multiple genetic loci
> > simultaneously. Thermal stress affects the function of virtually all
> > enzymes, starvation affects every metabolic pathway which requires
> > energy to function, dehydration same thing.
>
> To the extent that no mutation at any single locus is beneficial?

Some selection pressures are like that. Now antibiotics and antiviral agents have the special property that they (generally) only target a single genetic locus.

>
> >> There is a lot of natural selection implicit in this scenario, and Harshman
> >> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> >> that your knowledge of natural selection is close to nonexistent,
> >> and a concrete example like this could lay these insinuations to rest.
>
> > I agree with you, the suppression of the evolution of HIV by three
> > drug therapy is the quintessential example of why multiple selection
> > pressures inhibit the rmns phenomenon. HIV has everything going for it
> > when it comes to rmns. It has huge populations, high mutation rates, it
> > does recombination but the multiplication rule of probabilities stops it.
>
> This is a peripheral issue, but I have my doubts about the frequency of
> recombination. Does HIV recombine genomes infesting different cells or
> only the same cell? If the latter, then the effective population size
> for recombination events is much smaller than you imply.

John, I wrote a paper explaining when random recombination has a reasonable probability of giving a multi-drug resistant offspring. If you took a course in basic probability theory, you might understand it.

>
> >> Don't expect anyone else to do it for you. I am too busy with Harshman
> >> and Martinez's chicanery in other issues, and most of the others are
> >> not going to lift a finger to help you: Martinez keeps trying to nail
> >> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> >> and almost everyone else is too interested in seeing another creationist
> >> --yourself-- go down in flames to be objective about what is really going on.
>
> > The only thing going down in flames here is the theory of evolution.
> > In exchange, people will learn how to reduce or prevent multi-drug
> > resistant microbes, multi-herbicide resistant weeds, and produce more
> > durable cancer treatments. Not a bad exchange for a theory atheists
> > use to argue God doesn't exist.
> And so we see Alan's true motivation, only thinly disguised, if at all.
> He's closer to Ray that either of them supposes.

This is a bad time for atheists who understand basic probability theory.

[John Harshman]

All these multiple selective agents have a benefit to response to just
one of them. I think you have shot yourself in the foot there, as the
logical implication of your claim is that there is no selection in
nature, just extinction of pretty much every population.

>>> Those using this argument don't realize that most
>>> "natural" selection pressures generally target multiple genetic loci
>>> simultaneously. Thermal stress affects the function of virtually all
>>> enzymes, starvation affects every metabolic pathway which requires
>>> energy to function, dehydration same thing.
>>
>> To the extent that no mutation at any single locus is beneficial?

> Some selection pressures are like that. Now antibiotics and antiviral
> agents have the special property that they (generally) only target a
> single genetic locus.

Those two sentences don't relate to each other, though I know you think
they do. Your viral scenario relies on the assumption that in a
multidrug environment, resistance to a single drug confers no benefit.

>>>> There is a lot of natural selection implicit in this scenario, and Harshman
>>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
>>>> that your knowledge of natural selection is close to nonexistent,
>>>> and a concrete example like this could lay these insinuations to rest.
>>
>>> I agree with you, the suppression of the evolution of HIV by three
>>> drug therapy is the quintessential example of why multiple selection
>>> pressures inhibit the rmns phenomenon. HIV has everything going for it
>>> when it comes to rmns. It has huge populations, high mutation rates, it
>>> does recombination but the multiplication rule of probabilities stops it.
>>
>> This is a peripheral issue, but I have my doubts about the frequency of
>> recombination. Does HIV recombine genomes infesting different cells or
>> only the same cell? If the latter, then the effective population size
>> for recombination events is much smaller than you imply.

> John, I wrote a paper explaining when random recombination has a
> reasonable probability of giving a multi-drug resistant offspring. If
> you took a course in basic probability theory, you might understand
> it.

That doesn't answer my question. I'm assuming you don't actually know
the answer. Correct?

>>>> Don't expect anyone else to do it for you. I am too busy with Harshman
>>>> and Martinez's chicanery in other issues, and most of the others are
>>>> not going to lift a finger to help you: Martinez keeps trying to nail
>>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
>>>> and almost everyone else is too interested in seeing another creationist
>>>> --yourself-- go down in flames to be objective about what is really going on.
>>
>>> The only thing going down in flames here is the theory of evolution.
>>> In exchange, people will learn how to reduce or prevent multi-drug
>>> resistant microbes, multi-herbicide resistant weeds, and produce more
>>> durable cancer treatments. Not a bad exchange for a theory atheists
>>> use to argue God doesn't exist.
>> And so we see Alan's true motivation, only thinly disguised, if at all.
>> He's closer to Ray that either of them supposes.

> This is a bad time for atheists who understand basic probability theory.

It's a bad time for creationists who understand anything about geology,
paleontology, or biology. Now I know you aren't interested in those
subjects and know nothing about them, so it's not a bad time for you.
Ignorance is strength.

[Alan Kleinman MD PhD]

Even with a single selection pressure targeting a single gene, the multiplication rule of probabilities governs. Microevolutionary changes don't add, they are linked by the multiplication rule of probability.

>
> As for examples, is it your claim that most species in the wild are
> experiencing only a single selection pressure at any time?

Absolutely not.

>
> >>> There is a lot of natural selection implicit in this scenario, and Harshman
> >>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> >>> that your knowledge of natural selection is close to nonexistent,
> >>> and a concrete example like this could lay these insinuations to rest.
> >>
> >> You mistake several things here. You mistake my main point, which is not
> >> that Alan doesn't know anything about natural selection (though that
> >> also seems true) but that his mathematical models do not consider it. He
> >> thinks that "rmns" is a single thing, models "rm" and thinks he has
> >> modeled "ns" too.
>
> > John, what term in either Haldane's model or Kimura's model is the "natural selection" term?
>
> Fitness, the term that deals with expected change in frequency (or
> numbers if you prefer) from one generation to the next.

And I use number of replication correctly to determine the probability of a beneficial mutation occurring.

>
> >>> Don't expect anyone else to do it for you. I am too busy with Harshman
> >>> and Martinez's chicanery in other issues, and most of the others are
> >>> not going to lift a finger to help you: Martinez keeps trying to nail
> >>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> >>> and almost everyone else is too interested in seeing another creationist
> >>> --yourself-- go down in flames to be objective about what is really going on.
> >>
> >> What is really going on? I don't think you have a clue.
>
> > Hmmm
>
> You would seem to agree, based on your comments to him above.

Peter sees a linkage between the number of members of each variant and relative frequency of variants. And he is correct. He just needs to learn how this linkage effects rmns. The Kishony experiment uncouples this effect by having a large environment that doesn't cause competition between lineages. The Lenski experiment doesn't uncouple this effect and causes the rmns process to slow down as lineages compete for the limited resources.

[John Harshman]

That has nothing to do with selection. That's mutation. Selection is
about differential reproductive success of genotypes, not about the
probability of new mutations.

>> As for examples, is it your claim that most species in the wild are
>> experiencing only a single selection pressure at any time?

> Absolutely not.

Then it must be your claim that there is not selection occurring in the
wild, since it's also your claim that evolution can't occur when there
are three or more selection pressures.

>>>>> There is a lot of natural selection implicit in this scenario, and Harshman
>>>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
>>>>> that your knowledge of natural selection is close to nonexistent,
>>>>> and a concrete example like this could lay these insinuations to rest.
>>>>
>>>> You mistake several things here. You mistake my main point, which is not
>>>> that Alan doesn't know anything about natural selection (though that
>>>> also seems true) but that his mathematical models do not consider it. He
>>>> thinks that "rmns" is a single thing, models "rm" and thinks he has
>>>> modeled "ns" too.
>>
>>> John, what term in either Haldane's model or Kimura's model is the "natural selection" term?
>>
>> Fitness, the term that deals with expected change in frequency (or
>> numbers if you prefer) from one generation to the next.

> And I use number of replication correctly to determine the probability of a beneficial mutation occurring.

But you don't consider change in the number of replications, which is
what selection is. And the probability of a mutation occurring is
mutation, not selection.

>>>>> Don't expect anyone else to do it for you. I am too busy with Harshman
>>>>> and Martinez's chicanery in other issues, and most of the others are
>>>>> not going to lift a finger to help you: Martinez keeps trying to nail
>>>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
>>>>> and almost everyone else is too interested in seeing another creationist
>>>>> --yourself-- go down in flames to be objective about what is really going on.
>>>>
>>>> What is really going on? I don't think you have a clue.
>>
>>> Hmmm
>>
>> You would seem to agree, based on your comments to him above.

> Peter sees a linkage between the number of members of each variant
> and relative frequency of variants.

So does anyone with a working brain. So?

[Alan Kleinman MD PhD]

Tell that to the farmer who uses a herbicide only to watch his crops die from insects, droughts, floods, thermal stress, birds (animals with feathers), nutrient deficitis, diseases...

>
> >>> Those using this argument don't realize that most
> >>> "natural" selection pressures generally target multiple genetic loci
> >>> simultaneously. Thermal stress affects the function of virtually all
> >>> enzymes, starvation affects every metabolic pathway which requires
> >>> energy to function, dehydration same thing.
> >>
> >> To the extent that no mutation at any single locus is beneficial?
>
> > Some selection pressures are like that. Now antibiotics and antiviral
> > agents have the special property that they (generally) only target a
> > single genetic locus.
> Those two sentences don't relate to each other, though I know you think
> they do. Your viral scenario relies on the assumption that in a
> multidrug environment, resistance to a single drug confers no benefit.

They do if you understand how rmns works which you don't or are unwilling to.

>
> >>>> There is a lot of natural selection implicit in this scenario, and Harshman
> >>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> >>>> that your knowledge of natural selection is close to nonexistent,
> >>>> and a concrete example like this could lay these insinuations to rest.
> >>
> >>> I agree with you, the suppression of the evolution of HIV by three
> >>> drug therapy is the quintessential example of why multiple selection
> >>> pressures inhibit the rmns phenomenon. HIV has everything going for it
> >>> when it comes to rmns. It has huge populations, high mutation rates, it
> >>> does recombination but the multiplication rule of probabilities stops it.
> >>
> >> This is a peripheral issue, but I have my doubts about the frequency of
> >> recombination. Does HIV recombine genomes infesting different cells or
> >> only the same cell? If the latter, then the effective population size
> >> for recombination events is much smaller than you imply.
>
> > John, I wrote a paper explaining when random recombination has a
> > reasonable probability of giving a multi-drug resistant offspring. If
> > you took a course in basic probability theory, you might understand
> > it.
> That doesn't answer my question. I'm assuming you don't actually know
> the answer. Correct?

John, my paper gives the upper limits of the probability of random recombination giving a multi-drug resistant offspring. If you want to think of reasons to lower that probability, be my guest.

>
> >>>> Don't expect anyone else to do it for you. I am too busy with Harshman
> >>>> and Martinez's chicanery in other issues, and most of the others are
> >>>> not going to lift a finger to help you: Martinez keeps trying to nail
> >>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> >>>> and almost everyone else is too interested in seeing another creationist
> >>>> --yourself-- go down in flames to be objective about what is really going on.
> >>
> >>> The only thing going down in flames here is the theory of evolution.
> >>> In exchange, people will learn how to reduce or prevent multi-drug
> >>> resistant microbes, multi-herbicide resistant weeds, and produce more
> >>> durable cancer treatments. Not a bad exchange for a theory atheists
> >>> use to argue God doesn't exist.
> >> And so we see Alan's true motivation, only thinly disguised, if at all.
> >> He's closer to Ray that either of them supposes.
>
> > This is a bad time for atheists who understand basic probability theory.
>
> It's a bad time for creationists who understand anything about geology,
> paleontology, or biology. Now I know you aren't interested in those
> subjects and know nothing about them, so it's not a bad time for you.
> Ignorance is strength.

I don't mind seeing an academic mud wrestling match. I think the soft sciences like biology are at a big disadvantage though. I can see this disadvantage when I debate with you.

[Oxyaena]

On Tuesday, September 12, 2017 at 7:10:05 PM UTC-4, Alan Kleinman MD PhD wrote:
> On Tuesday, September 12, 2017 at 2:50:05 PM UTC-7, Peter Nyikos wrote:
> > On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > > .
> > > “By the way, Alan, Harshman has been needling me to comment on 
> > > how natural selection enters into those peer-reviewed articles of 
> > > yours. Would you like to give me a synopsis so I can accurately 
> > > judge it for myself?  ”
> >
> > What I meant was "a synopsis of those two papers." You obviously took
> > it differently.
> My first paper address the mathematics which govern rmns for a single selection pressure targeting a single gene. The derivation of the equations was based on empirical example published by Weinreich and others: "Darwinian evolution can follow only very few mutational paths to fitter proteins." My other paper on rmns addressed the mathematics of multiple simultaneous selection pressures and was based on the empirical example published by our friend Bill Rogers, "Failure of artesunate–mefloquine combination therapy for uncomplicated Plasmodium falciparum malaria in southern Cambodia"

You haven't demonstrated that your hypothesis has any bearing on evolution outside of obvious facts cloaked by unnecessary technobabble and bullshit statistics. Evolution is mostly unpredictable, and therefore math doesn't do you much good when it comes to evolution (or even biology in general), unlike, say, in physics.

> >
> >
> > > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.
> >
> > Why "leads to"? Some of the other variants might hang on to their
> > relative frequency or even improve it while the formerly populous ones might
> > undergo a severe decline.
> Agreed, that's why relative frequency is not the appropriate variable to model rmns. The absolute number of replications is the appropriate variable because the replication is the random trial which is key to determining the probability of a beneficial mutation occurring.

I don't need to repeat it, evolution isn't predictable, you can't "predict" the probability of a beneficial mutation occurring due to random variables such as mutation rates, environmental factors and so on.

> >
> >
> >
> >
> > > Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > > .
> > > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
> > >
> >
> > You call this "conceptual conflict," I call it "attention to different
> > contributions to overall relative frequency."
> The relative frequency of different variants in a population has no bearing on the mathematics of rmns.
> >
> > This applies to your next paragraph also.
> >
> > > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> > > .
> >
> > > This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur.
> >
> > Of course. But if one variant has a big advantage over another [like in
> > your drug cocktail scenario, where the resistant variants pretty much
> > take over the huge environment of the host from the nonresistant ones]
> > then there will be replications galore.
> The drug sensitive variants don't disappear from the world, they just can't grow in the environment with the drug. I can tell you from direct personal experience that there are still plenty of drug sensitive bacteria around.

They're being out-competed by antibiotic resistant bacteria, that seems to me like they're "disappearing". Antibiotic resistance is a beneficial mutation (to the bacteria, go figure) which selection acts upon, the antibiotic resistant bacteria then out-compete non-antibiotic resistant bacteria because they can better withstand certain antibiotic drugs, which enables them to infect victims better than non-antibiotic resistant pathogens.

> >
> > In his typically hasty reply to my first post to this thread, Harshman
> > claimed "that Alan thinks his drug cocktail scenario is analogous to
> > every possible instance of natural selection." Yet here, you don't
> > seem to give it the weight it deserves!
> If you read my publications, I explain that the reason HIV combination therapy works is because of the mathematics I presented. This quote is from my first publication: "Any disruption of this beneficial mutation/amplification of beneficial mutation cycle will stifle the

Your mathematics is bullshit. Evolution is unpredictable, there are too many factors to account for when trying to "predict" evolution. Mathematics just doesn't cut it in biology the way it does in physics.

> evolutionary process. Introducing a second selection pressure, which targets a different genetic locus than the first selection pressure along with the first selection pressure, forces the population to attempt to take two evolutionary trajectories simultaneously and will disrupt the evolutionary cycle. It is this principle, which has led to the success of combination therapy for the treatment of HIV [4]"
> >
> >
> >
> > > Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region.
> >
> > Here too you are downplaying the advantages of the resistant variant.
> Au contraire, I'm explaining how resistant variants come about. Once you understand this, you can take the proper steps to prevent it.

No, you're not. You're about as effective at explaining how resistant variants come about as religious apologists are when explainng why their particular theology is true, that is, not at all.

> >
> >
> > > This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
> > > .
> >
> > The two can be very strongly linked, see above.
> The linkage occurs when one lineage is competing for the same resources of the environment as other lineages. The empirical example of this is the Lenski experiment. His bacteria are not only evolving against starvation pressure but are competing for the limiting resource of the environment. But the mathematics of rmns is not changed in this circumstance, it only slows the ability of the slightly more fit variant to accumulate the replications necessary for the next beneficial mutation.
> >
> >
> > > The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step.

CITE? Can you link the paper in which these supposed equations took place? While it is true that sexual selection has a role in evolution, you're ignoring other, vital factors as well, such as genetic drift or environmental factors such as climate change. Those also have an equal, if not greater effect on evolution than sexual selection, especially since not all organisms reproduce by sexual reproduction. I also know that evolution is for the most part unpredictable, so math really doesn't apply to evolution the way it does to theoretical physics.

> >
> > But the reproductive success is intimately tied in with overall
> > frequency.
> Only when the environment puts limits on resources as well as applying other stressors to the different lineages. The mathematics of rmns remains the same. The ability of a lineage to accumulate beneficial mutations against a selection pressure is dependent only on its reproductive success on each evolutionary step. Other competitors in the environment limit their reproductive success and slow the evolutionary process. This is why Lenski's experiment takes more than 1000 generations for each beneficial mutation.

There's no argument here; why pretend this is revolutionary when it's been established fact for a long time? Lenksi's experiment doesn't show any "cycle", it only shows the amount of generations observed for the evolutionary change to take place.

> >
> >
> > > Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.

"Evolutionary step?" Don't you mean "generation"? Of course, this is assuming that the shit you are smearing across the wall has any meaning. It would be a lot simpler to just write "those variations that don't have sufficient reproductive success, that is, they fail to pass on their genes to the next generation, are selected against. Those variations which are able to successfully spread their genes are less likely to go extinct."

[John Harshman]

Does that happen with "pretty much every population"? If not, what keeps
if from happening?

>>>>> Those using this argument don't realize that most
>>>>> "natural" selection pressures generally target multiple genetic loci
>>>>> simultaneously. Thermal stress affects the function of virtually all
>>>>> enzymes, starvation affects every metabolic pathway which requires
>>>>> energy to function, dehydration same thing.
>>>>
>>>> To the extent that no mutation at any single locus is beneficial?

>>> Some selection pressures are like that. Now antibiotics and antiviral
>>> agents have the special property that they (generally) only target a
>>> single genetic locus.

>> Those two sentences don't relate to each other, though I know you think
>> they do. Your viral scenario relies on the assumption that in a
>> multidrug environment, resistance to a single drug confers no benefit.

> They do if you understand how rmns works which you don't or are unwilling to.

If they relate, please explain.

>>>>>> There is a lot of natural selection implicit in this scenario, and Harshman
>>>>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
>>>>>> that your knowledge of natural selection is close to nonexistent,
>>>>>> and a concrete example like this could lay these insinuations to rest.
>>>>
>>>>> I agree with you, the suppression of the evolution of HIV by three
>>>>> drug therapy is the quintessential example of why multiple selection
>>>>> pressures inhibit the rmns phenomenon. HIV has everything going for it
>>>>> when it comes to rmns. It has huge populations, high mutation rates, it
>>>>> does recombination but the multiplication rule of probabilities stops it.
>>>>
>>>> This is a peripheral issue, but I have my doubts about the frequency of
>>>> recombination. Does HIV recombine genomes infesting different cells or
>>>> only the same cell? If the latter, then the effective population size
>>>> for recombination events is much smaller than you imply.
>>
>>> John, I wrote a paper explaining when random recombination has a
>>> reasonable probability of giving a multi-drug resistant offspring. If
>>> you took a course in basic probability theory, you might understand
>>> it.

>> That doesn't answer my question. I'm assuming you don't actually know
>> the answer. Correct?

> John, my paper gives the upper limits of the probability of random
> recombination giving a multi-drug resistant offspring. If you want to
> think of reasons to lower that probability, be my guest.

A simple "I don't know" would suffice. There's no shame in not knowing.

>>>>>> Don't expect anyone else to do it for you. I am too busy with Harshman
>>>>>> and Martinez's chicanery in other issues, and most of the others are
>>>>>> not going to lift a finger to help you: Martinez keeps trying to nail
>>>>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
>>>>>> and almost everyone else is too interested in seeing another creationist
>>>>>> --yourself-- go down in flames to be objective about what is really going on.
>>>>
>>>>> The only thing going down in flames here is the theory of evolution.
>>>>> In exchange, people will learn how to reduce or prevent multi-drug
>>>>> resistant microbes, multi-herbicide resistant weeds, and produce more
>>>>> durable cancer treatments. Not a bad exchange for a theory atheists
>>>>> use to argue God doesn't exist.
>>>> And so we see Alan's true motivation, only thinly disguised, if at all.
>>>> He's closer to Ray that either of them supposes.
>>
>>> This is a bad time for atheists who understand basic probability theory.
>>
>> It's a bad time for creationists who understand anything about geology,
>> paleontology, or biology. Now I know you aren't interested in those
>> subjects and know nothing about them, so it's not a bad time for you.
>> Ignorance is strength.

> I don't mind seeing an academic mud wrestling match. I think the soft
> sciences like biology are at a big disadvantage though. I can see
> this disadvantage when I debate with you.

I would also like to point out that it isn't atheists who are your
opponents here, it's biologists, geologists, paleontologists, even
physicists, and there are many theists among them.

[Alan Kleinman MD PhD]

Google this "how many species have gone extinct"

>
> >>>>> Those using this argument don't realize that most
> >>>>> "natural" selection pressures generally target multiple genetic loci
> >>>>> simultaneously. Thermal stress affects the function of virtually all
> >>>>> enzymes, starvation affects every metabolic pathway which requires
> >>>>> energy to function, dehydration same thing.
> >>>>
> >>>> To the extent that no mutation at any single locus is beneficial?
>
> >>> Some selection pressures are like that. Now antibiotics and antiviral
> >>> agents have the special property that they (generally) only target a
> >>> single genetic locus.
>
> >> Those two sentences don't relate to each other, though I know you think
> >> they do. Your viral scenario relies on the assumption that in a
> >> multidrug environment, resistance to a single drug confers no benefit.
>
> > They do if you understand how rmns works which you don't or are unwilling to.
>
> If they relate, please explain.

Selection pressure which targets only a single gene and is used singly can have single mutations which give improved reproductive fitness. Think about what happens when a single selection pressure targets multiple genetic loci simultaneously.

>
> >>>>>> There is a lot of natural selection implicit in this scenario, and Harshman
> >>>>>> could easily see it if he weren't so stubbornly lazy. He keeps insinuating
> >>>>>> that your knowledge of natural selection is close to nonexistent,
> >>>>>> and a concrete example like this could lay these insinuations to rest.
> >>>>
> >>>>> I agree with you, the suppression of the evolution of HIV by three
> >>>>> drug therapy is the quintessential example of why multiple selection
> >>>>> pressures inhibit the rmns phenomenon. HIV has everything going for it
> >>>>> when it comes to rmns. It has huge populations, high mutation rates, it
> >>>>> does recombination but the multiplication rule of probabilities stops it.
> >>>>
> >>>> This is a peripheral issue, but I have my doubts about the frequency of
> >>>> recombination. Does HIV recombine genomes infesting different cells or
> >>>> only the same cell? If the latter, then the effective population size
> >>>> for recombination events is much smaller than you imply.
> >>
> >>> John, I wrote a paper explaining when random recombination has a
> >>> reasonable probability of giving a multi-drug resistant offspring. If
> >>> you took a course in basic probability theory, you might understand
> >>> it.
>
> >> That doesn't answer my question. I'm assuming you don't actually know
> >> the answer. Correct?
>
> > John, my paper gives the upper limits of the probability of random
> > recombination giving a multi-drug resistant offspring. If you want to
> > think of reasons to lower that probability, be my guest.
> A simple "I don't know" would suffice. There's no shame in not knowing.

I can think of lots of reasons why recombination is not helping HIV evolve resistance to combination therapy. The crossing over point may not allow the correct genes to give resistance to be passed to the descendant. My model assumes that it does and still shows why this is a low probability occurrence. John, it's ok to say that you don't understand probability theory, everyone knows it anyway.

>
> >>>>>> Don't expect anyone else to do it for you. I am too busy with Harshman
> >>>>>> and Martinez's chicanery in other issues, and most of the others are
> >>>>>> not going to lift a finger to help you: Martinez keeps trying to nail
> >>>>>> you as an "Atheist," Steady Eddie is too addicted to tweet-length posts,
> >>>>>> and almost everyone else is too interested in seeing another creationist
> >>>>>> --yourself-- go down in flames to be objective about what is really going on.
> >>>>
> >>>>> The only thing going down in flames here is the theory of evolution.
> >>>>> In exchange, people will learn how to reduce or prevent multi-drug
> >>>>> resistant microbes, multi-herbicide resistant weeds, and produce more
> >>>>> durable cancer treatments. Not a bad exchange for a theory atheists
> >>>>> use to argue God doesn't exist.
> >>>> And so we see Alan's true motivation, only thinly disguised, if at all.
> >>>> He's closer to Ray that either of them supposes.
> >>
> >>> This is a bad time for atheists who understand basic probability theory.
> >>
> >> It's a bad time for creationists who understand anything about geology,
> >> paleontology, or biology. Now I know you aren't interested in those
> >> subjects and know nothing about them, so it's not a bad time for you.
> >> Ignorance is strength.
>
> > I don't mind seeing an academic mud wrestling match. I think the soft
> > sciences like biology are at a big disadvantage though. I can see
> > this disadvantage when I debate with you.
>
> I would also like to point out that it isn't atheists who are your
> opponents here, it's biologists, geologists, paleontologists, even
> physicists, and there are many theists among them.

Which ones understand probability theory because those are the ones who will understand how rmns works.

[r3p...@gmail.com]

Alan, I watched the Kishony video "EXTRA MINUTES - SUPERBBUGS (Harvard Experiment explained)" all 3 minutes and 39 seconds of it. Hard to understand his broken English, but I get his point, nothing really new here.

Ray

[Alan Kleinman MD PhD]

On Tuesday, September 12, 2017 at 6:40:05 PM UTC-7, r3p...@gmail.com wrote:
> Alan, I watched the Kishony video "EXTRA MINUTES - SUPERBBUGS (Harvard Experiment explained)" all 3 minutes and 39 seconds of it. Hard to understand his broken English, but I get his point, nothing really new here.
>
> Ray

If you look real carefully you can see is bacteria playing dice.

[r3p...@gmail.com]

There isn't any mathematics or genetics in the Origin. Yet Darwin's successors in science retained the identity that he gave to his theory----"natural selection"----because they confirmed his conceptual foundation and framework to be correct. This occurred in the 1930s and 1940s during the rise of the genetical theory; that is, 70-80 years after the theory was originally published in 1859. If Darwin had not been correct then the theory would require a different identity to reflect the fact that his original formulation did not survive, but that's definitely not the case.

In case the main point is somehow missed: It took biology 70-80 years to ratify Darwin's theory of natural-cumulative selection published originally without any supporting mathematics. That's seven to eight decades of careful scrutiny THEN universal scientific acceptance of natural-cumulative selection occurred when genetics and mathematics had become integrated into the theory.

See Gould 2002 "The Structure Of Evolutionary Theory" to find out what happened during those 80 years. Gould will also tell you that the major restructuring of the theory that has occurred cannot be described as replacement, but an extension of Darwin (reference available upon request).

Your mathematics and the few who understand has very little chance of negating the observations, inferences, extrapolations, and internal logic that supports Darwin's theory of natural-cumulative selection, just saying.

I hate people who say the concept of natural selection exists in nature, but not the concept of cumulative selection. Sound logic says its an all or nothing proposition because the latter is based on the former.

Ray (anti-selectionist)

[r3p...@gmail.com]

Did Kishony say that?

I don't remember him using that analogy.

Ray

[John Harshman]

You haven't thought that through. If your claims are correct, just every
species now living should be in the process of going extinct, and that
fairly quickly. In contrast, the historical rate of extinction is quite
low, and most species last for millions of years.

But at least you have finally shown some interest in paleontology,
though you perhaps you didn't realize you were talking about paleontology.

>>>>>>> Those using this argument don't realize that most
>>>>>>> "natural" selection pressures generally target multiple genetic loci
>>>>>>> simultaneously. Thermal stress affects the function of virtually all
>>>>>>> enzymes, starvation affects every metabolic pathway which requires
>>>>>>> energy to function, dehydration same thing.
>>>>>>
>>>>>> To the extent that no mutation at any single locus is beneficial?
>>
>>>>> Some selection pressures are like that. Now antibiotics and antiviral
>>>>> agents have the special property that they (generally) only target a
>>>>> single genetic locus.
>>
>>>> Those two sentences don't relate to each other, though I know you think
>>>> they do. Your viral scenario relies on the assumption that in a
>>>> multidrug environment, resistance to a single drug confers no benefit.
>>
>>> They do if you understand how rmns works which you don't or are unwilling to.
>>
>> If they relate, please explain.

> Selection pressure which targets only a single gene and is used
> singly can have single mutations which give improved reproductive
> fitness. Think about what happens when a single selection pressure
> targets multiple genetic loci simultaneously.

What happens? I would suppose that a good mutation in any of the
targeted loci ought to improve fitness.

HIV doesn't cross over. The recombination isn't like that. The actual
reason recombination doesn't do much for HIV is that it's a rare event
that only happens when different strains infect the same cell. Here, 10
seconds with google turned this up. You're welcome.

D. S. Burke. Recombination in HIV: an important viral evolutionary
strategy. Emerg. Infect Dis. 1997 Jul-Sep; 3(3): 253–259.

Abstract: Human immunodeficiency virus (HIV) is a diploid virus: each
virion carries two complete RNA genomic strands. Homologous
recombination can occur when a cell is coinfected with two different but
related strains. Naturally occurring recombinant HIV strains have been
found in infected patients in regions of the world where multiple
genotypic variants cocirculate. One recombinant HIV strain has spread
rapidly to millions of persons in Southeast Asia. Recombination is a
mechanism whereby high level and multidrug-resistant strains may be
generated in individual treated patients. Recombination also poses
theoretical problems for the development of a safe HIV vaccine. Certain
features of HIV replication, such as syncytium formation and
transactivation, may be best understood as components of a sexual
reproductive cycle. Recombination may be an important HIV evolutionary
strategy.

So, cool. It's diploid, which I didn't know, and recombination occurs
only when cells are infected by different strains, which I suspected.
You would think that an expert such as yourself would already have known
this.

Only the biologists are your opponents regarding "rmns". The others
don't like different aspects of your creationism. How old do you think
the earth is?

[Peter Nyikos]

On Tuesday, September 12, 2017 at 1:05:05 AM UTC-4, John Harshman wrote:
> On 9/11/17 6:44 PM, Alan Kleinman MD PhD wrote:
> > On Monday, September 11, 2017 at 6:20:05 PM UTC-7, John Harshman wrote:

> >> On 9/11/17 5:26 PM, Alan Kleinman MD PhD wrote:

> >>> Too bad, I thought you were going to tell us how generations don't
> >>> have any effect on fitness. What if fitness is zero? How many
> >>> generations for that variant?

> >> You're just exposing your ignorance of selection. What you're saying
> >> there isn't even wrong, just nonsensical assembly of words.

His opening sentence was intelligible, and you could have met it
head on.

His two questions make perfect sense, and they are a cinch to
answer. Why must you keep playing dominance games?

> > I get your argument, no generations, no replicators, no replicators, no selection. Your argument makes great guacamole.
>
> Whenever you can't deal with an issue, you deflect, avoid, make a joke.

What's to deflect and avoid about your non sequitur about what he
last said?

You made no argument, just an *ipse dixit* followed by a literally
false claim about what Alan had written.

Before you go jumping to conclusions, recall that I have often told you that
I suffer fools gladly, knaves with difficulty or not at all.

You have been a knave from almost the first day I returned to talk.origins.
That much is blindingly obvious in hindsight.

I haven't seen enough of Alan to be sure whether he is a fool or a knave
(or maybe neither -- you aren't shedding any light on that here).

> >>> John, the population size does not remain constant, reproductive
> >>> fitness determines what that population size is. That's how natural
> >>> selection works for rmns.
> >>
> >> Then why did you model a constant population size? And yes, natural
> >> selection can happen in a population of constant size; why would you
> >> think otherwise?
>
> > Apparently, I did it to confuse you. And it has worked way beyond
> > what I had hoped. Any chance I can get a pillow filled with
> > Caudipteryx feathers?
>
> Deflection, avoidance, joke.

...pot...kettle. You do the same from time to time. You came up with
something that easily matches or outdoes what Alan did here, very soon
after I returned in 2010 after almost a decade of absence.

And, to sweeten the pot, you did it shortly after you falsely accused
me of playing a dominance game. And if you deny any of this, you
will see how your lame joke and your earlier false accusation of over
seven years ago were very well connected, with a gradual evolution
from one to the other.


But Alan isn't off the hook here either. I'll be watching the
sequel of this particular deflection of his closely.

Peter Nyikos

[Peter Nyikos]

On Tuesday, September 12, 2017 at 7:10:05 PM UTC-4, Alan Kleinman MD PhD wrote:
> On Tuesday, September 12, 2017 at 2:50:05 PM UTC-7, Peter Nyikos wrote:
> > On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > > .
> > > “By the way, Alan, Harshman has been needling me to comment on 
> > > how natural selection enters into those peer-reviewed articles of 
> > > yours. Would you like to give me a synopsis so I can accurately 
> > > judge it for myself?  ”
> >
> > What I meant was "a synopsis of those two papers." You obviously took
> > it differently.
> My first paper address the mathematics which govern rmns for a single selection pressure targeting a single gene. The derivation of the equations was based on empirical example published by Weinreich and others: "Darwinian evolution can follow only very few mutational paths to fitter proteins."

This is suspicious on the face of it. How can you know the number
of paths when you haven't a good idea of what kinds of proteins --
out of, say, 4^200 -- are going to be "fitter"?

However, this is beside the point of how natural selection enters
into your paper. Good thing you gave a little hint of that below.

> My other paper on rmns addressed the mathematics of multiple simultaneous selection pressures

Funny... Harshman, who is usually so quick on the draw, has
not replied to this e-mail to "point out" that your second paper
does not deal with selection pressures at all.

And yet, this post of yours was made FIFTEEN HOURS ago, during
which he did a rapid-fire back-and-forth with you elsewhere
on this thread.

Will wonders never cease? :-)

Do you suppose John thinks that Oxyaena's quick-on-the-draw
reply to this post of yours lets him off the hook? [Despite the
fact that Oxyaena wrote about completely different things!]

> > > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.
> >
> > Why "leads to"? Some of the other variants might hang on to their
> > relative frequency or even improve it while the formerly populous ones might
> > undergo a severe decline.
> Agreed, that's why relative frequency is not the appropriate variable to model rmns. The absolute number of replications is the appropriate variable because the replication is the random trial which is key to determining the probability of a beneficial mutation occurring.

True. But here is where Oxyaena came in, by pointing out that we don't
know enough yet to determine the probability. What he neglected to note
is that it can be done IN PRINCIPLE, and that is what is relevant to
your dispute with Harsman, the one which should have elicited a
"pointing out" from him 14 hours ago, were he up to speed.

> > > Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > > .
> > > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
> > >
> >
> > You call this "conceptual conflict," I call it "attention to different
> > contributions to overall relative frequency."

> The relative frequency of different variants in a population has no bearing on the mathematics of rmns.

But there is plenty of relevance in the opposite direction, and
so there is no "conceptual conflict".

> > This applies to your next paragraph also.
> >
> > > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> > > .
> >
> > > This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur.
> >
> > Of course. But if one variant has a big advantage over another [like in
> > your drug cocktail scenario, where the resistant variants pretty much
> > take over the huge environment of the host from the nonresistant ones]
> > then there will be replications galore.

> The drug sensitive variants don't disappear from the world, they just can't grow in the environment with the drug.

What's relevant here is the astronomical numbers of the drug resistant
ones in the patients receiving the drug. Penicillin is a well known
example where patients all around the world have received it, many times.

>I can tell you from direct personal experience that there are still plenty of drug sensitive bacteria around.


> > In his typically hasty reply to my first post to this thread, Harshman
> > claimed "that Alan thinks his drug cocktail scenario is analogous to
> > every possible instance of natural selection." Yet here, you don't
> > seem to give it the weight it deserves!

> If you read my publications, I explain that the reason HIV combination therapy works is because of the mathematics I presented. This quote is from my first publication: "Any disruption of this beneficial mutation/amplification of beneficial mutation cycle will stifle the
> evolutionary process. Introducing a second selection pressure, which targets a different genetic locus than the first selection pressure along with the first selection pressure, forces the population to attempt to take two evolutionary trajectories simultaneously and will disrupt the evolutionary cycle. It is this principle, which has led to the success of combination therapy for the treatment of HIV [4]"

Well, you are certainly talking about selection pressures here, which
leads me to suspect that Harshman has never read your two papers and
has been talking off the top of his head all this time about there
being nothing about natural selection mentioned in them.

The ball is definitely in his court here.

Peter Nyikos
Professor, Department of Math. -- standard disclaimer --
U. of South Carolina at Columbia
http://www.math.sc.edu/~nyikos/

[John Harshman]

On 9/13/17 5:31 AM, Peter Nyikos wrote:
> On Tuesday, September 12, 2017 at 1:05:05 AM UTC-4, John Harshman wrote:
>> On 9/11/17 6:44 PM, Alan Kleinman MD PhD wrote:
>>> On Monday, September 11, 2017 at 6:20:05 PM UTC-7, John Harshman wrote:
>>>> On 9/11/17 5:26 PM, Alan Kleinman MD PhD wrote:
>
>>>>> Too bad, I thought you were going to tell us how generations don't
>>>>> have any effect on fitness. What if fitness is zero? How many
>>>>> generations for that variant?
>
>>>> You're just exposing your ignorance of selection. What you're saying
>>>> there isn't even wrong, just nonsensical assembly of words.
>
> His opening sentence was intelligible, and you could have met it
> head on.
>
> His two questions make perfect sense, and they are a cinch to
> answer. Why must you keep playing dominance games?

We disagree on whether he made sense.

>>> I get your argument, no generations, no replicators, no replicators, no selection. Your argument makes great guacamole.
>>
>> Whenever you can't deal with an issue, you deflect, avoid, make a joke.
>
> What's to deflect and avoid about your non sequitur about what he
> last said?

You have eliminated the context that would have told you.

Will you have anything to contribute beyond telling me what a bad person
I am?

[John Harshman]

On 9/13/17 8:22 AM, Peter Nyikos wrote:
>
> Well, you are certainly talking about selection pressures here, which
> leads me to suspect that Harshman has never read your two papers and
> has been talking off the top of his head all this time about there
> being nothing about natural selection mentioned in them.
>
> The ball is definitely in his court here.

I didn't say that there was nothing about natural selection mentioned. I
said there was no modeling of natural selection in his math. And in fact
he postulates scenarios in which there is no selection, that is in which
no existing genotype is more fit than any other. In his scenarios, an
increase in fitness only arrives, and selection only happens, after the
last in a series of mutations. And his models are not concerned with
what happens after the last mutation.

Now all you have to do is go read his papers and find where in his math
the natural selection terms are. I would be interested, since I couldn't
find any. Is there perhaps a fitness parameter I missed? Something about
rate of change in population size of particular variants?

[Steady Eddie]

On Sunday, 10 September 2017 22:55:05 UTC-6, Alan Kleinman MD PhD wrote:
> Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> .
> “By the way, Alan, Harshman has been needling me to comment on 
> how natural selection enters into those peer-reviewed articles of 
> yours. Would you like to give me a synopsis so I can accurately 
> judge it for myself?  ”

> .

> Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection

> This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.

> .
> However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)

> .
> Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)?

Aren't both options saying the same thing? All things being equal, on average, reproductive success causes an increase in the relative frequency of a variant, and increased fitness causes an increase in reproductive success, doesn't it?

Really, imo, natural selection is not a separate "force", or influence, acting on populations; it is a trivial truism:
Variants that are more fit for their environment will be "selected" to survive to reproduce more.

This goes without saying, and Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.

The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> .

[John Harshman]

On 9/13/17 9:14 AM, Steady Eddie wrote:

> Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.

Excellent. Perhaps you can tell me where in Alan's calculations he
considers natural selection. Please be specific: what parameter or
parameters in his equations represent fitness?

[Steady Eddie]

Fitness is an intrinsic part of why any mutation would "amplify" in a population.
In any given model, one can assume that we are talking about mutations that increase fitness. It goes without saying.

[Steady Eddie]

On Wednesday, 13 September 2017 10:25:05 UTC-6, John Harshman wrote:

Actually, Alan is giving you everything but the kitchen sink in his calculations; he is examining situations where the selection pressures are lethal, so in his analyses, he is granting that NS is working at virtually 100% efficiency. In most cases of adaptation, NS has more modest effects on population, probably virtually 0.

So, Alan's research is showing that, even GRANTING about a 100% 'selection' rate for a beneficial mutation, it is the rate of ORIGINATION of beneficial mutations that is the determining factor for the probabilistic analysis of the capacity of Darwinism to produce novel structures, processes, and systems.

[John Harshman]

No, it doesn't go without saying. It has to be in the math.
"Amplification" isn't in the math either. But I'm willing to be wrong.
How does he represent fitness, amplification, or anything else that has
to do with selection?

[John Harshman]

On 9/13/17 10:31 AM, Steady Eddie wrote:
> On Wednesday, 13 September 2017 10:25:05 UTC-6, John Harshman wrote:
>> On 9/13/17 9:14 AM, Steady Eddie wrote:
>>
>>> Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.
>>
>> Excellent. Perhaps you can tell me where in Alan's calculations he
>> considers natural selection. Please be specific: what parameter or
>> parameters in his equations represent fitness?
>
> Actually, Alan is giving you everything but the kitchen sink in his
> calculations; he is examining situations where the selection
> pressures are lethal, so in his analyses, he is granting that NS is
> working at virtually 100% efficiency. In most cases of adaptation, NS
> has more modest effects on population, probably virtually 0.

I see that you don't understand selection either. Selection involves
genotypes of different fitnesses. No differences, no selection.

> So, Alan's research is showing that, even GRANTING about a 100%
> 'selection' rate for a beneficial mutation, it is the rate of
> ORIGINATION of beneficial mutations that is the determining factor
> for the probabilistic analysis of the capacity of Darwinism to
> produce novel structures, processes, and systems.

Good try, but no. His math features only mutation, not selection.

[Alan Kleinman MD PhD]

I came across an interesting paper by someone who did a mathematical analysis of Burke's work.
https://www.ncbi.nlm.nih.gov/pubmed/25645658
You might learn something about random recombination if you read this paper.

Again with changing the subject. Are you tired of talking about rmns?

[Alan Kleinman MD PhD]

On Wednesday, September 13, 2017 at 8:25:09 AM UTC-7, Peter Nyikos wrote:
> On Tuesday, September 12, 2017 at 7:10:05 PM UTC-4, Alan Kleinman MD PhD wrote:
> > On Tuesday, September 12, 2017 at 2:50:05 PM UTC-7, Peter Nyikos wrote:
> > > On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > > > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > > > .
> > > > “By the way, Alan, Harshman has been needling me to comment on 
> > > > how natural selection enters into those peer-reviewed articles of 
> > > > yours. Would you like to give me a synopsis so I can accurately 
> > > > judge it for myself?  ”
> > >
> > > What I meant was "a synopsis of those two papers." You obviously took
> > > it differently.
> > My first paper address the mathematics which govern rmns for a single selection pressure targeting a single gene. The derivation of the equations was based on empirical example published by Weinreich and others: "Darwinian evolution can follow only very few mutational paths to fitter proteins."
>
> This is suspicious on the face of it. How can you know the number
> of paths when you haven't a good idea of what kinds of proteins --
> out of, say, 4^200 -- are going to be "fitter"?

Weinreich measured the paths. He identified sequences of 5 mutations which give resistance to the antibiotic. Each of the successful sequences of mutations occurs with ever increasing fitness to the antibiotic selection pressure. I wrote the general probability equation that governs any of these evolutionary trajectories. Weinreich calculates 5! possible trajectories, I think he should have calculated 4^5 possible trajectories.

>
> However, this is beside the point of how natural selection enters
> into your paper. Good thing you gave a little hint of that below.
>
>
> > My other paper on rmns addressed the mathematics of multiple simultaneous selection pressures
>
> Funny... Harshman, who is usually so quick on the draw, has
> not replied to this e-mail to "point out" that your second paper
> does not deal with selection pressures at all.

I don't think John understands either of my papers and is unwilling to put in the effort to do so.

>
> And yet, this post of yours was made FIFTEEN HOURS ago, during
> which he did a rapid-fire back-and-forth with you elsewhere
> on this thread.
>
> Will wonders never cease? :-)
>
> Do you suppose John thinks that Oxyaena's quick-on-the-draw
> reply to this post of yours lets him off the hook? [Despite the
> fact that Oxyaena wrote about completely different things!]
>
> > > > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > > > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.
> > >
> > > Why "leads to"? Some of the other variants might hang on to their
> > > relative frequency or even improve it while the formerly populous ones might
> > > undergo a severe decline.
> > Agreed, that's why relative frequency is not the appropriate variable to model rmns. The absolute number of replications is the appropriate variable because the replication is the random trial which is key to determining the probability of a beneficial mutation occurring.
>
> True. But here is where Oxyaena came in, by pointing out that we don't
> know enough yet to determine the probability. What he neglected to note
> is that it can be done IN PRINCIPLE, and that is what is relevant to
> your dispute with Harsman, the one which should have elicited a
> "pointing out" from him 14 hours ago, were he up to speed.

rmns is a simple binary probability problem. Does the beneficial mutation occur or not. Identify the random trial (the replication), compute the probability for a particular mutation to occur at a particular site and then extend the computation for more than a single replication and for more than a single generation. In principle, a rudimentary probability problem which real populations obey.

>
>
> > > > Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > > > .
> > > > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
> > > >
> > >
> > > You call this "conceptual conflict," I call it "attention to different
> > > contributions to overall relative frequency."
>
> > The relative frequency of different variants in a population has no bearing on the mathematics of rmns.
>
> But there is plenty of relevance in the opposite direction, and
> so there is no "conceptual conflict".

From a mathematical point of view, relative frequency does not appear in the probability equations which govern rmns. From a physical point of view, the relative frequency can impact rmns. I've given to empirical examples which demonstrate this. The Kishony experiment doesn't force his population to compete with one another for resources of the environment, therefore amplification of beneficial mutations occurs very rapidly in his environment. The Lenski experiment which uses starvation forces his populations to compete for the resources of the environment slowing the amplification of any beneficial mutation which might occur.

>
>
> > > This applies to your next paragraph also.
> > >
> > > > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas
> > > > .
> > >
> > > > This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur.
> > >
> > > Of course. But if one variant has a big advantage over another [like in
> > > your drug cocktail scenario, where the resistant variants pretty much
> > > take over the huge environment of the host from the nonresistant ones]
> > > then there will be replications galore.
>
> > The drug sensitive variants don't disappear from the world, they just can't grow in the environment with the drug.
>
> What's relevant here is the astronomical numbers of the drug resistant
> ones in the patients receiving the drug. Penicillin is a well known
> example where patients all around the world have received it, many times.

I'm well aware of this, I have successfully treated hundreds, perhaps thousands of patients with MRSA. I have also treated many patients with penicillin sensitive staph infections as well. The relevant point here is understanding exactly how penicillin and other drug resistant bacteria are formed. Then a correct strategy for preventing this from happening can be formulated.

>
>
> >I can tell you from direct personal experience that there are still plenty of drug sensitive bacteria around.
>
>
> > > In his typically hasty reply to my first post to this thread, Harshman
> > > claimed "that Alan thinks his drug cocktail scenario is analogous to
> > > every possible instance of natural selection." Yet here, you don't
> > > seem to give it the weight it deserves!
>
> > If you read my publications, I explain that the reason HIV combination therapy works is because of the mathematics I presented. This quote is from my first publication: "Any disruption of this beneficial mutation/amplification of beneficial mutation cycle will stifle the
> > evolutionary process. Introducing a second selection pressure, which targets a different genetic locus than the first selection pressure along with the first selection pressure, forces the population to attempt to take two evolutionary trajectories simultaneously and will disrupt the evolutionary cycle. It is this principle, which has led to the success of combination therapy for the treatment of HIV [4]"
>
> Well, you are certainly talking about selection pressures here, which
> leads me to suspect that Harshman has never read your two papers and
> has been talking off the top of his head all this time about there
> being nothing about natural selection mentioned in them.

John doesn't understand the basic concepts of probability theory. Even if he read my papers he couldn't understand them. The editor of the journal who published my work asked me to write a layman's abstract describing the work for those without the mathematical skills to understand them. It was published here:
http://www.statisticsviews.com/details/news/10604248/Laymans-abstract-Random-mutation-and-natural-selection-a-predictable-phenomenon.html

>
> The ball is definitely in his court here.

John hasn't gotten to the ball park yet.

[Alan Kleinman MD PhD]

You think that if the relative frequency is not in the equation that natural selection is not modeled. That's your scientific blunder. My equation uses the number of replications as the measure of reproductive fitness. If that number is high, the probability of another beneficial mutation occurring is high, if that number is low, the probability of another beneficial mutation occurring is low.

[John Harshman]

Paywalled. Did you cite that paper of Burke's? If so, why didn't you
know the answer to my question? When dealing with recombination, did you
take into account the probability that two strains would infect the same
cell and the probability that two strains within a cell would recombine?

You have never managed to talk about natural selection, but I keep
hoping. But I didn't change the subject. I just added a subject. And of
course you're the one who introduced the subject of atheists.

[John Harshman]

On 9/13/17 1:10 PM, Alan Kleinman MD PhD wrote:
> On Wednesday, September 13, 2017 at 8:50:04 AM UTC-7, John Harshman wrote:
>> On 9/13/17 8:22 AM, Peter Nyikos wrote:
>>>
>>> Well, you are certainly talking about selection pressures here, which
>>> leads me to suspect that Harshman has never read your two papers and
>>> has been talking off the top of his head all this time about there
>>> being nothing about natural selection mentioned in them.
>>>
>>> The ball is definitely in his court here.
>>
>> I didn't say that there was nothing about natural selection mentioned. I
>> said there was no modeling of natural selection in his math. And in fact
>> he postulates scenarios in which there is no selection, that is in which
>> no existing genotype is more fit than any other. In his scenarios, an
>> increase in fitness only arrives, and selection only happens, after the
>> last in a series of mutations. And his models are not concerned with
>> what happens after the last mutation.
>>
>> Now all you have to do is go read his papers and find where in his math
>> the natural selection terms are. I would be interested, since I couldn't
>> find any. Is there perhaps a fitness parameter I missed? Something about
>> rate of change in population size of particular variants?

> You think that if the relative frequency is not in the equation that
> natural selection is not modeled.

No, I never said that.

> That's your scientific blunder. My
> equation uses the number of replications as the measure of
> reproductive fitness. If that number is high, the probability of
> another beneficial mutation occurring is high, if that number is low,
> the probability of another beneficial mutation occurring is low.

The number of replications can't be a measure of reproductive fitness.
It's the expected change in number (or frequency, if the population is
static) of replications per generation that's a measure of fitness. You
do not represent this in any way. And the probability of a mutation is
not a measure of fitness either.

[Alan Kleinman MD PhD]

On Wednesday, September 13, 2017 at 9:15:05 AM UTC-7, Steady Eddie wrote:
> On Sunday, 10 September 2017 22:55:05 UTC-6, Alan Kleinman MD PhD wrote:
> > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > .
> > “By the way, Alan, Harshman has been needling me to comment on 
> > how natural selection enters into those peer-reviewed articles of 
> > yours. Would you like to give me a synopsis so I can accurately 
> > judge it for myself?  ”
> > .
> > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
> > .
> > However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
> > .
> > Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)?
>
> Aren't both options saying the same thing? All things being equal, on average, reproductive success causes an increase in the relative frequency of a variant, and increased fitness causes an increase in reproductive success, doesn't it?

No, relative frequency does not necessarily give a correct measure of reproductive success. Here are a couple of simple examples which demonstrates this. Take an environment that has only a single variant with a population of 100 in a given generation. The relative frequency of that variant is 100/100=1 and the population is 100. The next generation the population doubles to 200. The relative frequency is 200/200=1 but now the population is 200. Reproductive success based on relative frequency says no change but based on absolute population size, the population has doubled. A second example, take an environment that has two equally fit variants each with a population of 50 in a given generation. The relative frequency of each variant is 50/100=0.5. The next generation each variant doubles to 100. The relative frequencies for each variant are 100/200=0.5. Again, based on relative frequency, no increase in reproductive success but based on absolute number of each variant, each has doubled their number. rmns is dependent on absolute number of members of a particular variant. The greater the number, the more players for the next beneficial mutation.

>
> Really, imo, natural selection is not a separate "force", or influence, acting on populations; it is a trivial truism:
> Variants that are more fit for their environment will be "selected" to survive to reproduce more.

Correct. That's why if a mutation is beneficial, that new variant will have improved reproductive success in the given environment. If that happens, you have more players for the next beneficial mutation.

>
> This goes without saying, and Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.

Remember, his entire academic career is at stake and the entire philosophy of evolutionism.

[Alan Kleinman MD PhD]

Well said! And mutations that don't increase fitness will not improve reproductive fitness (the absolute number of members).

[Alan Kleinman MD PhD]

John, in my papers, I plot the probability curves as a function of replications. Replications are the measure of reproductive fitness. Variants that have low reproductive fitness are not good candidates for another beneficial mutation (they don't do enough replications to improve the probability), on the other hand, variants which have good reproductive fitness are good candidates for the next beneficial mutation, they do lots of replications.

[Alan Kleinman MD PhD]

On Wednesday, September 13, 2017 at 1:35:04 PM UTC-7, John Harshman wrote:
> On 9/13/17 1:10 PM, Alan Kleinman MD PhD wrote:
> > On Wednesday, September 13, 2017 at 8:50:04 AM UTC-7, John Harshman wrote:
> >> On 9/13/17 8:22 AM, Peter Nyikos wrote:
> >>>
> >>> Well, you are certainly talking about selection pressures here, which
> >>> leads me to suspect that Harshman has never read your two papers and
> >>> has been talking off the top of his head all this time about there
> >>> being nothing about natural selection mentioned in them.
> >>>
> >>> The ball is definitely in his court here.
> >>
> >> I didn't say that there was nothing about natural selection mentioned. I
> >> said there was no modeling of natural selection in his math. And in fact
> >> he postulates scenarios in which there is no selection, that is in which
> >> no existing genotype is more fit than any other. In his scenarios, an
> >> increase in fitness only arrives, and selection only happens, after the
> >> last in a series of mutations. And his models are not concerned with
> >> what happens after the last mutation.
> >>
> >> Now all you have to do is go read his papers and find where in his math
> >> the natural selection terms are. I would be interested, since I couldn't
> >> find any. Is there perhaps a fitness parameter I missed? Something about
> >> rate of change in population size of particular variants?
>
> > You think that if the relative frequency is not in the equation that
> > natural selection is not modeled.
>
> No, I never said that.

I did say you said that, I said that you think that.

>
> > That's your scientific blunder. My
> > equation uses the number of replications as the measure of
> > reproductive fitness. If that number is high, the probability of
> > another beneficial mutation occurring is high, if that number is low,
> > the probability of another beneficial mutation occurring is low.
> The number of replications can't be a measure of reproductive fitness.
> It's the expected change in number (or frequency, if the population is
> static) of replications per generation that's a measure of fitness. You
> do not represent this in any way. And the probability of a mutation is
> not a measure of fitness either.

You had better learn that there is more than one way to measure reproductive fitness. One way is by absolute number, the measure used in rmns, and the other way is relative frequency, the measure used in random recombination.

[John Harshman]

Makes sense, except that your math does not say anything about either
good or poor reproductive fitness. "Number of replications" is not a
measure of fitness. Fitness differences would produce a change in number
of replications from one generation to the next, and you don't deal with
that at all.

[Alan Kleinman MD PhD]

The number of replications is reproductive fitness. Variants with poor reproductive fitness will have a small number of replications. Variants with good reproductive fitness will have a large number of replications. I understand that this is not the way you have seen it defined but this is the correct way to use reproductive fitness when modeling rmns.

[John Harshman]

I presume you meant to type "didn't" rather than "did". No, I don't
think that either. However, fitness is a relative quantity and must be
compared to something. Usually, it's compared to the fitnesses of other
genotypes in the population, but I suppose that in a pinch it could be
compared to the fitnesses of other genotypes not in the population, so
if the original genotype became extinct, we could still consider the
fitnesses of mutant genotypes by comparison with it. Of course you don't
consider fitness at all, in any way, so that's all moot.

>>> That's your scientific blunder. My
>>> equation uses the number of replications as the measure of
>>> reproductive fitness. If that number is high, the probability of
>>> another beneficial mutation occurring is high, if that number is low,
>>> the probability of another beneficial mutation occurring is low.
>> The number of replications can't be a measure of reproductive fitness.

>> It's the expected change in number (or frequency, if the population is
>> static) of replications per generation that's a measure of fitness. You
>> do not represent this in any way. And the probability of a mutation is
>> not a measure of fitness either.

> You had better learn that there is more than one way to measure
> reproductive fitness. One way is by absolute number, the measure used
> in rmns, and the other way is relative frequency, the measure used in
> random recombination.

None of that is relevant to what I said. Absolute number is not a
measure of fitness. Expected change across generations in absolute
number could be, but you don't do that.

[John Harshman]

You can say that as often as you like, but repeating a claim doesn't
make it true.

> Variants with
> poor reproductive fitness will have a small number of replications.
> Variants with good reproductive fitness will have a large number of
> replications.

....per generation, and by comparison with each other. You do not
consider any of that.

> I understand that this is not the way you have seen it
> defined but this is the correct way to use reproductive fitness when
> modeling rmns.

You do not model ns, just rm. Saying "n exists" or even "n*nG exists"
doesn't model fitness. Not even if you list a great many possible values
of n or n*nG in a table somewhere.

[Alan Kleinman MD PhD]

You got what I was thinking, correct, meant to say "didn't". But I also got correctly what you think. You see fitness as compared with other genotypes, therefore measured as relative frequency. But consider the Kishony experiment. You have colonies of e coli growing on either side of his petri dish. How does the reproductive fitness of a colony on one side have any impact on the evolution of a colony on the other side? You have colonies increasing in number until one of the members in the colony gets a beneficial mutation and can now grow in a high drug concentration region.

>
> >>> That's your scientific blunder. My
> >>> equation uses the number of replications as the measure of
> >>> reproductive fitness. If that number is high, the probability of
> >>> another beneficial mutation occurring is high, if that number is low,
> >>> the probability of another beneficial mutation occurring is low.
> >> The number of replications can't be a measure of reproductive fitness.
>
> >> It's the expected change in number (or frequency, if the population is
> >> static) of replications per generation that's a measure of fitness. You
> >> do not represent this in any way. And the probability of a mutation is
> >> not a measure of fitness either.
>
> > You had better learn that there is more than one way to measure
> > reproductive fitness. One way is by absolute number, the measure used
> > in rmns, and the other way is relative frequency, the measure used in
> > random recombination.
>
> None of that is relevant to what I said. Absolute number is not a
> measure of fitness. Expected change across generations in absolute
> number could be, but you don't do that.

Not only is the absolute number a measure of reproductive fitness, it is the correct measure for rmns. Until you understand what a sample space is, you will have great difficulty understanding this concept.

[John Harshman]

That was very confused. There is no selection until the mutation that
enables growth in the new environment happens, at which point the new
strain has a selective advantage in the new environment and proceeds to
demonstrate it by growing in that environment while other strains can't.

>>>>> That's your scientific blunder. My
>>>>> equation uses the number of replications as the measure of
>>>>> reproductive fitness. If that number is high, the probability of
>>>>> another beneficial mutation occurring is high, if that number is low,
>>>>> the probability of another beneficial mutation occurring is low.
>>>> The number of replications can't be a measure of reproductive fitness.
>>
>>>> It's the expected change in number (or frequency, if the population is
>>>> static) of replications per generation that's a measure of fitness. You
>>>> do not represent this in any way. And the probability of a mutation is
>>>> not a measure of fitness either.
>>
>>> You had better learn that there is more than one way to measure
>>> reproductive fitness. One way is by absolute number, the measure used
>>> in rmns, and the other way is relative frequency, the measure used in
>>> random recombination.
>>
>> None of that is relevant to what I said. Absolute number is not a
>> measure of fitness. Expected change across generations in absolute
>> number could be, but you don't do that.

> Not only is the absolute number a measure of reproductive fitness, it
> is the correct measure for rmns. Until you understand what a sample
> space is, you will have great difficulty understanding this concept.

Until you have some clue about what "natural selection" means, you will
have great difficulty understanding why neither absolute nor relative
number is a measure of fitness.

[r3p...@gmail.com]

On Wednesday, September 13, 2017 at 11:20:05 AM UTC-7, John Harshman wrote:
> On 9/13/17 10:31 AM, Steady Eddie wrote:
> > On Wednesday, 13 September 2017 10:25:05 UTC-6, John Harshman wrote:
> >> On 9/13/17 9:14 AM, Steady Eddie wrote:
> >>
> >>> Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.
> >>
> >> Excellent. Perhaps you can tell me where in Alan's calculations he
> >> considers natural selection. Please be specific: what parameter or
> >> parameters in his equations represent fitness?
> >
> > Actually, Alan is giving you everything but the kitchen sink in his
> > calculations; he is examining situations where the selection
> > pressures are lethal, so in his analyses, he is granting that NS is
> > working at virtually 100% efficiency. In most cases of adaptation, NS
> > has more modest effects on population, probably virtually 0.
> I see that you don't understand selection either. Selection involves
> genotypes of different fitnesses. No differences, no selection.

John observes once again that neither Eddie nor Alan understands natural selection. Yet Eddie has shown himself very knowledgeable concerning topic and Alan is a PhD. And Alan has accused John of not understanding natural selection either. And John is a PhD. I understand natural selection as complete nonsense, which means in the eyes of John, Alan, and Eddie I don't understand natural selection. Moreover, Darwinists in general have always complained that opponents do not understand natural selection.

Thus, no opponent understands natural selection, despite the fact that opponents possess the capacity to understand.

Point: The God of the Bible designed the human brain to misunderstand any claim about reality that says He is not Designer and Creator.

Ray

[Alan Kleinman MD PhD]

Unless the current variant has sufficient reproductive fitness, that variant will not be in a lineage with a good probability for another beneficial mutation.

>
> >>>>> That's your scientific blunder. My
> >>>>> equation uses the number of replications as the measure of
> >>>>> reproductive fitness. If that number is high, the probability of
> >>>>> another beneficial mutation occurring is high, if that number is low,
> >>>>> the probability of another beneficial mutation occurring is low.
> >>>> The number of replications can't be a measure of reproductive fitness.
> >>
> >>>> It's the expected change in number (or frequency, if the population is
> >>>> static) of replications per generation that's a measure of fitness. You
> >>>> do not represent this in any way. And the probability of a mutation is
> >>>> not a measure of fitness either.
> >>
> >>> You had better learn that there is more than one way to measure
> >>> reproductive fitness. One way is by absolute number, the measure used
> >>> in rmns, and the other way is relative frequency, the measure used in
> >>> random recombination.
> >>
> >> None of that is relevant to what I said. Absolute number is not a
> >> measure of fitness. Expected change across generations in absolute
> >> number could be, but you don't do that.
>
> > Not only is the absolute number a measure of reproductive fitness, it
> > is the correct measure for rmns. Until you understand what a sample
> > space is, you will have great difficulty understanding this concept.
>
> Until you have some clue about what "natural selection" means, you will
> have great difficulty understanding why neither absolute nor relative
> number is a measure of fitness.

Increase in absolute number will always be a good measure of reproductive fitness, increase in relative frequency does not guarantee that the particular variant will be in a lineage with a good probability of a beneficial mutation.

[jillery]

On Wed, 13 Sep 2017 15:57:42 -0700 (PDT), r3p...@gmail.com wrote:

>On Wednesday, September 13, 2017 at 11:20:05 AM UTC-7, John Harshman wrote:
>> On 9/13/17 10:31 AM, Steady Eddie wrote:
>> > On Wednesday, 13 September 2017 10:25:05 UTC-6, John Harshman wrote:
>> >> On 9/13/17 9:14 AM, Steady Eddie wrote:
>> >>
>> >>> Harshman's protestations that you don't consider NS in your calculations betray his simple-mindedness.
>> >>
>> >> Excellent. Perhaps you can tell me where in Alan's calculations he
>> >> considers natural selection. Please be specific: what parameter or
>> >> parameters in his equations represent fitness?
>> >
>> > Actually, Alan is giving you everything but the kitchen sink in his
>> > calculations; he is examining situations where the selection
>> > pressures are lethal, so in his analyses, he is granting that NS is
>> > working at virtually 100% efficiency. In most cases of adaptation, NS
>> > has more modest effects on population, probably virtually 0.
>> I see that you don't understand selection either. Selection involves
>> genotypes of different fitnesses. No differences, no selection.
>
>John observes once again that neither Eddie nor Alan understands natural selection. Yet Eddie has shown himself very knowledgeable concerning topic

I must have missed those posts, where Steadly has shown himself very
knowledgeable about anything. Cite?

>and Alan is a PhD. And Alan has accused John of not understanding natural selection either. And John is a PhD. I understand natural selection as complete nonsense, which means in the eyes of John, Alan, and Eddie I don't understand natural selection. Moreover, Darwinists in general have always complained that opponents do not understand natural selection.
>
>Thus, no opponent understands natural selection, despite the fact that opponents possess the capacity to understand.
>
>Point: The God of the Bible designed the human brain to misunderstand any claim about reality that says He is not Designer and Creator.
>
>Ray

--
I disapprove of what you say, but I will defend to the death your right to say it.

Evelyn Beatrice Hall
Attributed to Voltaire

[Oxyaena]

On Wednesday, September 13, 2017 at 11:25:09 AM UTC-4, Peter Nyikos wrote:
> On Tuesday, September 12, 2017 at 7:10:05 PM UTC-4, Alan Kleinman MD PhD wrote:
> > On Tuesday, September 12, 2017 at 2:50:05 PM UTC-7, Peter Nyikos wrote:

> > > On Monday, September 11, 2017 at 12:55:05 AM UTC-4, Alan Kleinman MD PhD wrote:
> > > > Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
> > > > .
> > > > “By the way, Alan, Harshman has been needling me to comment on 
> > > > how natural selection enters into those peer-reviewed articles of 
> > > > yours. Would you like to give me a synopsis so I can accurately 
> > > > judge it for myself?  ”
> > >

> > > What I meant was "a synopsis of those two papers." You obviously took
> > > it differently.
> > My first paper address the mathematics which govern rmns for a single selection pressure targeting a single gene. The derivation of the equations was based on empirical example published by Weinreich and others: "Darwinian evolution can follow only very few mutational paths to fitter proteins."
>
> This is suspicious on the face of it. How can you know the number
> of paths when you haven't a good idea of what kinds of proteins --
> out of, say, 4^200 -- are going to be "fitter"?
>

> However, this is beside the point of how natural selection enters
> into your paper. Good thing you gave a little hint of that below.
>
>
> > My other paper on rmns addressed the mathematics of multiple simultaneous selection pressures
>
> Funny... Harshman, who is usually so quick on the draw, has
> not replied to this e-mail to "point out" that your second paper
> does not deal with selection pressures at all.
>

> And yet, this post of yours was made FIFTEEN HOURS ago, during
> which he did a rapid-fire back-and-forth with you elsewhere
> on this thread.
>
> Will wonders never cease? :-)
>
> Do you suppose John thinks that Oxyaena's quick-on-the-draw
> reply to this post of yours lets him off the hook? [Despite the
> fact that Oxyaena wrote about completely different things!]
>

> > > > Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection
> > > > This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population.
> > >

> > > Why "leads to"? Some of the other variants might hang on to their
> > > relative frequency or even improve it while the formerly populous ones might
> > > undergo a severe decline.
> > Agreed, that's why relative frequency is not the appropriate variable to model rmns. The absolute number of replications is the appropriate variable because the replication is the random trial which is key to determining the probability of a beneficial mutation occurring.
>
> True. But here is where Oxyaena came in, by pointing out that we don't
> know enough yet to determine the probability. What he neglected to note
> is that it can be done IN PRINCIPLE, and that is what is relevant to
> your dispute with Harsman, the one which should have elicited a
> "pointing out" from him 14 hours ago, were he up to speed.

First off, you're showing your true nature by taking a few quick shots at me without actually addressing my reply to "Dr." Kleinsman, and second off, "IN PRINCIPLE" (unnecessary caps, as usual, Peter) is different from "in actuality". Just because something can be done on paper doesn't mean it can be done in real life, more so with evolution which is inherently unpredictable, so your point is moot anyways. While I can make educated guesses about the future of life on Earth (which goes with the "IN PRINCIPLE" filler you spouted), it's nothing more than mere speculation due to the nature of evolution and the random factors that can affect it, which is why "Dr." Kleinsman's pseudomathematical bullshit isn't worth the electronic paper it's printed on.


[snip mindless drivel]

[John Harshman]

Certainly true, but you don't consider the reproductive fitnesses of
variants in your model, and beneficial mutations aren't selection.

You don't model increase in absolute number, just absolute number. There
is a difference.

[Bill Rogers]

Let's try to make this simple.

Let's say each organism with the wild type genome produces, on average, 1.1 offspring per generation, in whatever environment you are working in. And then a mutant organism occurs. And this mutation is beneficial, in this environment, to the extent that each mutant organism produces, on average, 1.2 offspring per generation. Maybe, at the generation the mutation occurs, there are 10^8 wild type organisms and only one mutant organism.

Now, you can call the fitness of the wild type 1.1 and the fitness of the mutant 1.2. Or you can call the relative fitness of the mutant with respect to the wild type 1.2/1.1 = 1.09. It doesn't matter.

If now you let multiple generations pass, in the absence of a limit to the total population size, the number of wild type organisms will increase, and the number of mutant organisms will increase a bit more. Therefore the relative frequency of the mutant will increase. That's what everybody else means by selection. The same thing will also happen if the population size is constrained by some carrying capacity of the environment, but that's not critical to the argument.

When you say "beneficial mutation" that's what it means - a mutation such that the fitness of the organism carrying that mutant is greater than the fitness of the wild type (or greater than the weighted average of the fitness of all the genotypes in the population).

Your model does not model fitness. All you do is wave your hands and say that the number of organisms with the beneficial mutant goes up. That's a slogan, not a model.

Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.

[Alan Kleinman MD PhD]

Do remember that n*nG term? Do you remember what it represents?

Sure I do, just go to the right on P(X),n*nG graph.

[Alan Kleinman MD PhD]

And your understanding of selection is incorrect when trying to describe rmns. The probability of a beneficial mutation is not dependent on the relative fitness of one variant with respects to other variants. The only requirement for rmns to work is that a lineage on a particular evolutionary trajectory must have sufficient reproductive fitness at each evolutionary step to increase its number of members to have a reasonable probability of the next beneficial mutation. Other lineages on different evolutionary trajectories do not contribute to the probabilities to the lineage on the particular evolutionary trajectory.

>
> When you say "beneficial mutation" that's what it means - a mutation such that the fitness of the organism carrying that mutant is greater than the fitness of the wild type (or greater than the weighted average of the fitness of all the genotypes in the population).

OK

>
> Your model does not model fitness. All you do is wave your hands and say that the number of organisms with the beneficial mutant goes up. That's a slogan, not a model.

All evolutionary trajectories by rmns are represented by nested binomial probability equations where the individual probability equations are linked by the multiplication rule of probabilities. The only evolutionary trajectories which give a reasonable probability of occurring are ones where each binomial probability is high. This means that the number of replications at each step must be sufficient to give that high probability.

>
> Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.

Each evolutionary step on an evolutionary trajectory is of the form:
P(X)=(1 − (1 − P(Beneficial)𝜇)^(n∗nG))
where P(X) is the probability that the beneficial mutation occurs, P(Beneficial) is the probability of all the possible mutations that can occur at the particular site it is the beneficial mutation, 𝜇 is the mutation rate, n is the subpopulation that would benefit from the particular beneficial mutation, nG is the number of generations which that subpopulation replicates. Note that n*nG is simply the total number of replications. The n*nG term could be written more generally as a double summation of the sum of replications of the particular variant in a particular generation summed over all generations. The fitness of the particular variant is given by the n*nG term. A low reproductive fit variant will have low n*nG and therefore low P(X) of taking the particular evolutionary step. A high reproductive fit variant will have high n*nG and therefore high P(X) of taking the particular evolutionary step.
.
For your example of the emergence of drug-resistance to combination therapy for malaria, where double beneficial mutations must occur to improve fitness to reproduce, the equations are very similar but require extremely large population (such as malaria can attain) to have a reasonable probability of a double beneficial mutation.

[John Harshman]

Yes, I do. And I do. Not selection.

"Go to the right" isn't part of your model. Like I said, your model
doesn't consider selection.

[Alan Kleinman MD PhD]

Poor John. For a while, I thought you would be the first evolutionary biologist to understand rmns. Now I think you will be the last.

[Bill Rogers]

As always, you are only thinking about rm, not about ns.

> >
> > When you say "beneficial mutation" that's what it means - a mutation such that the fitness of the organism carrying that mutant is greater than the fitness of the wild type (or greater than the weighted average of the fitness of all the genotypes in the population).
> OK
> >
> > Your model does not model fitness. All you do is wave your hands and say that the number of organisms with the beneficial mutant goes up. That's a slogan, not a model.
> All evolutionary trajectories by rmns are represented by nested binomial probability equations where the individual probability equations are linked by the multiplication rule of probabilities. The only evolutionary trajectories which give a reasonable probability of occurring are ones where each binomial probability is high. This means that the number of replications at each step must be sufficient to give that high probability.

Your model does not model fitness. There is no term in the model which represents the fitness.

> >
> > Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.
> Each evolutionary step on an evolutionary trajectory is of the form:
> P(X)=(1 − (1 − P(Beneficial)𝜇)^(n∗nG))
> where P(X) is the probability that the beneficial mutation occurs, P(Beneficial) is the probability of all the possible mutations that can occur at the particular site it is the beneficial mutation, 𝜇 is the mutation rate, n is the subpopulation that would benefit from the particular beneficial mutation, nG is the number of generations which that subpopulation replicates. Note that n*nG is simply the total number of replications. The n*nG term could be written more generally as a double summation of the sum of replications of the particular variant in a particular generation summed over all generations.

>The fitness of the particular variant is given by the n*nG term. A low reproductive fit variant will have low n*nG and therefore low P(X) of taking the particular evolutionary step.

No. The fitness of the particular variant is *not* given by the n*nG term. The fitness is given by the derivative of n with respect to time. The n*nG term just tells you the total number of organisms in the subpopulation that have lived until the present. Velocity is not the distance traveled.

>A high reproductive fit variant will have high n*nG and therefore high P(X) of taking the particular evolutionary step.

OK. So show us what happens over time in your model. How do the populations of the wild type and mutants change over time? But those changes have to come out of the model; you can't just wave your hands and say "now move to the right along the n*nG axis." If you can't do that, you don't have a model.

> .
> For your example of the emergence of drug-resistance to combination therapy for malaria, where double beneficial mutations must occur to improve fitness to reproduce, the equations are very similar but require extremely large population (such as malaria can attain) to have a reasonable probability of a double beneficial mutation.

You really should stop talking about that malaria paper. In the first place, it's a very simple paper about monitoring the spread of drug resistant malaria in one part of Cambodia. There are plenty of more important papers in the field if you actually care about malaria. Second, you misunderstand it. The resistance we see is not occurring because of the de novo occurrence of multiple mutations in parasites with the individual patients we treated. The strains were resistant prior to treatment (as you could have figured out if you'd get off your one trick, "multiplication rule of probabilities" pony). Drug treatment enriched for the resistant parasites, but they were there from the start.

[John Harshman]

Oh, I understand you all too well.

[solar penguin]

On 13/09/17 22:04, Alan Kleinman MD PhD wrote:

> You had better learn that there is more than one way to measure
> reproductive fitness. One way is by absolute number, the measure used
> in rmns,

Absolute number? How the f*** do you put an absolute number to
reproductive fitness?

Cats, bacteria, oak trees, etc. all fit into the environment in very
different ways. They all have very different types of fitness. How can
you make one absolute scale that measures all those different fitnesses
in the same absolute way?

[Burkhard]

Yup. Intelligence is rarely enough if not supplemented with knowledge,
and that requires investment in terms of time and work. Nothing really
surprising about this, same holds true for any discipline.

[Burkhard]

John Harshman wrote:
> On 9/12/17 3:54 PM, r3p...@gmail.com wrote:
>
>> I've also pointed out that unintelligent processes (RMNS) cannot be
>> stumping our most brilliant medical minds and their computers
>> concerning, for example, antibiotic resistance, mutating disease. As
>> a matter of fact unintelligence cannot be stumping intelligence,
>> simply impossible; therefore the causal agents must be described
>> accordingly in teleological terms, not counter factually in
>> non-teleological terms. The Bible most clearly states that God
>> unleashes disease into the biosphere. Alan has never answered these
>> points except to restate his claim that RMNS exists in his medical
>> laboratory.
> God is trying to kill us?
>

Possibly not quite, but the difference is difficult to spot without a
microscope. Ray is an adherent of the Manichean heresy, and Satan is
co-creator with God (so a God for all practical purposes)

[Burkhard]

r3p...@gmail.com wrote:
> On Monday, September 11, 2017 at 1:35:05 PM UTC-7, Alan Kleinman MD PhD wrote:
>> On Monday, September 11, 2017 at 1:10:05 PM UTC-7, John Harshman wrote:
>>> On 9/11/17 12:39 PM, Alan Kleinman MD PhD wrote:
>>>> On Monday, September 11, 2017 at 12:05:05 PM UTC-7, John Harshman wrote:
>>>>> On 9/11/17 11:14 AM, Alan Kleinman MD PhD wrote:
>>>>>> On Monday, September 11, 2017 at 10:15:05 AM UTC-7, John Harshman wrote:
>>>>>>> On 9/11/17 9:27 AM, Alan Kleinman MD PhD wrote:
>>>>>>>> On Monday, September 11, 2017 at 9:00:05 AM UTC-7, John Harshman wrote:
>>>>>>>>> On 9/11/17 7:48 AM, Alan Kleinman MD PhD wrote:
>>>>>>>>>> On Monday, September 11, 2017 at 6:35:05 AM UTC-7, John Harshman wrote:

>>>>>>>>>>> On 9/10/17 9:52 PM, Alan Kleinman MD PhD wrote:
>>>>>>>>>>>> Peter Nyikos has asked for a synopsis on my understanding of natural selection and how I use it in my calculations. I believe this topic warrants its own thread. The following quote was taken from Peter's post on September 9 in thread “Evolutionary Theorist Concedes: Evolution “Largely Avoids” Biggest Questions of Biological Origins ”
>>>>>>>>>>>> .
>>>>>>>>>>>> “By the way, Alan, Harshman has been needling me to comment on
>>>>>>>>>>>> how natural selection enters into those peer-reviewed articles of
>>>>>>>>>>>> yours. Would you like to give me a synopsis so I can accurately
>>>>>>>>>>>> judge it for myself? ”

>>>>>>>>>>>> .

>>>>>>>>>>>> Traditionally, natural selection is defined as, “Natural selection is the differential survival and reproduction of individuals due to differences in phenotype.” https://en.wikipedia.org/wiki/Natural_selection

>>>>>>>>>>>> This definition leads to mathematical models where the increase in the relative frequency of one variant leads to a decrease in relative frequency of the other variants in a population. Haldane and Kimura use this idea to develop their mathematical models of substitution and fixation of the most fit variants.
>>>>>>>>>>>> .
>>>>>>>>>>>> However, this conflicts conceptually with the meaning of reproductive fitness which is defined “Fitness (often denoted w or ω in population genetics models) is the quantitative representation of natural and sexual selection within evolutionary biology. It can be defined either with respect to a genotype or to a phenotype in a given environment. In either case, it describes individual reproductive success and is equal to the average contribution to the gene pool of the next generation that is made by individuals of the specified genotype or phenotype.” https://en.wikipedia.org/wiki/Fitness_(biology)
>>>>>>>>>>>> .

>>>>>>>>>>>> Does natural selection for rmns work by changing relative frequencies of different variants in a population or is the change in relative frequencies a consequence of the change in reproductive success (the absolute number of members in each of the particular variants)? The answer to this question is given by the empirical evidence. Consider the Kishony experiment. https://www.youtube.com/watch?v=Irnc6w_Gsas

>>>>>>>>>>>> .
>>>>>>>>>>>> This experiment demonstrates that the evolutionary process by rmns is dependent on the increase in the number of members of the particular variants (the bacterial colonies must grow (amplify)) sufficiently until there are enough replications for a beneficial mutation to occur. Once a beneficial mutation occurs, that progenitor of the new variant can grow in an area of higher drug concentration. But that new variant must again amplify (increase in number) for there to be a reasonable probability of the next beneficial mutation and for that new variant to grow in the next higher drug concentration region. This is the cycle of beneficial mutation/amplification of beneficial mutation. The probability of a beneficial mutation occurring on some member of a lineage is dependent on the number of replications (its reproductive success), not its relative frequency in the population.
>>>>>>>>>>>> .
>>>>>>>>>>>> The mathematical description of every lineage on a particular evolutionary trajectory is given by a set of nested binomial probability equations where each binomial probability equation is linked to the others by the multiplication rule of probabilities. The reproductive success of each step on the evolutionary trajectory determines the probability for the next evolutionary step. Those variants which don't have sufficient reproductive success on each evolutionary step are selected out. Those variants which have reproductive success on each evolutionary step continue the cycle.
>>>>>>>>>>>>

>>>>>>>>>>> Note that the only actual mathemetics he mentions here is about the
>>>>>>>>>>> probability of mutation and has nothing to do with selection.
>>>>>>>>>
>>>>>>>>>> John, you still don't get it. The last two sentences of the post
>>>>>>>>>> address how natural selection enters into my model. I think you are
>>>>>>>>>> stuck on the notion of differential survival as the key to natural
>>>>>>>>>> selection where for rmns, reproductive success is actually the key.
>>>>>>>>>> Perhaps more detail will help you understand.
>>>>>>>>> If natural selection enters into your model, where does it? You have a
>>>>>>>>> variable for population size, and you have a variable for number of
>>>>>>>>> generations. But where's your variable representing change in population
>>>>>>>>> size over time (what you call "amplification")?
>>>>>>>
>>>>>>>> John, debating with you is like debating with Ray. You don't
>>>>>>>> understand the basics of probability theory and you are unwilling to
>>>>>>>> learn these basic principles. For those who do understand the basic
>>>>>>>> principles of probability theory, the term n*nG is simply the number
>>>>>>>> of replications and the number of replications is the measure of
>>>>>>>> reproductive fitness.
>>>>>>>
>>>>>>> No, n*nG is the number of replications given a constant number of
>>>>>>> replications per generation, and it's the change in number of
>>>>>>> replications per generation that's a measure of fitness. You don't model
>>>>>>> that change at all.
>>>>>
>>>>>> So a population of e6 can only replicate for 100 generations, and not 1000 generations?
>>>>>
>>>>> Of course it could. But that doesn't model fitness.
>>>> Really? So more generations of replication does not contribute more to the gene pool?
>>>
>>> Really? So avocados aren't green and bumpy?
>> Some avocados are black, but more generations of replications do contribute more to the gene pool.
>>>
>>>>>>>> It is the total number of replications which
>>>>>>>> determines the probability of the next beneficial mutation occurring.
>>>>>>>> My model does not address the rate at which these mutations occur,
>>>>>>>> only the number of replications necessary for there to be a
>>>>>>>> reasonable probability of a beneficial mutation occurring.
>>>>>>>
>>>>>>> How can you consider that probability without knowing the rate? Or
>>>>>>> perhaps you are using your personal definition of "rate" here?
>>>>>
>>>>>> John, what is the difference in the sample space if you roll 5 dice
>>>>>> 10 times in 2 minutes and rolling 5 dice 10 times in 10 minutes?
>>>>> Ah, I see that you are using your personal definition of "rate". A
>>>>> mutation rate is generally considered to be something like "number of
>>>>> expected mutations per site per generation".
>>>> Oh well, another missed point by someone who hasn't taken an introductory course in probability theory.
>>>
>>> Have you ever taken an introductory course in evolutionary biology?
>> It wouldn't help in understanding how rmns works. If it would help, you wouldn't be so confused on the subject despite all the courses you've taken in evolutionary biology. An introductory course in probability theory would help you much more if you want to understand stochastic processes like rmns.
>>>
>>>>>>>> For any
>>>>>>>> readers of this thread, perhaps you can find a way to explain this to
>>>>>>>> John but I doubt he will understand any explanation until he takes
>>>>>>>> the time to learn introductory probability theory.
>>>>>>> I'd be interesting in anyone explaining that to me.
>>>>>
>>>>>> I certainly am having a hard time doing it with someone who hasn't had an introductory course in probability theory.
>>>>>
>>>>> Typical response: having nothing you say, you attempt to denigrate the
>>>>> person you're talking to.
>>>> John, if you want me to stop responding to you like this, take an introductory course in probability theory.
>>>
>>> I don't mind. I just point out that it's a non-response that avoids
>>> engaging with the subject.
>> John, debating with you is like debating with Ray. The challenge I have with you is trying to find a way of explaining this phenomenon when you refuse to try to understand the mathematical principles which govern it. But I don't mind either. It makes me try to find other ways to explain it.
>>
>
> You've said this twice now, using my name inaccurately. So I have the right to barge in and remind the group of what I have said against your claims:
>
> Alan claims that he has seen RMNS in his medical laboratory. Yet I've pointed out that the accepted **conceptual** model of natural selection in the wild consists of multiple inferences which then result in the occurrence of micro-evolution over generational time.


And given how short a generation of bacteria is, what Alan (or any
medical researcher) observes is indeed evolution over generational time.

>So Alan has not seen natural selection in his medical laboratory; rather, he has seen "natural selection."

>
> I've also pointed out that unintelligent processes (RMNS) cannot be stumping our most brilliant medical minds and their computers concerning, for example, antibiotic resistance, mutating disease. As a matter of fact unintelligence cannot be stumping intelligence, simply impossible;

That is simply a glaring non-sequitur. Why on earth should the mere fact
that we are intelligent mean that we can solve all problems? Your
theology is pretty heretical as it is, but this argument always assumes
humans are omniscient, and is particularly weird even for you.

Problem solving requires more than intelligence, it requires
information, or knowledge. Knowledge is gained through a slow,
cumulative (if we are lucky and don't have an external crisis that
interrupts it) process.

This means that while we can hope to solve more problems as time passes,
there is no guarantee that a specific problem is solvable in this
generation, the next, or indeed ever

>therefore the causal agents must be described accordingly in teleological terms, not counter factually in non-teleological terms. The Bible most clearly states that God unleashes disease into the biosphere. Alan has never answered these points except to restate his claim that RMNS exists in his medical laboratory.
>

> And the only thing I said about Alan's use of mathematics and probability theory is the fact that he has exalted the same into a place of preeminence, when in fact observation and logic occupy the place of preeminence.

Since the mid 19th century logic is simply a part of math

>Alan has the proverbial cart before the horse because he really believes mathematics and probability theory are preeminent, when in fact the same are always supplementary to observation and logic.


Well, yes and no. If a theory has a logic contradiction, it is wrong.
But equally, if it has a mathematical contradiction, or if the
mathematical machinery predicts something that is patently wrong or
impossible. So far his approach is not wrong.

But all this assumes that logic and mathematics were applied correctly
to the problem at hand. If a theory is well confirmed (explains lots of
the data r observations) and someone claims an apparent logical or
mathematical mistake, chances are that the fault is not with the theory,
but the way that person build its mathematical or logical model.

As is the case here. Alan's argument, in a nutshell, is
a) evolution of antibiotic resistance happens very quickly
b) it can be prevented only through intelligently designed interference
(multi-drug therapy) by humans and the social infrastructures they build
therefore
c) evolution almost never happens

It does not take a lot of math, or a genius, to see that something is a
bit odd about this arguement.



>
> Moreover, Alan has admitted that certain persons, presumably possessing Ph.Ds, who peer-reviewed some of his work, were unqualified to do so because they did not understand the mathematics! In this context I observed that Alan's claims have ZERO chance of enacting harm onto macro-evolutionary theory because only a handful can understand his claims. Yet we are told that science, unlike religion, is for everyone. But in Alan's world he and a few others are king. Sorry, Alan, observation and words (logic) are king.
>
> I wish Alan all the luck in the world. I hope he obtains some degree of satisfying success. But I've studied evolutionary theory long enough to know that Darwinists will not abandon the fact of evolution, via nested hierarchies, just because the numbers don't add up.
>
> Ray (Old Earth; species immutabilist)
>

[Alan Kleinman MD PhD]

So you think reproductive fitness is not a consequence of ns?

>
> > >
> > > When you say "beneficial mutation" that's what it means - a mutation such that the fitness of the organism carrying that mutant is greater than the fitness of the wild type (or greater than the weighted average of the fitness of all the genotypes in the population).
> > OK
> > >
> > > Your model does not model fitness. All you do is wave your hands and say that the number of organisms with the beneficial mutant goes up. That's a slogan, not a model.
> > All evolutionary trajectories by rmns are represented by nested binomial probability equations where the individual probability equations are linked by the multiplication rule of probabilities. The only evolutionary trajectories which give a reasonable probability of occurring are ones where each binomial probability is high. This means that the number of replications at each step must be sufficient to give that high probability.
>
> Your model does not model fitness. There is no term in the model which represents the fitness.

What do replications give? If more replications do not give more members of a particular variant (reproductive fitness), what do they give, chop liver?

>
>
> > >
> > > Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.
> > Each evolutionary step on an evolutionary trajectory is of the form:
> > P(X)=(1 − (1 − P(Beneficial)𝜇)^(n∗nG))
> > where P(X) is the probability that the beneficial mutation occurs, P(Beneficial) is the probability of all the possible mutations that can occur at the particular site it is the beneficial mutation, 𝜇 is the mutation rate, n is the subpopulation that would benefit from the particular beneficial mutation, nG is the number of generations which that subpopulation replicates. Note that n*nG is simply the total number of replications. The n*nG term could be written more generally as a double summation of the sum of replications of the particular variant in a particular generation summed over all generations.
>
> >The fitness of the particular variant is given by the n*nG term. A low reproductive fit variant will have low n*nG and therefore low P(X) of taking the particular evolutionary step.
>
> No. The fitness of the particular variant is *not* given by the n*nG term. The fitness is given by the derivative of n with respect to time. The n*nG term just tells you the total number of organisms in the subpopulation that have lived until the present. Velocity is not the distance traveled.

Is that so? So if a one population of bacteria with a generation time of 20min and doubles and another population of bacteria (same number) with a generation time of 30min and doubles have different reproductive fitness? Probabilities in rmns are not dependent on the rate at which replications occur only on the total number of replications. You should understand this.

>
> >A high reproductive fit variant will have high n*nG and therefore high P(X) of taking the particular evolutionary step.
>
> OK. So show us what happens over time in your model. How do the populations of the wild type and mutants change over time? But those changes have to come out of the model; you can't just wave your hands and say "now move to the right along the n*nG axis." If you can't do that, you don't have a model.

The movement to the right along the n*nG axis simply gives the number of replications that a particular variant has made. What John does not understand and you should is that the sample space for this problem is dependent solely on the number of replications. The rate at which the replications are occurring does not affect the probabilities because it doesn't affect the sample space. If I roll two dice 10 times over 5 minutes gives the same sample space if I roll two dice 10 times over 10 minutes.

>
> > .
> > For your example of the emergence of drug-resistance to combination therapy for malaria, where double beneficial mutations must occur to improve fitness to reproduce, the equations are very similar but require extremely large population (such as malaria can attain) to have a reasonable probability of a double beneficial mutation.
>
> You really should stop talking about that malaria paper. In the first place, it's a very simple paper about monitoring the spread of drug resistant malaria in one part of Cambodia. There are plenty of more important papers in the field if you actually care about malaria. Second, you misunderstand it. The resistance we see is not occurring because of the de novo occurrence of multiple mutations in parasites with the individual patients we treated. The strains were resistant prior to treatment (as you could have figured out if you'd get off your one trick, "multiplication rule of probabilities" pony). Drug treatment enriched for the resistant parasites, but they were there from the start.

It is possible that drug resistant variants were there from the start, I acknowledge this in my paper. But then I ask (and answer) the question, is it possible that drug resistance occurred de novo, that is drug resistant variants were not in your study. And the answer is yes. You will need effective three drug therapy to have a reasonable probability that you won't see drug resistance emerging. Malaria can attain a population of e12 and in this population size, there is a reasonable probability you will see double beneficial mutations. Three drug therapy will require population size of around e18 before there is a reasonable probability of a triple beneficial mutation. I like your paper, it lays out the problem of drug resistance to combination therapy nicely.

[Alan Kleinman MD PhD]

It is the absolute number of replications of a particular variant. Each variant has its own particular evolutionary trajectory. And of course, the more fit variants will have more replications and the less fit variants will have fewer replications.

[Alan Kleinman MD PhD]

Not quite, any evolutionary process by rmns to more than one selection pressure at a will be slowed multiplicatively. If Lenski tried his experiment with both starvation and thermal stress acting on his populations simultaneously, it would take far more than 1000 generations for each beneficial mutation (that is if he doesn't drive his populations to extinction).

[Bill Rogers]

Sure, but you treat the number of replications as a parameter that you enter into your model as you like.

> >
> >
> > > >
> > > > Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.
> > > Each evolutionary step on an evolutionary trajectory is of the form:
> > > P(X)=(1 − (1 − P(Beneficial)𝜇)^(n∗nG))
> > > where P(X) is the probability that the beneficial mutation occurs, P(Beneficial) is the probability of all the possible mutations that can occur at the particular site it is the beneficial mutation, 𝜇 is the mutation rate, n is the subpopulation that would benefit from the particular beneficial mutation, nG is the number of generations which that subpopulation replicates. Note that n*nG is simply the total number of replications. The n*nG term could be written more generally as a double summation of the sum of replications of the particular variant in a particular generation summed over all generations.
> >
> > >The fitness of the particular variant is given by the n*nG term. A low reproductive fit variant will have low n*nG and therefore low P(X) of taking the particular evolutionary step.
> >
> > No. The fitness of the particular variant is *not* given by the n*nG term. The fitness is given by the derivative of n with respect to time. The n*nG term just tells you the total number of organisms in the subpopulation that have lived until the present. Velocity is not the distance traveled.
> Is that so? So if a one population of bacteria with a generation time of 20min and doubles and another population of bacteria (same number) with a generation time of 30min and doubles have different reproductive fitness?

Of course they do. The genotype with a doubling time of 20 min is more fit than one with a doubling time of 30 minutes. Put them in a chemostat and pull out a sample every day and in short order the only thing in the sample will be the genotype with the 20 minute doubling time.

>Probabilities in rmns are not dependent on the rate at which replications occur only on the total number of replications. You should understand this.

I understand perfectly that the number of a random mutants that you get depends on the number of replications. That's rm. Natural selection is what happens after the mutants occur.

> >
> > >A high reproductive fit variant will have high n*nG and therefore high P(X) of taking the particular evolutionary step.
> >
> > OK. So show us what happens over time in your model. How do the populations of the wild type and mutants change over time? But those changes have to come out of the model; you can't just wave your hands and say "now move to the right along the n*nG axis." If you can't do that, you don't have a model.
> The movement to the right along the n*nG axis simply gives the number of replications that a particular variant has made. What John does not understand and you should is that the sample space for this problem is dependent solely on the number of replications. The rate at which the replications are occurring does not affect the probabilities because it doesn't affect the sample space. If I roll two dice 10 times over 5 minutes gives the same sample space if I roll two dice 10 times over 10 minutes.

Once again you ignore selection.

> >
> > > .
> > > For your example of the emergence of drug-resistance to combination therapy for malaria, where double beneficial mutations must occur to improve fitness to reproduce, the equations are very similar but require extremely large population (such as malaria can attain) to have a reasonable probability of a double beneficial mutation.
> >
> > You really should stop talking about that malaria paper. In the first place, it's a very simple paper about monitoring the spread of drug resistant malaria in one part of Cambodia. There are plenty of more important papers in the field if you actually care about malaria. Second, you misunderstand it. The resistance we see is not occurring because of the de novo occurrence of multiple mutations in parasites with the individual patients we treated. The strains were resistant prior to treatment (as you could have figured out if you'd get off your one trick, "multiplication rule of probabilities" pony). Drug treatment enriched for the resistant parasites, but they were there from the start.
> It is possible that drug resistant variants were there from the start, I acknowledge this in my paper. But then I ask (and answer) the question, is it possible that drug resistance occurred de novo, that is drug resistant variants were not in your study. And the answer is yes.

Empirically the answer is no. The resistance mutations were present in the samples taken from the patients prior to treatment.

>You will need effective three drug therapy to have a reasonable probability that you won't see drug resistance emerging. Malaria can attain a population of e12 and in this population size, there is a reasonable probability you will see double beneficial mutations. Three drug therapy will require population size of around e18 before there is a reasonable probability of a triple beneficial mutation. I like your paper, it lays out the problem of drug resistance to combination therapy nicely.

If you want to talk about malaria drug resistance you really need to read more papers and bone up on the details. For many drugs resistance is not a single point mutation (resistance to atovaquone, in fact, is te only anti-malarial resistance that I can think of that is, in fact, due to a single point mutation, and the only one for which selection for resistance from a cloned population of sensitive, wild-type parasites can occur in the course of treatment.) Chloroquine and fansidar resistance both require multiple different point mutations. Mefloquine resistance requires progressive amplifications of n entire gene. That the peer reviewers at Stat Med failed to pick up on any of that makes me quite sure that while they may have an excellent background in statistics, they weren't really up on evolutionary biology or drug resistance.

[Vincent Maycock]

On Thu, 14 Sep 2017 13:19:56 -0700 (PDT), Alan Kleinman MD PhD
<klei...@sti.net> wrote:

>On Thursday, September 14, 2017 at 6:35:05 AM UTC-7, Burkhard wrote:

snip

>> But all this assumes that logic and mathematics were applied correctly
>> to the problem at hand. If a theory is well confirmed (explains lots of
>> the data r observations) and someone claims an apparent logical or
>> mathematical mistake, chances are that the fault is not with the theory,
>> but the way that person build its mathematical or logical model.
>>
>> As is the case here. Alan's argument, in a nutshell, is
>> a) evolution of antibiotic resistance happens very quickly
>> b) it can be prevented only through intelligently designed interference
>> (multi-drug therapy) by humans and the social infrastructures they build
>> therefore
>> c) evolution almost never happens
>Not quite, any evolutionary process by rmns to more than one selection pressure at a will be slowed multiplicatively. If Lenski tried his experiment with both starvation and thermal stress acting on his populations simultaneously,

How do bacteria ever evolve resistance at all if thermal stress and
starvation occur in the wild, preventing the beneficial mutations from
amplifying?

[Alan Kleinman MD PhD]

And what is wrong with this? However, I would call replication a variable, not a parameter. The mutation rate is more consistent with the meaning of a parameter. The probability of a particular mutation occurring at a particular site is a function of the number of replications of the particular variant. The mutation rate is a parameter of this functional relationship.

>
> > >
> > >
> > > > >
> > > > > Just write down the equation for us from your model that shows the fitness (offspring per generation) for the wild type and for the beneficial mutant. You can take a ratio and call it relative fitness, if you like, but you don't have to, if it makes you uncomfortable.
> > > > Each evolutionary step on an evolutionary trajectory is of the form:
> > > > P(X)=(1 − (1 − P(Beneficial)𝜇)^(n∗nG))
> > > > where P(X) is the probability that the beneficial mutation occurs, P(Beneficial) is the probability of all the possible mutations that can occur at the particular site it is the beneficial mutation, 𝜇 is the mutation rate, n is the subpopulation that would benefit from the particular beneficial mutation, nG is the number of generations which that subpopulation replicates. Note that n*nG is simply the total number of replications. The n*nG term could be written more generally as a double summation of the sum of replications of the particular variant in a particular generation summed over all generations.
> > >
> > > >The fitness of the particular variant is given by the n*nG term. A low reproductive fit variant will have low n*nG and therefore low P(X) of taking the particular evolutionary step.
> > >
> > > No. The fitness of the particular variant is *not* given by the n*nG term. The fitness is given by the derivative of n with respect to time. The n*nG term just tells you the total number of organisms in the subpopulation that have lived until the present. Velocity is not the distance traveled.
> > Is that so? So if a one population of bacteria with a generation time of 20min and doubles and another population of bacteria (same number) with a generation time of 30min and doubles have different reproductive fitness?
>
> Of course they do. The genotype with a doubling time of 20 min is more fit than one with a doubling time of 30 minutes. Put them in a chemostat and pull out a sample every day and in short order the only thing in the sample will be the genotype with the 20 minute doubling time.

If your concept was correct, the replicator with the shortest generation time will always be the most fit. But your concept is not correct. Your scenario is equivalent to the Lenski experiment where one variant competes with and drives another variant to extinction. For rmns to work, all a variant has to do is replicate a sufficient number of times for there to be a reasonable probability of a beneficial mutation. It doesn't matter how much time it takes for those replications to occur.

>
> >Probabilities in rmns are not dependent on the rate at which replications occur only on the total number of replications. You should understand this.
>
> I understand perfectly that the number of a random mutants that you get depends on the number of replications. That's rm. Natural selection is what happens after the mutants occur.

The number of mutations which occur depends on the mutation rate. Natural selection will amplify any mutation that has improved fitness in the particular environment. This amplification is measured by the number of replications. And if there is sufficient amplification, one of the members of the variant will have a reasonable probability of another beneficial mutation.

>
> > >
> > > >A high reproductive fit variant will have high n*nG and therefore high P(X) of taking the particular evolutionary step.
> > >
> > > OK. So show us what happens over time in your model. How do the populations of the wild type and mutants change over time? But those changes have to come out of the model; you can't just wave your hands and say "now move to the right along the n*nG axis." If you can't do that, you don't have a model.
> > The movement to the right along the n*nG axis simply gives the number of replications that a particular variant has made. What John does not understand and you should is that the sample space for this problem is dependent solely on the number of replications. The rate at which the replications are occurring does not affect the probabilities because it doesn't affect the sample space. If I roll two dice 10 times over 5 minutes gives the same sample space if I roll two dice 10 times over 10 minutes.
>
> Once again you ignore selection.

Reproductive fitness of a particular variant is measured by the number of replications. I do not ignore the number of replication in my equations. In fact, the probability of a beneficial mutation occurring is a function of the number of replications.

>
>
> > >
> > > > .
> > > > For your example of the emergence of drug-resistance to combination therapy for malaria, where double beneficial mutations must occur to improve fitness to reproduce, the equations are very similar but require extremely large population (such as malaria can attain) to have a reasonable probability of a double beneficial mutation.
> > >
> > > You really should stop talking about that malaria paper. In the first place, it's a very simple paper about monitoring the spread of drug resistant malaria in one part of Cambodia. There are plenty of more important papers in the field if you actually care about malaria. Second, you misunderstand it. The resistance we see is not occurring because of the de novo occurrence of multiple mutations in parasites with the individual patients we treated. The strains were resistant prior to treatment (as you could have figured out if you'd get off your one trick, "multiplication rule of probabilities" pony). Drug treatment enriched for the resistant parasites, but they were there from the start.
> > It is possible that drug resistant variants were there from the start, I acknowledge this in my paper. But then I ask (and answer) the question, is it possible that drug resistance occurred de novo, that is drug resistant variants were not in your study. And the answer is yes.
>
> Empirically the answer is no. The resistance mutations were present in the samples taken from the patients prior to treatment.
>
> >You will need effective three drug therapy to have a reasonable probability that you won't see drug resistance emerging. Malaria can attain a population of e12 and in this population size, there is a reasonable probability you will see double beneficial mutations. Three drug therapy will require population size of around e18 before there is a reasonable probability of a triple beneficial mutation. I like your paper, it lays out the problem of drug resistance to combination therapy nicely.
>
> If you want to talk about malaria drug resistance you really need to read more papers and bone up on the details. For many drugs resistance is not a single point mutation (resistance to atovaquone, in fact, is te only anti-malarial resistance that I can think of that is, in fact, due to a single point mutation, and the only one for which selection for resistance from a cloned population of sensitive, wild-type parasites can occur in the course of treatment.) Chloroquine and fansidar resistance both require multiple different point mutations. Mefloquine resistance requires progressive amplifications of n entire gene. That the peer reviewers at Stat Med failed to pick up on any of that makes me quite sure that while they may have an excellent background in statistics, they weren't really up on evolutionary biology or drug resistance.

Each beneficial mutation required to give resistance to a selection pressure is governed by a binomial probability equation. The individual binomial probability equations are linked to one another by the multiplication rule of probabilities. All rmns process works like this. Your example is no different.

[Alan Kleinman MD PhD]

On Thursday, September 14, 2017 at 2:20:05 PM UTC-7, Vincent Maycock wrote:
> On Thu, 14 Sep 2017 13:19:56 -0700 (PDT), Alan Kleinman MD PhD

> wrote:
>
> >On Thursday, September 14, 2017 at 6:35:05 AM UTC-7, Burkhard wrote:
>
> snip
>
> >> But all this assumes that logic and mathematics were applied correctly
> >> to the problem at hand. If a theory is well confirmed (explains lots of
> >> the data r observations) and someone claims an apparent logical or
> >> mathematical mistake, chances are that the fault is not with the theory,
> >> but the way that person build its mathematical or logical model.
> >>
> >> As is the case here. Alan's argument, in a nutshell, is
> >> a) evolution of antibiotic resistance happens very quickly
> >> b) it can be prevented only through intelligently designed interference
> >> (multi-drug therapy) by humans and the social infrastructures they build
> >> therefore
> >> c) evolution almost never happens
> >Not quite, any evolutionary process by rmns to more than one selection pressure at a will be slowed multiplicatively. If Lenski tried his experiment with both starvation and thermal stress acting on his populations simultaneously,
>
> How do bacteria ever evolve resistance at all if thermal stress and
> starvation occur in the wild, preventing the beneficial mutations from
> amplifying?

Replicators that can't adapt and don't have sufficient reproductive fitness in a particular environment go extinct. Replicators that can't improve fitness by rmns but have sufficient reproductive fitness in a particular environment will drift (eg HIV to combination therapy). You assume that populations will always evolve by rmns regardless of the stressors in an environment. rmns only works in very specialized instances.

[Burkhard]

So we need not bother with all the multiple drug therapies is what you
are saying? After all, in addition to the single drug, the strain would
also be under all the other stresses that exist in the wild, so
obviously cannot evolve resistance.

[r3p...@gmail.com]

"The Lord’s hand was heavy on the people of Ashdod and its vicinity; he brought devastation on them and afflicted them with tumors" (1 Samuel 5:6; NIV). Other translations say painful boils.

I could provide many more examples of God introducing disease into the biosphere. Let's not derail this thread; seems like Alan and John are making some progress; perhaps start a new topic?

Ray

[Alan Kleinman MD PhD]

Absolutely not. What I am saying is that if you are going to gamble with bacteria, play a game where the bacteria has to win multiple simultaneous lotteries in order to win the game. Single drug therapy usually works in the community environment because you are working with a generally healthier population than seen in the hospital. These people have good functioning immune systems. On the other hand, hospitalized patients (where drug resistance is a greater problem) don't have as good functioning immune systems. I always take into account the immune status of my patients. The sicker patients will get double and occasionally triple antibiotic therapy.

[Vincent Maycock]

On Thu, 14 Sep 2017 14:29:15 -0700 (PDT), Alan Kleinman MD PhD
<klei...@sti.net> wrote:

>On Thursday, September 14, 2017 at 2:20:05 PM UTC-7, Vincent Maycock wrote:
>> On Thu, 14 Sep 2017 13:19:56 -0700 (PDT), Alan Kleinman MD PhD
>> wrote:
>>
>> >On Thursday, September 14, 2017 at 6:35:05 AM UTC-7, Burkhard wrote:
>>
>> snip
>>
>> >> But all this assumes that logic and mathematics were applied correctly
>> >> to the problem at hand. If a theory is well confirmed (explains lots of
>> >> the data r observations) and someone claims an apparent logical or
>> >> mathematical mistake, chances are that the fault is not with the theory,
>> >> but the way that person build its mathematical or logical model.
>> >>
>> >> As is the case here. Alan's argument, in a nutshell, is
>> >> a) evolution of antibiotic resistance happens very quickly
>> >> b) it can be prevented only through intelligently designed interference
>> >> (multi-drug therapy) by humans and the social infrastructures they build
>> >> therefore
>> >> c) evolution almost never happens
>> >Not quite, any evolutionary process by rmns to more than one selection pressure at a will be slowed multiplicatively. If Lenski tried his experiment with both starvation and thermal stress acting on his populations simultaneously,
>>
>> How do bacteria ever evolve resistance at all if thermal stress and
>> starvation occur in the wild, preventing the beneficial mutations from
>> amplifying?
>Replicators that can't adapt and don't have sufficient reproductive fitness in a particular environment go extinct.

Right.

> Replicators that can't improve fitness by rmns but have sufficient reproductive fitness in a particular environment will drift (eg HIV to combination therapy). You assume that populations will always evolve by rmns regardless of the stressors in an environment. rmns only works in very specialized instances.

You still haven't answered the question of how bacteria are able to
develop resistance if there are all those selection pressures acting
on them as they "try" to evolve resistance to antibiotics.

I mean, empirically, they have been observe to become resistant,
without going extinct; so how do they do it?