Are you not a student of population genetics? Go back to your text book and you will find that population geneticists distinguish three types of selection; they are stabilizing selection, directional selection and disruptive selection. Stabilizing selection favors an intermediate optimum phenotype and selects against phenodeviants. Directional selection is selection directed toward a new intermediate optimum, not at the mean of the population, in response to a new selective challenge. Disruptive selection, rather than favoring any one phenotype, favors two quite different forms and selects against the intermediates; this may be viewed as directional selection of two separate subpopulations, in response to two different sets of environmental conditions. These definitions are taken from the text Genetics in Medicine by Thompson & Thompson.
I think I am beginning to understand the depth of your misunderstandings. I will honestly do my best to educate you, as, in this case, it is an easy mistake to make. Please forgive my headings and the like; I'm in a strange mood.
=Introduction of key terms=
There are few key terms that we must understand before continuing, these being "variation", "fitness" and "selection".
-Variation-
Variation is a key concept in evolution, one which is vital for evolution if evolution is to occur. If we restrict our example to a single locus on a genome, which we will call the (made up)
evo gene. This imaginary gene is 100 basepairs long. To measure
genetic variation in a population (the only form of variation I am dealing with because, essentially, all variation arrises from this), we need to know the exact sequence of this gene. In our example, the sequence is the basepair Guanine (G) 100 times.
Genetic variation occurs when there exists different sequences for the
evo gene in the population. Say, 20% of our population contains 80 G and then 20 T (lets call this the
B allele, and the other (100% G) the
A allele). The more variations of the same gene (called "
alleles"), the more variation exists in the population.
-Fitness-
Fitness is fairly easy to define, and very difficult to quantify. Fitness, simply put, is a measure of "how well" an organism does compared to another. Generally, this is measured by how many offspring an organism could produce, but naturally this is very hard to judge. In practice, "fitness" is thought of in terms of alleles, not organisms. If we take some samples, and find that the 20% of our imaginary population which have the B allele, when compared to those with the A allele, produce 25% more offspring, then we could conclude (depending on statistical support for these results of course) that the fitness of the A allele is 0.75 that of the B allele. There is a reason why we do not consider the fitness of the B allele to be 1.25 that of the A allele, which I shall get into later.
In mathematical models, the fitness of an allele is given by "1-s", where "s" is the
selection coefficient. In our case above, the selection coefficient would be 0.25, thus making the fitness 0.75 (1-0.25=0.75).
-Natural Selection (or "What now?")-
This is where the real bulk of evolution happens; through natural selection. In common language, selection is the force which "weeds out" unsuccessful traits in a population. It is often said that selection "selects against" unsuccessful traits while "selecting for" successful traits. Unfortunately, this is not quite the truth, but rather a close approximation which is easier to explain. Firstly, selection is not a "force" at all, but should rather be considered a
phenomenon of nature. It occurs everywhere and always. Secondly, and most importantly, the phenomenon we call "natural selection" is simply the fact that those organisms which are less fit (i.e. have a higher selection coefficient) the others are, on average, less able to reproduce. Thus, they have a lower contrabution of genetic variation to the next generation then those with higher fitness (lower selection coefficients). Over time, therefore, the proportion of alleles in a population with higher levels of fitness will grow, and those with lower fitness will fall. It is this change in allele frequences which is evolution.
Now, kleinman, we come to your (implied) question. What is the difference between "stabilizing selection, directional selection and disruptive selection". The key to understanding this is that the selection occuring in these three different situations is not different in any way. Selection in all three cases provides "negative pressure" on alleles with lower fitnesses. What these different things are, are
models of the
results of selection, given different allele fitnesses. But natural selection does not change in any way. The equations are still the same. Alleles with high selection coefficients will be strongly selected against, and alleles with low selection coefficients will be weakly selected against. Obviously, alleles with no selection coefficient (i.e. the most "successful" allele) will not be selected against at all. This is important. The "best" allele
is not selected "for", it is just not selected against. The difference is very subtle, the result is often the same, but it is fundamental to understanding your misconceptions.
Why does selection not select "for" anything? It is because natural selection is blind. If it were to select "for" something, then it would have an "end product" in mind. It would have a plan. But this is not how natural selection (and evolution), works. It is completely blind, and as such, can only "weed out" the weakest. It is this which drives the change in allele frequences.
To conclude, it is a common misconception that selection is "for" and "against" different alleles (or traits, or whatever). It is often explained as if this is the case. But when you look at the models of selection, and at the science behind selection, what we really see is that selection is only "against" and "not against". The understanding of this difference is key to the understanding of evolutionary theory as a whole.
So, to answer your (implied) questions in a straight forward manner, in the different "modes" of selection, selection is not, in fact, different at all. All that is different is the result, based upon allele fitness.
Dr Richard’s link also includes “neutral selection” which essentially occurs with mutations which don’t affect fitness.
It is an old debate, and one I'm not going to get into, but the general argument goes something like "If it doesn't affect fitness, how can it be selected for? And if it can't be selected for, how is it selection?".
The goal for directed selection is a new intermediate optimum on the fitness landscape.
And right there is your misconception. There is no "goal" for evolution. Nor is there any "goal" for natural selection. Natural selection is only "against" and "not against".
The reason you don’t understand this is that you don’t understand the different types of selection pressures. Go back to the Ka/Ks equation. When Ka/Ks = 1 you have neutral selection pressure, that is mutations occur which don’t affect the frequency of the particular gene in the gene pool, when Ka/Ks is « 1 you have a stabilizing selection pressure, that is a mutation causes a phenodeviant which is detrimental to the creature and is selected against. When Ka/Ks is « 1, the frequency of that particular mutation is reduced in the gene pool. Only when Ka/Ks > 1 do you have selection which increases the frequency of that mutation in the gene pool.
Again, it is your misunderstanding. Selection behaves no different in all three of these cases, only the
result of that selection is different. When Ka/Ks = 1, the allele has no selection acting on it at all. When Ka/Ks < 1, the allele is being selected against. When Ka/Ks > 1, selection is not selecting against it. You have to remember that fitness is relative. This is one of the reasons why it is so hard to give a numerical value to fitness. Fitness is not a finite number, but a relative value
based on the "optimal" allele fitness. In other words, all alleles' fitnesses are being compared to the "best" allele. The exact same allele (i.e. the same allele with the same genetic sequence) can be Ka/Ks < 1 in one population, and Ka/Ks > 1 in another. I.e. the same allele can have a low fitness in one population, and a high fitness in another. This is because fitness is relative to all other possible alleles in a population, and why all selection models always have one allele with a fitness of 1 (and thus, no selection coefficient).
These types of mutations lead to a new optimum on the fitness landscape and are thus directional in nature.
No. Selection is not directional, because selection does not have a "goal in mind" when it "selects". Selection, and evolution, are blind.
The authors have indicated that only in rare cases are selection pressures positive (or directional) (Ka/Ks > 1).
It is very true that beneficial mutations are rare.
Neutral or stabilizing selection processes do not lead to new strains in the gene pool.
If you mean "Neutral and stabilizing selection processes do not lead to
a change in allele frequency", then I agree with one small correction. Alleles which are undergoing neutral selection most certainly do change in frequency. Look up "genetic drift".
Only when you have directional selection do you introduce new strains in the gene pool and again, these authors indicate that directional selection for HIV is rare. When antiretroviral medications are used on HIV, these apply directional selection pressure on the virus. Note that when a resistant strain evolves, that directional selection pressure will become a stabilizing selection pressure if the virus is still exposed to the drug.
Only directional selection pressure leads to genetic evolution. Of the hundreds of selection pressures Dr Richard refers to, only a tiny fraction of these are directional selection pressures. If you want to evolve drug resistance in HIV, use a single directional selection pressure and you can achieve this quickly because of the viruses’ high mutation rate and rapid reproduction. Combine multiple directional selection pressures and you will slow this process.
What slows is not evolution, but the arisal of the correct variation in the population. Evolution does not slow, the levels of variation decrease. Please note that this is only the case in instances where the population is significantly reduced. However, evolution (the change in allele frequency over time) does not slow. What is happening is the level of "unique features" in a smaller population is less.
You better go back and re-read your textbooks before you take your final exams.
Just for your information, I have no more finals. I already have my degree.
Unless you mean my finals for my graduate studies, in which case I can assure you that they will not deal with such basic issues which we are discussing. This stuff was covered in my second year.
I don’t know what to tell you other than go back to your textbooks and learn what the different forms of selection do.
It is you who does not understand, kleinman. The different forms of selection do not "do" anything, they "just are". The different forms are different forms
of results, not different ways in which selection works.
If you teach what you are saying here, you will be misleading your students on how mutation and selection works. What can you expect from an evolutionist?
I would expect any evolutionary geneticist to understand these basic principles of genetic evolution, and so naturally I would teach it.
You still don’t understand the mathematics of mutation and selection. Once you understand what the different forms of selection are, perhaps you will be able to begin to understand the mathematics of mutation and selection.
Kleinman, you are wrong. Simply, plainly, and completely, wrong. I
do understand "the mathematics of mutation and selection" (insofar as that is actually a meaningful term), and I
do understand how selection works. It is you who does not understand, and refuses to learn.
If you ever graduate Sesame Street and learn how to count, you will understand how mutation and selection really works. Your own link to The HIV positive selection mutation database shows this but it must be too technical for you to understand.
This is what happens when you try to discuss mathematics with a Sesame Street drop out.
He has a far greater grasp of all of this then you do, kleinman.
![picrender.fcgi](http://[/SIZE][/FONT][URL="http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1669717&blobname=gkl855e1.jpg"][FONT=Times New Roman][SIZE=3][COLOR=#0000ff]http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1669717&blobname=gkl855e1.jpg[/COLOR][/SIZE][/FONT][/URL][FONT=Times New Roman][SIZE=3])
