On 8/1/2014 3:30 PM, Steady Eddie wrote:
> On Friday, 1 August 2014 13:13:44 UTC-6, Greg Guarino  wrote:
>> On 8/1/2014 2:46 PM, Steady Eddie wrote:
>>
>>
>>
>>> Why would NEUTRAL mutations be passed on to that individual's offspring?
>>
>>
>>
>> I likely have at least several dozen mutations in my DNA; "letters" that
>>
>> do not match either my mother's or father's genome at those sites. None
>>
>> of them were fatal, and none of them have prevented me from reproducing.
>>
>>
>>
>> Even though we could describe those mutations as "copying errors", they
>>
>> are nonetheless part of my genome now; copied (mostly) faithfully
>>
>> trillions of times into the cells in my body, notably including my sperm
>>
>> cells.
>>
>>
>>
>> What would prevent (some of) those mutations from being passed on to my
>>
>> offspring? "Selection" is no answer, as I have successfully reproduced.
>>
>> Statistically speaking, roughly half a parent's mutations should be
>>
>> inherited by each child.
>
> Yes, then half of that in the next generation, until your particular mutations exponentially approach zero.
> Meaning, as your progeny continues to reproduce, your particular mutations get CANCELLED OUT
>   quite quickly by the original genome in the population (unless, in the case of the wild, a mutation
> confers an evolutionary advantage).
I was going to post a one-liner about just how profoundly you're 
misunderstanding things but that's pointless.
Please think on what you wrote here. You are NOT describing something 
that's "unselected" (as you wrote in a different post). You are 
describing (without realizing it) a mutation under significant selection 
pressure- probably more powerful selection pressure than you'd ever see 
in the wild, other than for a lethal dominant allele.
Remember when you asked where Ernest Major got that number 10E-08? This 
is where it where it comes into play. 10E-08 seems to be a pretty good 
estimate of the mutation rate in eukaryotes. So every individual born, 
hatched, whatever, has a number of mutations. Some number of these are 
deleterious (but NOT lethal, please remember that for later), most are 
neutral, and a very few might be beneficial.
Let's stick to the mutations that are neutral. Do you know how a 
mutation can be neutral? A moment's calculation will tell you there are 
64 possible combinations of the four nucleotides (adenine, guanine, 
cytosine, thymine) that make the DNA triplets- each triplet codes for a 
particular amino acid. One triplet codes for AUG, the "start" codon, and 
three code for "stop" codons, so there's actually 60 triplets for 20 
amino acids (AUG also codes for methionine, so line up quick for pedant 
points...)
This allows for some redundancy. Many amino acids will be produced by 
more than one triplet. Proline is an example. Proline is coded for by 
several DNA triplets, including GGG and GGT. So we could have a mutation 
in a DNA triplet, changing G to T or vice versa, and it's completely 
neutral.
Another way to get a neutral mutation becomes obvious when you look at 
the structure of enzymes. Most enzymes are pretty large proteins. The 
most critical region is the binding site, where the substrate attaches. 
A change in an amino acid there is likely to alter or even wreck the 
function of the enzyme- the substrate just won't be able to bind there 
anymore. Second, the enzyme has to have a particular shape to it. Ever 
hear the phrase "form follows function"? It's true with a vengeance in 
enzymes. The shape is determined by other "levels" of protein structure, 
in particular things like van der Waal's forces and disulfide bonds 
between amino acids that are distant from the binding site. A change 
here is slightly less likely to result in change or loss of function, 
but mutations can be serious. Sickle-cell disease is an example: the 
mutation is not a change at the oxygen binding site, but the molecular 
structure is altered such that the hemoglobin is liable to collapse if 
it is depleted of oxygen too rapidly, causing the erythrocyte to change 
shape or "sickle".
Now, neutral mutations are not "cancelled out" by the population's 
genome. There's a real chance that a mutation will be lost due to drift 
("dumb luck" as you correctly put it). Remember- an organism might die 
without progeny, and even if it does reproduce, there's only a 50% 
chance that the mutation will be inherited by any given one of its 
offspring. But neutral mutations are accumulated over time. Heck, 
compare how long it takes to eliminate a _lethal_ recessive allele from 
a population and it becomes obvious that _neutral_ recessive mutations 
are just going to stick around forever.
How do they accumulate, you ask? Well, they just keep happening, over 
and over and over again. 10E-08, remember? How many nucleotides do you 
have in your genome? On average, we've all got about a dozen mutations 
that we're stuck with, and that doesn't even include the ones that 
happened in our parents, grandparents, great-grandparents etc. If you 
have any kids, each one will get about 6 of your mutations, and have a 
dozen of their own. Most are neutral, some might be deleterious, and 
some might be beneficial. Some might be back-mutations that actually 
revert an earlier mutation to its previous state.
Now, why did I say to remember the difference between deleterious and 
lethal? Lethal alleles are obviously a subset of deleterious alleles, 
but by no means are they the whole of the set. Too many people equate 
the two, and as we see in sickle-cell disease (and for that matter 
Huntington's disease, a lethal _dominant_ disorder, although there's a 
complicating factor there) even a disease that can kill you will not 
necessarily prevent you from reproducing. "Deleterious" in the sense of 
population genetics really just means "reduced fertility" or "lower 
reproductive success". Might kill you, might not, but on average, you'll 
have _fewer_ surviving progeny than someone not carrying that particular 
allele or pair of alleles, but probably not _zero_ progeny.
This obviously applies to Behe's work, and I have not seen anyone bring 
it up yet. A mutation may very well be deleterious (the mutations in 
_Plasmodium_ were apparently neutral, or nearly so) but so what? Unless 
it's a lethal dominant allele, it is NOT going to be removed from the 
population by selection in a single generation.It might persist for a 
very long time indeed.
And if it is a deleterious recessive allele, its very rarity will help 
it persist, wrt selection, at least. If it's recessive, it will not 
exert deleterious effects unless it is paired with another copy of the 
mutation- and that means it's highly unlikely to be exerting any 
ill-effects on the organism's reproductive success.
This is why inbreeding/incest can have consequences. It's not so much 
that one person might be a mutant- we're all mutants. But the more 
closely related you are to your mate, the more likely it is that you're 
carrying the SAME deleterious recessive alleles, rather than a 
completely different set.
Chris