Mutations, fitness and more
by BradfordTwo TT commenters, Zachriel and Salvador Cordova have an interesting exchange ongoing in the open thread. Salvador initiated the discussion with this comment. In one of his comments Zachriel mentioned the genetic load concept. He also cited the following papers:
Nachman & Crowell, Estimate of the mutation rate per nucleotide in humans, Genetics 2000
Whitlock & Bourguet, Factors affecting the genetic load in Drosophila: synergistic epistasis and correlations among fitness components, Evolution 2000
Commonly used terms like natural selection and fitness are part of the exchanges.
In a recent comment Salvador said the following:
At the very least, it means, selection can't even detect or correct 108 new dysfunctional mutations per generation. On what grounds then can we suppose selection could have incorporated these mutations into the population in the first place?
This brings up an interesting question. If deleterious mutations accumulate within genomes without fatally compromising the survival of individual organisms until a threshold is reached, then what does this say about the role natural selection could have played in fixing such biological properties in the first place? At first glance it seems counterintuitive to think that x was selected for when x's disablement does not appreciably impact the fitness of the affected organism. So how does theory, and more importantly data, relate to Salvador's point?
November 12th, 2009 at 2:50 pm
Thank you Bradford!!!!!!!!
First off before going into all the population genetics, make sure to see the animation.
It takes about 2 minutes to load and only 1 minute and 25 seconds to watch. It is an mpeg file. We hope to make it available on youtube soon.
http://www.idcsnetwork.com/aud...
The creatures are represented by ginger bread men. A harmful mutation is represented by a red dot. It might have been a little more illustrative to position the read dot in various places for each creature to emphasize the novelty, but this animation is only a draft cut.
Explosions of gingerbread men represent elimination by natural selection. The point is that even after selective sweeps, the collective genome remains irreversibly damaged (represented by the persistence of red dots on the surviving non-exploded gingerbread men).
The goal of the animation was to distill the essential points of graduate level population biology in regards to genetic deterioration to something even a high school senior could understand.
The current estimated mean rate of nearly neutral, but harmful mutations is 300 per human per generation. Even with a several standard deviations from the mean were still talking about easily 100 harmful mutations per individuals. The true number could be in the thousands, but I'm picking 100 as a conservative number.
The animation depicts the inability of selection to weed out such mutations even on the generous assumption we are dealing only with 1 mutation!
Actually, I cited that same paper repeatedly at UD and TT for at least the last 3 years. See:
http://telicthoughts.com/the-r...
That was November 5th in the "Wedge out of Closet", and it was continued in the open thread.
Zach cites the source paper that inspired the animation, but he cites it as if I had never even mentioned it!
I argue Zach has not interpreted the paper's real implications.
I also argue Zach has yet to characterize my claims accurately, much less refute them.
Comment by Salvador T. Cordova — November 12, 2009 @ 2:50 pm
November 12th, 2009 at 3:00 pm
Another way of looking at it:
Kimura,Crow, and a large number of geneticists say 90-99% molecular evolution was not the product of selection!
See:
http://en.wikipedia.org/wiki/N...
Sternberg points out 90%, maybe 100% of the genome is functional.
Ergo, if 90-99% of molecular evolution is non-Darwinian, 90-99% of evolution of functionality is non-Darwinian.
Masotoshi Nei makes the argument in a torturously written paper for the National Academy of Sciences (of which I think he is a member):
Selectionism vs Netrualism in Molecular Evolution
Note: the opposing views are not Darwinism vs. Creationism but Darwinism vs. Mutationism and Darwinism vs. Neutralism.
The neutralists and mutationists have been unwitting allies to the creationists (and some ID proponents).
The animation was to satirize how ridiculous selectionist ideas can be and to underscore the fact there are theoretical limits to the creative powers of Darwinism (aka The Norman Bates School of Hotel Management).
Comment by Salvador T. Cordova — November 12, 2009 @ 3:00 pm
November 12th, 2009 at 3:46 pm
This doesn't put our high school seniors in a very good ligt, lol.
Let's take selection out of the picture, because whether it's operating at the level of human individuals to weed out individual deleterious mutations
is somewhat questionable. Personal experiece would tell us that we're mostly all alive, even though we all have these bad mutations.
What if we just let the mutations accumulate and spread randomly through the gene pool? What then?
So wha is the probability of even one of these being fixed in the population?
Also, what if we look at this from a different perspective, at the population level, rather than individuals. Is the population as a hole becoming "less fit," and even if so, so what? Can we infer that the human population is going to go extinct due to the accumulation in the gene pool of slightly deleterious mutations?
Comment by Mung — November 12, 2009 @ 3:46 pm
November 12th, 2009 at 3:57 pm
I don't think it matters Sal. The vast majority of mutations are evolutionary dead ends, regardless of whether they are beneficial or not. And by "dead end" I mean they don't spread through the population to any significant level of frequency.
Do two questions need to be answered.
1. What is the probability of fixation of a single mutation.
2. Given 100 mutations, what is the probability of fixation of even one of the 100?
Comment by Mung — November 12, 2009 @ 3:57 pm
November 12th, 2009 at 4:10 pm
You might try reading that again.
The probability of fixation of neutral mutations is simply the rate of neutral mutations. The probability of fixation of a *particular* neutral mutation depends on population size, P = 1/2N (diploid). For an idealized, infinite population, it's zero.
Under selection, it's P ≈ 2s / ( 1− e^(−4 Ns) ). For small s and large N, that approximates to 2s.
P, probability
N, effective population
s, selection coeffecient
If there are large numbers of neutral mutations per individual in each generation, then we expect large numbers to accumulate. A deleterious mutation will be effectively neutral if s is much less than 1/2N.
Comment by Zachriel — November 12, 2009 @ 4:10 pm
November 12th, 2009 at 4:15 pm
Even assuming the probability of fixation of a harmful is low to zero, the issue is that for every one bad trait weeded of the population, several more appear. The population on the whole can be sicker even if not one harmful reaches fixation. How can this be? Everyone gets sicker but from different hereditary diseases.
Sykes at Oxford thinks so.
Sanford at Cornell thinks so.
But lets say for the sake of argument, humans don't go extinct, they might become sicker, functionally speaking.
Recall, in the world Darwin sickel cell anemia, tay-sachs disease, cystic fibrosis, and a host of other diseases are regarded in the world of Darwin, reproductively superior traits!
But a less ambitious claim, and one I think that is very defensible, is that Darwinism cannot possibly police 4 billion nucleotides simultaneously. One has to wonder how many nucleotides it can positively select for simultaneously.
Say we have
trait A: (speed)
trait B: (intelligence)
etc.
and each trait has a 1% selective advantage over the rest of the population. From an accounting standpoint, it is impossible to have 1000 traits with 1% reproductive advantage over the population.
This reminds me of a proverbial account (that has some empirical basis) that 99% of the individuals view themselves were "above average" even though it cannot likely be true that 99% of people are above average (or perhaps more precisely, above the median).
The point of the animation however was to help understand the hypothesis of genetic entropy.
Comment by Salvador T. Cordova — November 12, 2009 @ 4:15 pm
November 12th, 2009 at 4:22 pm
Harmfuls don't need to reach fixation to deteriorate the population.
The situation with beneficials versus harmfuls is not symmetric. To transform the population in the way Dawkins hypothesizes, the beneficial has to be fixed and acquired by every individual.
The situation for harmfuls is not the same. Every individual can have several rare defects, and the overall population will suffer (hence, we have health care issues for numerous hereditatry problems, and we might have many more as time goes on).
The situation between beneficials and harmfuls is not symmetric. Arguments which tacitly compare them in symmentric ways (such as probability of fixation), miss the point. The animation was intended to elucidate the problem.
Perhaps a future modification is to illustrate different positions for the red dots, and different phynotypic defects (like different missing limbs for each gingerbread man) to emphasize the point.
Comment by Salvador T. Cordova — November 12, 2009 @ 4:22 pm
November 12th, 2009 at 5:09 pm
Direct measurement of the human mutation rate is in good agreement with previous estimates based on the hypothesis that most mutations are effectively neutral.
Comment by Zachriel — November 12, 2009 @ 5:09 pm
November 12th, 2009 at 5:19 pm
Whatever survives, survives- meaning there are many reasons why an organism could live or die. Not all of which are tied to genetics.
You lose the ability to internally synthesize something you find an external source.
Comment by ID guy — November 12, 2009 @ 5:19 pm
November 12th, 2009 at 5:21 pm
Most mutations being "effectively neutral" helps out your position how?
I could see if we had direct evidence for network building mutations, but we only have evidence for network breakers, not makers.
So what is it, besides personal bias, that would lead anyone to infer that mutations can accumulate to build useful things, new molecular machines and new bodies?
Comment by ID guy — November 12, 2009 @ 5:21 pm
November 12th, 2009 at 6:08 pm
I don't think thee factors are irrelevant.
And let's not overlook that the estimated mutation rate has a per generation factor. You're treating it as if it's additive.
Here, we apply new sequencing technology to measure directly one mutation rate, that of base substitutions on the human Y chromosome. …led to a mutation-rate measurement of 3.0 × 10−8 mutations/nucleotide/generation (95% CI: 8.9 × 10−9–7.0 × 10−8),
Comment by Mung — November 12, 2009 @ 6:08 pm
November 12th, 2009 at 9:41 pm
They may not be irrelevant with respect to other points but Salvador is correct that a novel harmful for each newborn would signify a "net increase in harmfuls per individual."
Comment by Bradford — November 12, 2009 @ 9:41 pm
November 12th, 2009 at 9:53 pm
Gentlemen, how does an analysis differ for the following two types of "harmfuls?"
One type entails hidden effects on phenotype because the mutation would be recessive and not apparent unless two recessive alleles were present. Expression in the phenotype however is fatal (barring modern medicine). Hemophilia, for example.
The other type impairs an individual's fitness without causing its demise (color blindness). The accumulation of this type is more intriguing.
Comment by Bradford — November 12, 2009 @ 9:53 pm
November 12th, 2009 at 10:03 pm
Consider a very large population that randomly mates and each female produces a large number of children. Every child will then experience a single mutation. Assume strong selection in that the mutation is always fatal when homozygous, but of no consequence when heterozygous or nullizygous. The parents are heterozygous.
A) A quarter of the children will be stillborn. 25%
B) Half the children will have a recessive gene. 50%
C) And a quarter will be born with no mutant gene at all. 25%
Every child will now experience the inevitable mutation.
B) Half will die. 25%
C) All will be become heterozygous, just like their parents. 25%
So half the children will survive as heterozygotes. 50%
The result is that the offspring population has increased and yet has the same genetic composition as their parents.
Of course the actual biological situation is more complex. We're dealing with various shades of benefit or detriment, as well as other mechanisms, such as synergistic epistasis, drift and sexual selection. But it does show that the argument is not correct. The parameters do matter.
Comment by Zachriel — November 12, 2009 @ 10:03 pm
November 12th, 2009 at 10:19 pm
Now, if we change some parameters, we might get different results. The previous example was with an unbounded population. With a limited population, it could happen that no children survive—by chance alone. Increasing fecundity reduces this chance. Of course, we don't expect exactly one mutation per child, so if we distribute the mutations randomly across the population, some will be hit more than once, while others will have no mutation. There may even be reverse mutations. Finally, if there is a benefit to those who are nullizygous, then environmental or sexual selection may change the result.
Comment by Zachriel — November 12, 2009 @ 10:19 pm
November 12th, 2009 at 11:27 pm
Effectively neutral means effectively invisible to selection!!!! Ergo, most mutational changes to functional systems are invisible to selection.
Reproductive fitness is a horrible way to estimate the level of functionality. In the world of Darwinism, blindess in cavefish and broken pumps in bacteria (which provide antibiotic resistance) are "beneficial" traits. A decrease in function (such as losing the ability to see in cavefish) can be a negative functional change that results in a reproductive increase.
Thus, it is completely possible that something can be invisible to selection, or even disfavored by selection, but functional.
I gave some examples of systems that are invisible to selection, but are functional. See: Airplane Magnetos, contingency designs, and reasons ID will prevail
Comment by Salvador T. Cordova — November 12, 2009 @ 11:27 pm
November 12th, 2009 at 11:32 pm
No. Because the offspring have novel mutations that the parents are not carrying.
And if that is not the scenario you are describing it is a misinterpretation of the premise I put forward, namely, the offspring have a different genetic composition than their parents owing to the fact they are carrying a novel mutation which their parents don't have.
Comment by Salvador T. Cordova — November 12, 2009 @ 11:32 pm
November 13th, 2009 at 8:15 am
Bradford:
Sal:
I am utterly amazed that these facts have to be repeated. It really illustrates the magnitude of the communication gap that must be over come in these discussions.
It’s almost like the principals are having different conversations
peace
Comment by fifth monarchy man — November 13, 2009 @ 8:15 am
November 13th, 2009 at 8:23 am
That's right. Most mutations are effectively neutral. That means there are only a small number of mutations that can contribute to mutational load, directly contradicting your stance. None of the rest of your post concerns the original claim about mutational load.
Keep in mind the fact that the directly measured mutation rate is consistent with rates posited on neutral evolution between humans and chimpanzees. Most of the genome really is junk. (Any effect of redundancy would show since the divergence between humans and apes.)
Comment by Zachriel — November 13, 2009 @ 8:23 am
November 13th, 2009 at 8:40 am
What divergence?
There isn't any evidence that such a transition is even possible- especially given what you have been posting.
Does anyone else think that it is strange that we have plenty of evidence for neutral and harmful mutations but no evidence whatsoever for body plan changing mutations nor molecular machine building mutations?
It is as if the entire theory of evolution was "built" on imagination alone.
It is amazing what passes for science with the anti-ID mob…
Comment by ID guy — November 13, 2009 @ 8:40 am
November 13th, 2009 at 9:05 am
Here's a different scenario, again with a very large population and high fecundity. Every child receives negative mutations, but sometimes also receives a positive mutation (big teeth). Those with the most positive mutations (biggest teeth) eat those with the most negative mutations (their sickly siblings). This will act as a direct check on the number of negative mutations. All that is required to maintain the population is sufficient numbers of offspring.
Of course this is an exaggerated case, but is sufficient to disprove the assertion.
Comment by Zachriel — November 13, 2009 @ 9:05 am
November 13th, 2009 at 9:11 am
The hypothesis that humans and chimpanzees share a common ancestor with most of the differences due to neutral evolution (Nachman 2004) accurately predicted the direct measurement of the mutation rate in humans (Xue et al. 2009). Quite amazing!
Comment by Zachriel — November 13, 2009 @ 9:11 am
November 13th, 2009 at 9:27 am
I think we're on track, more-or-less. The idea is that slightly deleterious mutations tend to accumulate in genomes. This is a real phenomena that has been extensively modeled and studied. Selection tends to work at the level of the organism and can't look into the genome to purge minor differences. (This is not quite correct. Differential reproductive rates may make a finer gradation.) This is not an issue with many organisms, such as bacteria, because at least some of the offspring are exact clones of the parents. But with other organisms, such as slow-reproducing mammals, every child is a mutant.
My complaint with Salvador T. Cordova's argument is that he has over-simplified the problem. He claims that if every child has a deleterious mutation, then genomic overloading is inevitable regardless of all other factors. This is incorrect. There are a number of investigations that demonstrate the effectiveness of various mechanisms that can mitigate the accumulation of deleterious mutations, including sexual recombination, sexual selection, positive mutations, synergistic epistasis, along with the old stand-bys, environmental selection and differential reproduction.
There is apparently a maximum mutational load that these mechanisms can account for. But life tends to exist on the edge for the greatest evolutionary flexibility, so it is not unexpected that it would use 'every trick in the book' to push the limit.
Comment by Zachriel — November 13, 2009 @ 9:27 am
November 13th, 2009 at 10:38 am
No, all you have done is to repeat Darwinism's contorted view of reality where "beneficial" is tied to reproductive success and not to function or design. This is an attempt to solve the problem of the appearance of design by redefining, equivocating, and resorting to double speak distortions. It is a world that can redefine "harmful" in the engineering sense to "beneficial" in the Darwinian sense, and thus give the impression the problem of increasing harmfuls is somehow solved.
There is much to be said on this topic, and I hope to post more later since it is relevant.
In the world of Darwinism, gentic propensitities for rape (Thornhill and Palmer), murder (David Buss), sickle-cell anemia, obesity, high cholesterol, PANDAS (see Moalem) — are beneficials. Moalem's book became a NY Times best seller, his thesis was that Darwin was a genius for showing us how genetic diseases advance the human species! Moalem went on to extol the virtues of various genetic diseases.
Loss of function (such as the alternative pathways I described) could easily be neutral, but by any engineering standard it degrades the existing designs in an organism (see the discussion above about the misleading results of knockout experiments).
So no, my stance has not been contradicted.
Emergence of selective advantage is not the same as emergence of design, as Alan Orr notes.
PS
Orr was responding to Dennett's apparent attack on population biologists who were neutralists or at least non-Darwinian's. Dennett might have thought them as some sort of ally of the creationists.
Comment by Salvador T. Cordova — November 13, 2009 @ 10:38 am
November 13th, 2009 at 11:22 am
As I said, reproducitve "fitness" in the population sense is a horrible way to define functional architectures and designs. Are there designs which may not be functional, but are reconizable as designed architectures. Possibly.
For example, if I laid out coins in the following configuration (H=heads, T = tails)
HHH TT HHH TT HHH TT ……
or how about
HHH THHTTH HHH HHTTTH HHH ….
where the string HHH appears at regular intervals despite what string appears in between (the non bolded)….
These examples can't in any way be claimed as designed based on reproductive fitness or even engineering function, but they can be recognized as designs in the architectural sense.
It turns out the ENCODE project is detecting these sorts of designs. In the coming years, there will be much more elucidation of these sorts of designs that exist in the genome.
What could be especially damaging to the Darwinian case is if we find certain patterns between species.
Perhaps Dembski dream of detecting steganography will be realized!
It may be desireable to decouple the discussion of genetic deterioration from questions of reproductive "fitness" and functionality. How can this be done? One way is to tie it to the deterioration of detectable architectures with respect to the secondary designs which the ENCODE project is helping us to see.
If we see deterioration in these secondary designs, it will be consistent with the question Bradford's OP:
I suggested to Dr. Sanford this might be a fruitful area of research of genetic entropy since it bypasses the contentious issues of defining "fitness" and "function". It will be more in line with conceptions of statistical mechanics which we see in physics, thus much more amenable to empirical measurement.
If selection can't police these architectures, it would seem (circumstantially if not formally) that selection had nothing to do with their emergence in the first place. Hence, Darwinism can't account for certain designs in nature, and neither can stochastic processes. One would thus be tempted to think an intelligence of great ability was the Author.
Comment by Salvador T. Cordova — November 13, 2009 @ 11:22 am
November 13th, 2009 at 11:23 am
In the scenario, negative mutation is used in the common sense of weakening.
* You suggested that a genome can't possibly purge numerous new, slightly deleterious mutations. This is incorrect. We can show it depends on a number of factors.
* You said the average human has hundreds of new deleterious mutations per individual. I've posted evidence that this is not true.
Why were the estimates of the human mutation rate based on the hypothesis of the common ancestry of humans and chimpanzees so predictive of the direct measurements?
Comment by Zachriel — November 13, 2009 @ 11:23 am
November 13th, 2009 at 11:44 am
I think this recent paper may be relevant to the discussion:
http://www.genetics.org/cgi/ra...
Springman, R., Keller, T., Molineux, I.J., Bull, J.J. (2009) Evolution at a high imposed mutation rate: adaptation obscures the load in phage T7. Genetics epub ahead of print.
The authors subjected bacteriophage T7 high levels of mutagens to test the hypothesis that fitness would decline significantly, perhaps to extinction, as deleterious mutations overwhelm the beneficial ones. Instead, fitness increased. the authors observed a large number of deleterious mutations, but a relatively small number beneficial mutations were able to overwhelm the effect of the deleterious ones.
Simulations of mutational load need to take into account real-world observations like this.
Comment by Nick — November 13, 2009 @ 11:44 am
November 13th, 2009 at 1:13 pm
Evolution, like water and electricity, always takes the path of least resistance.
That's my hypothesis.
Comment by Daniel Smith — November 13, 2009 @ 1:13 pm
November 13th, 2009 at 2:20 pm
Thanks for the comment Nick. You're right that we need to take into account real-world observations. Quoting from Evolution at a high imposed mutation rate: adaptation obscures the load in phage T7 by R. Springman et al.
The authors made noteworthy points. But I hesitate to apply the conclusions to organisms like primates as the differences in physiology call into question the applicability of the study paradigm. I can go into further detail on this if the point is unclear to anyone. The reference by Nick is a step in the right direction responsive to the OP's call for data sources.
Comment by Bradford — November 13, 2009 @ 2:20 pm
November 13th, 2009 at 3:03 pm
I'm still trying to undersatnd why the simulation doesn't come down to the following:
G = Generation
P = Parent
Pf = father
Pm = mother
C = Child
C1 = first child
C2 = second child
: G1
Pf – 1 mutation
Pm – 1 mutation
C1 = 2 mutations (inherited) + 1 new mutation
C2 = 2 mutations (inherited) + 1 new mutation
The children become parents.
: G2
Pf – 3 mutations
Pm – 3 mutations
C1 = 6 mutations (inherited) + 1 new mutation
C2 = 6 mutations (inherited) + 1 new mutation
The only way for mutations to accumulate, which is essential to Sals argument, is to be inherited.
I think you can see how quickly the proposition becomes absurd.
So I think this is a reductio ad absurdam disproof of the argument.
Just try starting with 100 new mutations per generation for a real treat.
Comment by Mung — November 13, 2009 @ 3:03 pm
November 13th, 2009 at 3:12 pm
By the way, another conceptual problem with the simlulation is that when a certain number of mutations are deemed sufficient to lead to the termination of the lineage, it would be true of all "children" of the same generation and thus the simulation should not represent any new children.
You should have, mutation accumulate, more accumulate, more accumulate. BOOM! Everyone dies.
The only way to avoid this is:
1.) fluctuating mutation rates in different lineages in the simulation
2.) different levels of mutations required to "poof" a gingerbread-man out of the population.
Otherwise you have to explain how the accumulatiog "genetic load" isn't the same for all individuals and isn't fatal for all individuals at the same level of load.
I put "genetic load" in parens because I don't think it's a term that is properly applied to individuals.
It seems to me that Sal is taking a population argument and trying to apply it to individuals.
http://www.everythingbio.com/g...
Comment by Mung — November 13, 2009 @ 3:12 pm
November 13th, 2009 at 3:45 pm
It can purge some, it can't purge them all. You're distorting my argument.
Yes, like presuming a good fraction of the offspring don't have new muations, contrary to the premise I put forward.
You posted the mutations are nearly neutral with respect to reproductive fitness, that is not the same as non-harmful for reasons that were repeated several times but reasons which you refuse to acknowledge.
The problem with a Blind/Mindless Watchmaker is that it is myopic. Fitness in Darwin's world is defined with respect to immediate reproductive success, not long term reproductive success or function. The Mindless Watchmaker has no foresight, it resorts to expediency, and thus is quite willing to sacrifice functions that may be beneficial at some future time in another context either because it does not immediately see the benefit or there is a more pressing matter — such was the case with sickle cell anemia, cystic fibrosis, and who knows how many of the thousands (perhaps tens of thousands) of genetic disorders. That was the case with gammarus minus and broken pumps in antibiotic resistance. Here is another example of selection destroying functional complexity: Microsporidia (HT: Mike Gene). The organisms are "fit" in the Darwinian sense, they are deterioated in the functional sense.
To illustrate further, consider what "fit" means in this scenario:
Knock out path A, no lethal effect (effectively neutral). Knock out path B, no lethal effect (effectively neutral). Knock out A and B simultaneously, then a lethal effect.
The problem with the myopic definition of "fitness" which you are relying on is that it is fit with respect to immediate effects. A perceptive fitness would assign non-neutral effects to path A and B separetly. As it stands, one will come to the following absurd conclusion:
1. knock out path A, 0 effect on reproductive fitness
2. knockout path B, 0 effect on reproductive fitness
3. Sum the effects of knocking out A and B = 0 + 0 = zero effect on reproductive fitness.
Bad inference!
Thus your claim of being truly "neutral" is suspect. It is neutral only in expedient short term horizons, it is unlikely to be neutral in long term horizons, and most definitely not in terms of functional issues if Sternberg's claim that 90% or more of the genome is actually functional.
You're quick to claim synergistic epistasis in short term horizons, but where is it recognized in long term more complex horizons where combination mutations can accumulate and then be disastorous long after it is too late to purge them (if such a thing were possible). The case of Microsporidia is illustrative of how well blind mindless watchmakers can build watches.
I've pointed out before that the definition of "fit" in the Darwinian sense is so amorphous and incoherent that one has to wonder what use it is. Andreas Wagner and also Lewontin have pointed out the problem. See Lewontin:
Santa Fe Winter 2003
Translation: the notion of "fitness" is a conceptual mess. I've pointed some of the problems in characterizing fitness, Lewontin adds even more.
Comment by Salvador T. Cordova — November 13, 2009 @ 3:45 pm
November 13th, 2009 at 3:55 pm
Here's another way of viewing it. Biological trends can be viewed as competing but opposite tendencies. One would generate function consistent with the continuation of biological organisms. The opposite would cause degeneration of the same. The presumption is that the former tendency holds sway over the entire course of natural history. Indeed that is an assumption that leads to the destruction of plausible design concepts in the minds of some. Salvador prompts us to question the assumption.
Comment by Bradford — November 13, 2009 @ 3:55 pm
November 13th, 2009 at 4:26 pm
But I don't know why anyone would need to accept this presumption. Take extinction and punctuated equilibria, for example.
He should stick with the case then that there aren't enough beneficial mutations and/or enough time for them to have an effect.
But that's not what he's arguing. He's arguing that lineages will necessarily go extinct after a short amount of time due to accumulation of deleterius mutations.
In addition, the "simulation" is no such thing, and it doesn't represent anything connected to reality, adn shouldn't be shown to kindergartners, much less high school or college students. (Unless his goal is to make ID look bad.)
Look, I'm as willing as the next guy to take shots at MET. But it doesn't help the ID cause to look like IDiots.
Comment by Mung — November 13, 2009 @ 4:26 pm
November 13th, 2009 at 4:35 pm
This is the same formulation used throughout the previous thread. However, we agree that selection can purge some, but not all.
Each scenario explicitly stated that every child receives a negative mutation.
The only use of the term "nearly neutral" on this thread is by you. The term I used was "effectively neutral."
You're detouring, and not responding to the points. I've provided an open source model. I've provided cites to scientific journals. I've provided scenarios that directly contradict your position. When you complain about the definition of "harmful" with regards to fitness, I point out that I am using it in the common sense meaning of weakening.
That's why fitness is often explicitly defined in scientific papers. But more importantly, you have made specific claims based on the notion of fitness. Yet here you are now implying the concept is incoherent.
Scientists with their 'incoherent concepts' make specific empirical predictions. Why were the estimates of the human mutation rate based on the hypothesis of common ancestry and neutral theory so predictive of the direct measurements? Lucky guess?
Comment by Zachriel — November 13, 2009 @ 4:35 pm
November 13th, 2009 at 4:40 pm
That's correct. The idea is that new deleterious mutations are added over time. A simple example is Muller's Ratchet. The Ratchet applies to asexual non-recombining genomes. Most neutral or slightly deleterious mutations will never reach fixation and will eventually disappear. But occasionally, through chance alone, such a mutation may become fixed. And being fixed, being shared by all members of the population, there is no selective disadvantage. And then another and another.
Muller's Ratchet assumes there are no back mutations, which become more likely as the number of targets increase. There are many other complications, as well.
Comment by Zachriel — November 13, 2009 @ 4:40 pm
November 13th, 2009 at 5:09 pm
To frame the problem of the fitness metric in more familiar terms, consider an irreducibly complex 2-binding protein-protein interaction.
Say we need 10 nucleotides to be mutated for the binding system to be functional. Each mutation in and of itself is neutral and thus invisible to selection.
Conceptually speaking, the creature that has 8 of the 10 correct nucleotides in place is concptually more "fit" toward the emergence of this new feature, but in Darwinian terms, it is no more fit than the creatures with less of the pre-cursors in place. The mutations are neutral with respect to reproductive fitness, but not with respect to potential function.
Hence the notion of reproductive fitness is is not at all aligned with the conceptual notion of functionality in the case of such irreducible complex systems.
There is a comparable situation with irreducible catastrophe: numerous mutations might have to be in place for catastrope to occur, but the precursor mutations might be neutral in and of themseleves.
Zach seems unwilling to acknowledge this complication. He gives the strong impression that if something is immediately neutral it cannot eventually be harmful. At least that is what his argument imply independent of whether he wishes to concede the mistake or not.
Comment by Salvador T. Cordova — November 13, 2009 @ 5:09 pm
November 13th, 2009 at 5:22 pm
Where again is the scenario that a population purges it's harmful mutations so there is a net decrease in harmfuls over time when the premise is that every new born human introduces a novel mutation (novel as in his parents don't have the mutation).
Let's be specific Zach using your open source model which you say contradicts my thesis.
Every newborn child introduces 1 new harmful which his parents don't have. How many harmfuls on average per human does your open source model say each human will have after 100 generations.
Don't accuse me of ignoring your model. I'm now querrying you on a specific output of your model which could supposedly overturn my claim.
So how many harmfuls are there per individual after 100 generations given the parameter I've provided?
Provide for the readers an answer based on this model you claim refutes my position. If you can't do it for 100 generations, do it for however many you can run, but provide some specifics, not evasions and distortions and misdirections.
How many harmfuls per individual after 100 generations. Put whatever parameters you want with the only constraint being each child has a novel mutation (one the parents don't have).
So how many are their Zach. You accuse me of ignoring your model. On the contrary, I want to know what number your model comes up with.
Comment by Salvador T. Cordova — November 13, 2009 @ 5:22 pm
November 13th, 2009 at 5:22 pm
Mung:
As a genetic based explanation for extinction I see nothing wrong with the hypothesis when applied on a species by species basis.
Comment by Bradford — November 13, 2009 @ 5:22 pm
November 13th, 2009 at 5:30 pm
To give a scope of the difficulty given Sternberg's hypothesis of 90-100% functionality:
If there are 75,000,000 nucleotides of difference between humans and the last common ancestor with the chimp, and we assume that approximately all these are functional (per Sternberg's thesis), and one argues that selection created it, then one must account for how selection could put 75,000,000 nucleotides in place.
Given that selection is having a hard time seeing the compromise of even 100 mutations in functional positions per human per generation, does it seem beleiveable that 75,000,000 nucleotides could be created via selection? This is prima facie evidence that the selection coefficients are not anywhere near substantial enough as supposed.
Although not a formal disproof of selection, I think the consideration I put forward is reason to be skeptical on circumstantial grounds.
I don't know how to go about making the argument formal. I leave that for future investigation.
Comment by Salvador T. Cordova — November 13, 2009 @ 5:30 pm
November 13th, 2009 at 5:33 pm
Conceptually? That's an odd definition of fitness.
When we say they are selectively neutral, that means we can *predict* their behavior over time. And we can and have tested those predictions. Some evolution is clearly neutral. Biologists are also quite aware that neutral evolution can enable adaptive evolution.
Why were the estimates of the human mutation rate based on the hypothesis of common ancestry and neutral theory so predictive of the direct measurements? Lucky guess?
Comment by Zachriel — November 13, 2009 @ 5:33 pm
November 13th, 2009 at 6:17 pm
I don't know what you think this demonstrates. Are you saying humans and some other species evolved from a common ancestor, and that the human lineage did so by neutral mutations?
What about the non-humans? Did they also diverge from the common ancestor via neutral mutations?
One would think the human ancestor was very much like a human then, I guess.
Comment by Mung — November 13, 2009 @ 6:17 pm
November 13th, 2009 at 6:29 pm
I find it quite fascinating that in none of the previous comments has anyone mentioned two very important concepts: Mendelian dominance and diploidy. Unless these are included in calculations of the effects of mutation, no realistic analysis is possible.
All of the comments posted so far assume that deleterious mutations will be “visible” to selection, and that non-deleterious mutations will not be (and will necessarily persist in populations indefinitely. However, even severely deleterious mutations will not be visible at all unless they are dominant (or at least exhibit some dominant “penetrance”). If they are recessive (and show no penetrance) then they will not be eliminated from populations except when they appear in homozygous recessives.
For example, consider the empirical fact that the allele for cystic fibrosis (call it "f") is extraordinarily deleterious and recessive, whereas the F allele is both normal and dominant with virtually 100% penetrance. The f allele almost always results in death when homozygous (usually in childhood), but has no immediately obvious phenotypic effect when heterozygous. I know this for a fact, as my wife is heterozygous for the cystic fibrosis allele (i.e. she is Ff, whereas I am FF, yet she shows absolutely no phenotypic signs of having any of the symptoms of cystic fibrosis).
We know (as the result of molecular genetic analysis) that there are multiple versions of the f allele, all of which have the effect of disabling the chloride protein ion channels in animal cell membranes. People who are homozygous FF and Ff can make all the normal (i.e. functional) chloride channels they need. Only ff individuals cannot do this, and hence cannot regulate the transport of chloride ions across cell membranes.
Each of the versions of the f allele (and again there are quite a few) have apparently arisen separately from each other in several different human populations at different times. The frequency of the f allele among people whose ancestors are from Europe (i.e. Caucasians) is approximately one in twenty (or about 5%).
So, how quickly should the cystic fibrosis allele be removed from the human population, and is there some equilibrium value at which the frequency of the cystic fibrosis allele is so low that it only ever appears in heterozygotes (i.e. its frequency is so low that the probability of two heterozygotes interbreeding and having 1/4 of their offspring expressing cystic fibrosis is effectively zero)?
The answer always surprises my students: let's take just Europeans. Even though the allele is present in 5% of the population, the probability that two people who carry the allele will mate and produce children is only 1/20 X 1/20 = 1/400. Furthermore, the probability that two heterozygotes will have a child that has cycstic fibrosis is only 1/4 (i.e. 1/4 of their children will be homozygous normal and 1/2 = 2/4 will be heterozygous carriers). Ergo, the probability that two Europeans (chosen at random) will have a child that has cystic fibrosis is 1/20 X 1/20 X 1/4 = 1/1,600. This is why almost every case of cystic fibrosis in the United States is the very first appearance of this condition among either of the families of the parents of the affected child.
But that's the theoretical calculation. What is the actual frequency of cystic fibrosis? In 1997, about 1 in 3,300 Caucasian children in the United States was born with cystic fibrosis. In contrast, only 1 in 15,000 African American children suffered from cystic fibrosis, and in Asian Americans the rate was even lower at 1 in 32,000.
Why is the actual frequency of cystic fibrosis so much lower than the predicted frequency? Because the actual frequency of the cystic fibrosis allele varies from subpopulation to subpopulation. Indeed, if one does a calculation based on the assumption that the f allele is a complete recessive lethal allele, and the F allele is both completely dominant and has no deleterious effects, then the frequency of the f allele should be less than 1/10,000, and therefore the actual rate of appearance of cystic fibrosis should be 1/40,000 (i.e. 1/10,000 X 1/4). This calculation is based on the assumption that every single individual who is ff dies before reproducing, whereas every single individual who is FF and Ff dies of something else (i.e. unrelated to the presence or absence of the F allele).
So the interesting question is not “why is the frequency of the f allele not zero”, but rather “why is the frequency of the f allele so anomalously high among Europeans (but not African Americans or Asians, who still exhibit a frequency of the f allele that is slightly higher than that calculated on first principles alone).
The answer seems to be that, like the allele for sickle-cell anemia, the allele for cystic fibrosis increases the fitness of heterozygotes, relative to homozygous normal FF. Using standard population genetics calculations, it is possible to calculate the degree of this increased fitness and then factor it into the forgoing calculations. Intense research is now being conducted to determine precisely how the anomalously high frequency of the f allele is being maintained in some (but not all) human populations. Four hypotheses have been advanced for this, all based on the assumption that being Ff confers some resistance to at least one of the following diseases:
• Cholera: With the discovery that cholera toxin requires normal host chloride transport proteins to function properly, it was hypothesized that carriers of the mutant f allele benefited from resistance to cholera and other causes of diarrhea. Further studies have not confirmed this hypothesis.
• Typhoid: Normal chloride transport proteins are also essential for the entry of Salmonella typhi into cells, suggesting that carriers of mutant chloride transport genes might be resistant to typhoid fever. No in vivo study has yet confirmed this. In both cases, the low level of cystic fibrosis outside of Europe, in places where both cholera and typhoid fever are endemic, is not immediately explicable.
• Diarrhea: It has also been hypothesized that the prevalence of cystic fibrosis in Europe might be connected with the development of cattle domestication. In this hypothesis, carriers of a single mutant chloride transport chromosome had some protection from diarrhea caused by lactose intolerance, prior to the appearance of the mutations that caused lactose tolerance.
• Tuberculosis: Poolman and Galvani from Yale University have added another possible explanation – that carriers of the f allele have some resistance to TB.
My money is on the diarrhea hypothesis, as there is good evidence that the f allele first appeared among Europeans about 52,000 years ago (based on the neutral mutation rate of “silent” base-pairs in the f allele).
In the context of this thread, cystic fibrosis is as lethal as a mutation gets, yet its frequency has neither caused humans (including Europeans) to go extinct, nor has it been removed from the collective human genome. How might this eventually occur? The probability of the f allele persisting in the human genome is an equilibrium between the rate of its removal (via both expression among homozygous ff individuals and pure, random accidental disappearance as the result of heterozygotes failing to pass on the allele for reasons unrelated to its phenotypic effect) and the rate of its preservation as the result of the increase fitness of heterozygotes. It is virtually impossible for the f allele to cause humans to go extinct nor for the allele to completely disappear as long as the effective breeding population of humans remains relatively high (this also calculable using standard population genetics, but I won’t go into it here). However, if the effective breeding population declines enough, then inbreeding effects begin to cause the f allele to appear much more often as ff homozygotes. However, this has the effect of removing the f allele from the population, and if the population is small enough, it can completely disappear (this is the so-called “Sewall Wright effect”, which is now usually referred to as “genetic drift”).
Ergo, if effective breeding populations fluctuate in size enough that they get so small that deleterious alleles can disappear from them by accident as the result of genetic drift, then the Muller/Kondrashov/Sanford problem of “genetic load” (i.e. the accumulation of deleterious alleles, which appear only in heterozygotes) goes away.
Comment by Allen_MacNeill — November 13, 2009 @ 6:29 pm
November 13th, 2009 at 6:51 pm
Heck, if there were that many mutations, why not assume that they were all "essentially neutral" (according to Zachriel's thesis) and try explaining that!
Comment by Mung — November 13, 2009 @ 6:51 pm
November 13th, 2009 at 6:54 pm
Common descent is among the assumptions. Nachman measured the differences in pseudogenes between humans and chimpanzees, and the inferred mutation rate based on those assumptions was 2.5 x 10^-8 mutations per nucleotide. If the assumptions of common descent and neutral evolution in the pseudogenes were incorrect, then Nachman's result would be a number pulled out of thin air. Instead, we have a direct measure by Xue that confirms this rate.
Nachman & Crowell, Estimate of the Mutation Rate per Nucleotide in Humans, Genetics 2000.
Xue et al., Human Y Chromosome Base-Substitution Mutation Rate Measured by Direct Sequencing in a Deep-Rooting Pedigree, Current Biology 2009.
Not all mutations are neutral, of course. Selection is important to adaptation.
Comment by Zachriel — November 13, 2009 @ 6:54 pm
November 13th, 2009 at 7:05 pm
Quite so. Effective population size is often overlooked in these discussions. Not every member of a population breeds. In addition, this number can be subject to significant change over time, creating oscillations of intense then relaxed selection. This has important effects on evolution.
Comment by Zachriel — November 13, 2009 @ 7:05 pm
November 13th, 2009 at 7:07 pm
Allen:
Me:
Comment by Mung — November 13, 2009 @ 7:07 pm
November 13th, 2009 at 7:16 pm
This is pretty irrelevant. Every one that does breed has mutations, and all the members in the next generation are, I think it is safe to assume, the result of breeding. (None of them got there via a parent that didn't breed.)
Comment by Mung — November 13, 2009 @ 7:16 pm
November 13th, 2009 at 7:26 pm
Allen MacNeill:
That's the advantage of an open forum Allen. It gives you the chance to point out the missing pieces.
I did refer to the issue you described albeit not nearly as eloquently.
(snipping some informative paragraphs):
I'll call this the flushing hypothesis; a way of explaining how deleterious alleles can disappear. Two points. Do time frames enable us to effectively assess whether flushing neutralizes novel genetic defects? This process would take place possibly over eons of time so how do we test this?
Second, is there any reason to think that genomic equilibrium, i.e. a balance between the generation of deleterious mutations and their flushing, is an expected result of natural selection? Is the answer a no brainer when competing species stand ready to occupy a niche left open by genetically based extinctions?
Comment by Bradford — November 13, 2009 @ 7:26 pm
November 13th, 2009 at 10:35 pm
Is not extinction just as (more) likely in such a case?
peace
Comment by fifth monarchy man — November 13, 2009 @ 10:35 pm
November 14th, 2009 at 1:49 am
Zachriel:
So one study (Nachman) looked at pseudogenes in two species. Is it safe to assume that they were operating on the assumption that the pseudogenes they chose for their study were homologous? Where they using genees from the Y chromosome? Then they used those genes to come up with the composition of those genes in a hypothetical common ancestor. One assumes these were also pseudogenes in the common ancestor then, unaffected by selection? Then they used those hypothetical pseudgenes from the hypothetical ancestor to estimate a mutation rate, based upon an assumption that the changes were neutral. And how many hypothetical generations were they estimating between the hypothetical ancestor and the present?
That about got it? I could certainly have missed some assumptions.
And the Xue study, it also used pseudogenes, non-coding regions, etc., that were not subject to selection? And it covered how many generations?
Do you really think you have something definitive upon which to base a conclusion?
Comment by Mung — November 14, 2009 @ 1:49 am
November 14th, 2009 at 9:42 am
We have the mutation rate inferred from the posited common ancestry of chimpanzees and humans ~250,000 generations ago and that of known humans 13 generations ago. If you read the Nachman study, they discuss the variance based on different estimates of generation time, age of divergence, etc. If any of the assumptions had been incorrect, then we wouldn't expect a close match between independently derived measures.
The result is also consistent with another direct measure, the known error rate for DNA replication of 10^-10 (Drake et al. 1998, Tago et al. 2005). (There are about 400 replications between the human male zygote and sperm.) The mutation rate in humans has been measured consistently for generations by many different techniques (Haldane 1935).
It's not just a number pulled out of a hat. There is high congruence between the molecular evidence and unrelated fossil evidence which represents strong confirmation of common descent. (Not to mention the nested hierarchy.)
Comment by Zachriel — November 14, 2009 @ 9:42 am
November 14th, 2009 at 10:31 am
Thank you much for weighing in
My understanding is that bad recessives make the problem of bad genes persisting monstorously worse.
A killer recessive could float around for many generations and never be visible to selection, especially if there is limited net inbreeding.
My understanding is that animal breeders use inbreeding to purge such recessives out.
Genetics and Eugenics: A Textbook for Biology Students
But if a sufficient number of loci in humans have extremely bad recessives, even a vigorous program of inbreeding might still be infeasible (not to mention society would never accept breeding humans like breeding lab mice). The reason it would be infeasible (even on the unrealistic assumption society would even attempt a program of eugenics) is that the reproduction rate in humans is too small.
The central question is what is the tolerable harmful mutation rate in humans vs. the actual rate of harmfuls. Muller said the tolerable rate is 0.1.
I maintain that a rate of 1.0 harmful is for all practical purposes irreversible.
By harmful, I do not mean necessarily with respect to immediate reproductive fitness, but rather to function.
That is correct, and for the sake of argument, I have said:
So, I'm presuming a specific harmful eventually leaves the population via drift or selection.
The problem however is that for every one that disappears, more are added. I postulate this would be the situation with even 1 harmful per new born human being added.
If Dad has a bad mutation A and Mom have a bad mutation B, there is a chance some of the kids will not have any of mutation A or B simply by mendalian inheritance. Unfortunately, I don't have the formula to describe the exact outcome. Nachman and Crowel allude to it by saying with U=3, then 2 out of 40 kids will be mutation free.
But if Mom and Dad on average are not pumping out 40 kids (in utero selection for mildly harmfuls is awfully presumptuous), and then on top of that new mutations are being added (I postulate 1 new harmful per new born), we have an irreversible deterioration.
Nachman and Crowell postulate "positive epistasis" which means, as far as I understand, somehow magically the bad mutations in combination somehow become beneficial!
Dave Wisker pointed out a study by Lynch where there was supposed evidence of "compensatory mutations", but I objected and claimed this interpretation is suspect, even assuming the lab results are accurate. Why?
For example, Microspordia are supposedly very fit in the Darwinian sense, but have suffered enormous loss of function over time. If such a situation happened with humans, it would amount to saying as long as humans keep reproducing it is irrelevant how many organs or genetic diseases are floating around, they are "fit" in the reproductive sense. Thus Lynch's idea of compensatory mutations (from a population standpoint and prevention of extinction) is not necessarily what I had in mind as far as the problem of genetic deterioration.
These questions are important to the
1. creation vs evolution debate
2. ID vs Darwinism debate
3. neturalists vs. selectionist debate
4. the mutationists vs. the selectionist debate
But the main reason I think the question will persist and be explored is that it has central significance to medical understanding of the future human condition.
Thus, I don't think the work of anyone on the question genetic accounting should be dismissed, rather it should be vigorously studied and researched.
I must sadly, at this time, agree with geneticist Bryan Sykes of Oxford who was pioneer research on Y-chromosomal Adam and Mitochondrial Eve, the human race is vulnerable to extinction (100,000 years is his estimate). And if not extinction, some horrible condition ( as illustrated by microspordia).
Comment by Salvador T. Cordova — November 14, 2009 @ 10:31 am
November 14th, 2009 at 10:39 am
After all these comments neither Zachriel nor Dr MacNeill hass provided any scientific data which would show us that mutations can accumalate to BUILD new useful molecular machinery and bring about new body plans.
Everything with Common Descent relies on the assumption that it happened.
Nothing in biology says that the transitions are even possible.
Comment by ID guy — November 14, 2009 @ 10:39 am
November 14th, 2009 at 11:01 am
Yes, there will always be a certain level of deleterious mutations lurking as recessives. But that's not your claim. Your claim is the inevitability of genomic meltdown. That's not born out by the data.
You keep citing Sykes, but his contention only concerns the largely non-recombining y-chromosome, and even then he has failed to convince most of his colleagues (and he has backed off somewhat). There are other mechanisms at work, and ultimately the y-chromosome could pick up additional material from elsewhere in the genome, or transfer its sex determinant responsibility to other chromosomes. This may or may not result in speciation, and it doesn't necessarily mean the end of the lineage. In any case, the evidence indicates that the human y-chromosome still has the same functional genes since its divergence from chimpanzees about 5-6 million years ago.
Page et al., Conservation of Y-linked genes during human evolution revealed by comparative sequencing in chimpanzee, Nature 2005.
It's so odd to see your reliance on authority when you think they support your position, but ignore the vast majority of biologists who strongly support the Theory of Evolution.
Comment by Zachriel — November 14, 2009 @ 11:01 am
November 14th, 2009 at 11:26 am
Except for the fact that there isn't any scientific data which supports the transition.
You keep ignoring that. Interesting…
Science is not done via majority.
And not one of your alleged majority can support the claims made by the theory.
That you continue to ignore that says quite a bit about your agenda.
Comment by ID guy — November 14, 2009 @ 11:26 am
November 14th, 2009 at 11:43 am
Still waiting for your answer on your runs of the open source model. How many mutations on averege per person after 50 generations assuming 1 new harmful per new born? Use reasonable estimates for the other paramters, but the constraining parameter must be 1 new harmful per new born.
Or did you not really run it with that parameter?
Comment by Salvador T. Cordova — November 14, 2009 @ 11:43 am
November 14th, 2009 at 12:32 pm
The animation is now available (thanks to my IDCS Network partner in crime who goes by the handle "g-man". ) This now results in very fast loading and viewing.
http://www.youtube.com/watch?v...
Thanks to Zachriel and Allen for weighing in.
The arguments have refined my thinking on the matter and will result to amendments to the animation with appropriate qualifications and descriptions. There will be future animations and improvements to this one.
On the assumption of 1 new mutation per new born, in future animations, we can even incorporate Zach's figure for the number of mutations each inividual will carry after fifty generations.
But that will depend on how quickly Zach can pull that figure out of the open source model he claims refutes my conclusions.
Comment by Salvador T. Cordova — November 14, 2009 @ 12:32 pm
November 14th, 2009 at 12:37 pm
Gee whiz.
Nick even provided you a recent study showing the limitations of current models. Yet you claim that your simplistic model can overthrow the entire corpus of evolutionary biology.
Gregor's Bookkeeper is a qualitative model. It is sufficient only to understand and debunk certain "common sense" beliefs about evolutionary systems, such your simplistic scenario above.
Did you ever attempt an answer to this: Why were the estimates of the human mutation rate based on the hypothesis of the common ancestry of humans and chimpanzees so predictive of the direct measurements?
Comment by Zachriel — November 14, 2009 @ 12:37 pm
November 14th, 2009 at 1:04 pm
I'm sorry, in all your misdirections and evasions, I seem to have missed the number that your computer run of the open source model returned. Was the number
1?
2?
3?
45?
50?
A simple number will do. You complained I'm ignoring your open source model. Can you tell me what number outputted with respect to the queston at hand. I'd be happy to report at Uncommon Descent the results.
You claimed the data refutes my assertion. So how far was I off Zach. It would be helfpul if you can give a ball park estimate based on the open source model you claim I'm ignoring. Im not ignoring it. I'm wanting to know what numbers it is reporting so I can compare it to my own estimates.
Simply saying "Salvador is wrong" is not very helpful. I get told that by you all the time. I'd like to know what the ball park right answer is. Assuming 1 harmful per new born human, after fifty generations, what is the number of mutations existing in each individual?
Is it:
1?
2?
3?
….
45?
….
50?
For me to be wrong the number would have to be less than one. I have to say, fat chance of it being less than one unless you're invoking black magic.
But for the sake of everyone here, tell us what you think the number is.
To say "Salvador is positively wrong, but there is no accurate answer" sounds a bit like double speak. You could give us all a clue by giving an estimate of the number of mutations per individual after 50 generations on the assumption that 1 new harmful is introduced per new born.
I already pointed out, we can assume each harmful never reaches fixation. That absence of fixation does not solve the complication that I'm asserting.
On the assumption of 1 new harmful per new born is introduced, are you saying I would be wrong to claim after 50 generations there will be on average more than 1 harmful in each individual?
What does your open source model say will be the ball park average?
0 (possible if we invoke black magic)
.5 (possible if we invoke black magic)
1.0 (only if truncations selection is practiced in the wild and if there is sufficent reproductive excess — none of which are realistic assumptions)
1.00000000001 (oops, that means guaranteed deterioration)
1.00000000001 , 2 , 3, etc (then Salvador is right)
So how about it Zach? Are you so confident that after fifty generations on the assumption of 1 new harmful per new born that there will be less than 1 harmful per person after 50 generations? That's what it would take to arrest genetic deterioration.
Comment by Salvador T. Cordova — November 14, 2009 @ 1:04 pm
November 14th, 2009 at 1:20 pm
No. For the umpteenth time, I said your claim was unfounded.
Gregor's is an abstract, qualitative model. Do you not understand what the word means?
Comment by Zachriel — November 14, 2009 @ 1:20 pm
November 14th, 2009 at 1:32 pm
It depends on how harmful. It also depends on whether there are favorable mutations as well. If the mutations are neutral, or nearly so, they will fix at the rate of mutation, or nearly so.
If we have several slightly deleterious mutations to a gene (say -1%), then a significant favorable mutation (say +10%), could the gene have improved function?
Comment by Zachriel — November 14, 2009 @ 1:32 pm
November 14th, 2009 at 2:15 pm
How so. If the removal rate of harmfuls is not 1 or more per human on average, genetic deterioration won't be arrested. That is basic logic.
Care to provide the removal rate in numbers. Are you saying the average removal rate is greater than 1 per human (which if true would be nonsense, but I just want to make sure you're committed to defending such nonsense since you keep saying my claim is unfounded).
What does your open source model say will happen again after 50 generations? How many mutations per human after 50 generations. Provde the number, and then I'll show you how to calculate the predicted removal rate.
But until you cough some numbers from you open source model, it doesn't seem you have good reason to say my claim is unfounded.
Comment by Salvador T. Cordova — November 14, 2009 @ 2:15 pm
November 14th, 2009 at 2:22 pm
No, the essential conclusion does not depend on how harmful nor whether the harmfuls reach fixation. I already said for the sake of argument, we can assume no harmfuls reach fixation. The animation makes clear we can select away the most harmful or the least harfmul, the deterioration still happens.
The degree of harmfulness might only affects the severity of deterioration, it doesn't stop it from being inevitable, which is the central issue.
The presence of a harmful in the population is not the same as the harmful being fixed in the population (fixed = present in every individual).
Consider the 10,000 or so rare genetic diseases in the article I mentioned in another thread. The disease is not fixed in the population but it is present. The net result is the human genome, on average, is less functional than if these diseases didn't exist at all.
Comment by Salvador T. Cordova — November 14, 2009 @ 2:22 pm
November 14th, 2009 at 4:45 pm
But you haven't shown that mutations aren't removed. They are. Some are removed by simple drift, for instance. Others are removed by selection. The only question is the rate, which depends on a number of variables. Nor have you accounted for beneficial mutations and a number of other mechanisms.
If we have several slightly deleterious mutations to a gene (say -1%), then a significant favorable mutation (say +10%), could the gene have improved function?
It depends on that, it depends on the distribution of beneficial mutations, it depends on population. It depends on lots of things.
Why were the estimates of the human mutation rate based on the hypothesis of the common ancestry of humans and chimpanzees so predictive of the direct measurements?
When we started this conversation, I posted a significant amount of data as a starting point for discussion. Instead of looking at the data and seeing if it was relevant, you refused to even consider it. I've asked a number of questions, which you have consistently neglected to answer. It's your claim, but you want me to prove you wrong rather than you defending your own claim. I just can't see most readers giving that sort of 'argument' any currency.
Comment by Zachriel — November 14, 2009 @ 4:45 pm
November 14th, 2009 at 5:16 pm
A very large population and high fecundity. Every child receives negative mutations. Homozygotes for any trait are stillborn. The strongest eat the weakest. This will act as a direct check on the number of negative mutations. Recombination ensures that healthy children will be produced in each generation. As parents can't be homozygous for any trait, some children will be clear for that trait. Even if the health of the population slips, it can recover in the next generation, as long as population and fecundity are high.
This case is an exaggeration, but it's at the root of your misunderstanding. Once you understand that the one harmful mutation per individual per generation leads to genomic meltdown isn't an immutable law of logic, then we can consider more reasonable cases.
Comment by Zachriel — November 14, 2009 @ 5:16 pm
November 14th, 2009 at 5:28 pm
Consider these two parents. They have all sorts of negative mutations, indeed every gene has a damaged allele. (1 is a negative mutation, 0 is clear.)
#100111001110001 ...
11000110001110 ...
#200111011111111 ...
11000100000000 ...
Recombination will ensure that at least some of the genes will be clear in at least some of the children. Fitness improves without any positive mutations. Now their children will compete for the next generation, with only those with the lowest number of harmful mutations becoming parents.
Comment by Zachriel — November 14, 2009 @ 5:28 pm
November 14th, 2009 at 9:09 pm
Sal, you haven't made any conclusions (Hand-waving doesn't count.)
Comment by Mung — November 14, 2009 @ 9:09 pm
November 14th, 2009 at 9:14 pm
If Muller's Ratchet does not appear to be a problem for asexual species, why should we believe that something similar operates in sexual species and that it does become an issue for those species?
Comment by Mung — November 14, 2009 @ 9:14 pm
November 14th, 2009 at 10:11 pm
Muller's ratchet will apply to large linkage blocks in the genome and the Y-chromosome of humans, the Y-chromosome and linkage blocks exist in sexually reproducing species like humans.
Comment by Salvador T. Cordova — November 14, 2009 @ 10:11 pm
November 14th, 2009 at 10:13 pm
Zach,
What's the number again from your run of the open source model. I missed your answer.
I disagree with that characterization. But for the sake of argument, let's assume I didn't make any conclusions. Perhaps Zach can offer some insights.
After 50 generations, how many harmfuls on average will exist in each human given 1 new harmful per new born? What does Zach's run of the open source model predict? Any number greater than 1.0 (like say 1.0000001) would confirm the hypothesis I put forward.
I notice Zach seems awfully reluctant to provide results of the open source model he swears by.
Sal
Comment by Salvador T. Cordova — November 14, 2009 @ 10:13 pm
November 15th, 2009 at 3:08 pm
Sal wrote:
I agree that Müller's ratchet should apply to alleles located in the y chromosome of mammals (as well as to alleles in the chromosomes of haploid eukaryotes and the DNA of prokaryotes and viruses), but I do not see how it could apply to alleles located in "large linkage blocks" or anything else located in the autosomes or X chromosome of mammals or other eukaryotes. In my understanding, "large linkage blocks" are just another name for chromosomes. Indeed, that's how they were first defined in the early 20th century, before the congruence between linkage groups and chromosomes was empirically verified in the early 20th century by Bridges, Stephens, and Sturtevant.
If you disagree, please explain how Müller's ratchet might apply to alleles located in the autosomes or X chromosome of mammals (or provide a link to such an explanation). Thanks!
Comment by Allen_MacNeill — November 15, 2009 @ 3:08 pm
November 15th, 2009 at 9:20 pm
Sal:
For your model to be a reasonable representation of biological reality it must also take into account a variant of natural selection that can remove deleterious alleles from gene pools when effective breeding population sizes become very small. This form of natural selection is often referred to as either “negative selection” or “purging selection”. Here is a definition (from Wikipedia) with references, followed by three empirical studies showing that negative/purging selection does in fact take place:
NEGATIVE SELECTION:
Negative selection, in natural selection, is the selective removal of alleles that are deleterious. This can result in stabilizing selection through the purging of deleterious variations that arise. It is also known as purifying selection.
Purging of deleterious alleles can be achieved on the population genetics level, with as little as a single point mutation being the unit of selection. In such a case, individuals bearing the allele selected against might simply have less offspring on average generation after generation.
In the case of strong negative selection on a locus, the purging of deleterious variants will result in the occasional removal of linked variation, producing a decrease in the level of variation surrounding the locus under selection. The accidental purging of non-deleterious alleles due to such spatial proximity to deleterious alleles is called background selection.[1] This effect increases with higher mutation rate but decreases with higher recombination rate.[2]
References:
1. Charlesworth, B., Morgan, M. T. and Charlesworth, D. 1993. The effect of deleterious mutations on neutral molecular variation. Genetics 134, 1289-1303. Link
2. Hudson RR, Kaplan NL (December 1995). "Deleterious background selection with recombination". Genetics 141 (4): 1605–17. PMID 8601498.
http://en.wikipedia.org/wiki/N...
************
How are deleterious mutations purged? Drift versus nonrandom mating.
Glémin S.
Evolution. 2003 Dec;57(12):2678-87.
Accumulation of deleterious mutations has important consequences for the evolution of mating systems and the persistence of small populations. It is well established that consanguineous mating can purge a part of the mutation load and that lethal mutations can also be purged in small populations. However, the efficiency of purging in natural populations, due to either consanguineous mating or to reduced population size, has been questioned. Consequences of consanguineous mating systems and small population size are often equated under "inbreeding" because both increase homozygosity, and selection is though to be more efficient against homozygous deleterious alleles. I show that two processes of purging that I call "purging by drift" and "purging by nonrandom mating" have to be distinguished. Conditions under which the two ways of purging are effective are derived. Nonrandom mating can purge deleterious mutations regardless of their dominance level, whereas only highly recessive mutations can be purged by drift. Both types of purging are limited by population size, and sharp thresholds separate domains where purging is either effective or not. The limitations derived here on the efficiency of purging are compatible with some experimental studies. Implications of these results for conservation and evolution of mating systems are discussed.
http://www.ncbi.nlm.nih.gov/pu...
************
Testing alternative methods for purging genetic load using the housefly (Musca domestica L.).
Meffert LM, Regan JL, Hicks SK, Mukana N, Day SB.
Genetica. 2006 Sep-Nov;128(1-3):419-27.
When a population faces long-term inbreeding, artificial selection, in principle, can enhance natural selection processes for purging the exposed genetic load. However, strong purge pressures might actually decrease fitness through the inadvertent fixation of deleterious alleles and allelic combinations. We tested lines of the housefly (Musca domestica L.) for the effectiveness of artificial selection to promote the adaptation to small population size. Specifically, replicate populations were held at average census sizes of 54 for nine generations or 30 for 14 generations while being subjected to artificial selection pressure for increased fitness in overall mating propensity (i.e., the proportion of virgin male-female pairs initiating copulation within 30 min), while also undergoing selection to create differences among lines in multivariate components of courtship performance. In the 14-generation experiment, a subset of the lines were derived from a founder-flush population (i.e., derived from three male-female pairs). In both experiments, we also maintained parallel non-selection lines to assess the potential for natural purging through serial inbreeding alone. Sub-populations derived from a stock newly derived from the wild responded to artificial selection for increased mating propensity, but only in the short-term, with eventual rebounds back to the original levels. Serial inbreeding in these lines simply reduced mating propensity. In sub-populations derived from the same base population, but 36 generations later, both artificial selection and serial inbreeding increased mating propensity, but mainly to restore the level found upon establishment in the laboratory. Founder-flush lines responded as well as the non-bottlenecked controls, so we base our major conclusions on the comparisons between fresh-caught and long-term laboratory stocks. We suggest that the effectiveness of the alternative purge protocols depended upon the amount of genetic load already exposed, such that prolonged periods of relaxed or altered selection pressures of the laboratory rendered a population more responsive to purging protocols.
http://www.ncbi.nlm.nih.gov/pu...
************
Purging of inbreeding depression within the Irish Holstein-Friesian population
Sinéad Mc Parland, Francis Kearney and Donagh P Berry
Genetics Selection Evolution 2009, 41:16doi:10.1186/1297-9686-41-16
The objective of this study was to investigate whether inbreeding depression in milk production or fertility performance has been partially purged due to selection within the Irish Holstein-Friesian population. Classical, ancestral (i.e., the inbreeding of an individual's ancestors according to two different formulae) and new inbreeding coefficients (i.e., part of the classical inbreeding coefficient that is not accounted for by ancestral inbreeding) were computed for all animals. The effect of each coefficient on 305-day milk, fat and protein yield as well as calving interval, age at first calving and survival to second lactation was investigated. Ancestral inbreeding accounting for all common ancestors in the pedigree had a positive effect on 305-day milk and protein yield, increasing yields by 4.85 kg and 0.12 kg, respectively. However, ancestral inbreeding accounting only for those common ancestors, which contribute to the classical inbreeding coefficient had a negative effect on all milk production traits decreasing 305-day milk, fat and protein yields by -8.85 kg, -0.53 kg and -0.33 kg, respectively. Classical, ancestral and new inbreeding generally had a detrimental effect on fertility and survival traits. From this study, it appears that Irish Holstein-Friesians have purged some of their genetic load for milk production through many years of selection based on production alone, while fertility, which has been less intensely selected for in the population demonstrates no evidence of purging.
http://www.gsejournal.org/cont...
Comment by Allen_MacNeill — November 15, 2009 @ 9:20 pm
November 15th, 2009 at 9:57 pm
Me:
Salvador:
It's not a characterization.
So please tell us, try to be as brief as you can and still convey what precisely your conclusion is.
Perhaps I'm construing conclusion in too narrow a sense, so let us know what you mean by conclusion as well.
As near as I can tell, you've proffered an animation, and come to a "conclusion" of "this is a problem for evolution."
I don't think that qualifies as a conclusion because it doesn't follow from any premises. But granted, I may have misunderstood.
Comment by Mung — November 15, 2009 @ 9:57 pm
November 15th, 2009 at 10:02 pm
Allen:
Mitochondrial DNA?
Comment by Mung — November 15, 2009 @ 10:02 pm
November 15th, 2009 at 10:41 pm
Mung:
Yep, mitochondrial DNA, too, and also chloroplasts. In my (lame) defense, I usually think of mitochondria and chloroplasts as modified prokaryotes, as they almost certainly evolved from prokaryotic ancestors via serial endosymbiosis (as proposed by Copeland and expanded by Margulis).
Comment by Allen_MacNeill — November 15, 2009 @ 10:41 pm
November 15th, 2009 at 10:45 pm
Which raises an interesting point: since the mitochondria in every multicellular eukaryote are virtually always inherited via the maternal line (from mitochondria contained in the maternal egg cell), it would seem likely that there would be a "founder-flush" event every generation, as the sample of mitochondria in each egg cell would be an almost infinitesimal fraction of the total mitochondrial population of the eukaryotic mother (who made the egg cells). If this were indeed the case, it would explain why there is an unusually low frequency of deleterious mutations in mitochondria (with the exception of some rare forms of muscular dystrophy).
Comment by Allen_MacNeill — November 15, 2009 @ 10:45 pm
November 16th, 2009 at 2:09 am
I thought mitochondrial DNA because eit seems directly relevant to the human evolution scenario questioned by the OP (however inadequately).
But I am surprised to find you a supporter of the endosymbiotic theory. Is there a "best" candidate for the prokaryote ancestor? Isn't "prokaryote-first" old school?
Perhaps this could be taken up in another thread, as it would seem to be relevant to the front-loading hypothesis as well.
Comment by Mung — November 16, 2009 @ 2:09 am
November 16th, 2009 at 11:53 am
Okay. I made a few modifications to Gregor's Bookkeeper, Gregor's Teeth.
An original isogenic population with a fitness of one. Mutation is considered damage to an allele. If both alleles are damaged, then the fitness for the entire organism is zero. Otherwise, we can adjust the effect. With a slight deleterious effect of 0.1 per gene, the population is more stable than if the damage is completely hidden.
Here is a typical scenario, each with a stable population after a hundred generations.
Population = 100Offspring per individual = 2
Recessive = -0.1
Average Fitness = 0.62
Population = 100Offspring per individual = 1.5
Recessive = -0.1
Average Fitness = 0.41
Population = 200Offspring per individual = 1.5
Recessive = -0.1
Average Fitness = 0.53
No beneficial mutations, no variance, no phylogenetic noise. Doesn't change the overall result anyway.
Population = 100Offspring per individual = 2
Recessive = -0.0
Average Fitness = 0.58
With silent recessives (i.e. Recessive = -0.0), the Offsprings have to be at least ~1.9 per individual for a stable population.
Comment by Zachriel — November 16, 2009 @ 11:53 am
November 16th, 2009 at 11:53 am
Okay. I made a few modifications to Gregor's Bookkeeper, Gregor's Teeth.
An original isogenic population with a fitness of one. Mutation is considered damage to an allele. If both alleles are damaged, then the fitness for the entire organism is zero. Otherwise, we can adjust the effect. With a slight deleterious effect of 0.1 per gene, the population is more stable than if the damage is completely hidden.
Here is a typical scenario, each with a stable population after a hundred generations.
Population = 100Offspring per individual = 2
Recessive = -0.1
Average Fitness = 0.62
Population = 100Offspring per individual = 1.5
Recessive = -0.1
Average Fitness = 0.41
Population = 200Offspring per individual = 1.5
Recessive = -0.1
Average Fitness = 0.53
No beneficial mutations, no variance, no phylogenetic noise. Doesn't change the overall result anyway.
Population = 100Offspring per individual = 2
Recessive = -0.0
Average Fitness = 0.58
With silent recessives (i.e. Recessive = -0.0), the Offsprings have to be at least ~1.9 per individual for a stable population.
Comment by Zachriel — November 16, 2009 @ 11:53 am
November 16th, 2009 at 12:08 pm
Pushing the limit a bit on fecundity.
Population = 100Offspring per individual = 1.25
Recessive = -0.2
Average Fitness = 0.31
Comment by Zachriel — November 16, 2009 @ 12:08 pm
November 16th, 2009 at 12:27 pm
As interesting and important as it is for an amateur unencumbered with training or experience to reveal a flaw at the heart of the field of genetics, I'm looking for the telic angle. What is the purposeful mechanism which cleans up these mutations? I understand this is a recent proposal, so I'm not expecting the hundred years' worth of evidence we're discarding based on the logic that it couldn't possibly be accurate. Just a description of one directed mechanism and a test to confirm it, hopefully something that confirms it as telic, and not simply a non-telic mechanism that has been overlooked until now. Thanks.
Comment by don provan — November 16, 2009 @ 12:27 pm
November 16th, 2009 at 4:29 pm
Interesting comparison, after 1000 generations. (The previous examples used 20 genes. This is with 100 genes.)
Population = 100
Offspring per individual = 2.5
Recessive = 0
Average Fitness = 0.58
Average Fitness refers to average fitness of the children, so of 250 children, 58% have a fitness of one, more than enough to replenish the population of 100. The remaining 42% have a fitness of zero.
Population = 100
Offspring per individual = 2.5
Recessive = -0.001
Average Fitness = 0.98
The majority of children have a fitness between 0.993 and 0.999. Only a handful were stillborn.
Notice the large difference in average fitness due to the very slight effect of the recessive allele.
-
"Ah, I don't have to outrun the bear, dear Albert," said Niels. "I only have to outrun you."
Comment by Zachriel — November 16, 2009 @ 4:29 pm
November 16th, 2009 at 4:32 pm
NO. It's not even necessarily true when all mutations are deleterious and there are no beneficial mutations. It does depend.
Comment by Zachriel — November 16, 2009 @ 4:32 pm
November 16th, 2009 at 9:00 pm
The fact that a deleterious or neutral allele is linked to another allele in a “large linkage block” (i.e. in the same chromosome without an intervening crossover chiasma) doesn’t really apply here. Once again, it matters if the allele is recessive or dominant (and if dominant, what its degree of penetrance is). Furthermore, if a deleterious or neutral allele is linked to an allele that confers increased fitness (i.e. those individuals who have it survive and pass it on more often than other individuals who do not have it), then the increase in frequency of the beneficial allele will “drag” the deleterious and neutral allele frequencies along with it, so long as the benefit of the one allele outweighs the deleterious effect of the other.
This phenomenon is similar to genetic drift, and has hence been referred to as “genetic draft” by John Gillespie, the population geneticist who discovered it (based on a suggestion from Will Provine). That is, the slightly deleterious or neutral alleles will “draft” along with the beneficial allele(s) so long as they are all linked in the same non-recombining chromosome.
And again, if the deleterious allele is recessive, but beneficial when heterozygous, then your model is completely out to lunch (yum, yum, gingerbread men for dessert
.
Comment by Allen_MacNeill — November 16, 2009 @ 9:00 pm
November 17th, 2009 at 6:03 pm
I must have missed it:
What is the evidence that mutations can collect in a genome and lead to new useful molecular machinery and body parts?
Also do asked for a test for directed processes, but how can we test for non-telic processes?
Zachriel has already admitted that position is not scientific.
Comment by ID guy — November 17, 2009 @ 6:03 pm
November 17th, 2009 at 6:05 pm
Salvador et al:
This entire discussion of “genetic entropy” took place almost three quarters of a century ago, in response to Hermann Müller’s presentation of the phenomenon now commonly referred to as “Müller’s Ratchet”. John Sanford’s GE model is explicitly modeled on Müller’s Ratchet, but does not include the very phenomenon that Müller proposed as the solution to his dilemma: sexual reproduction via diploidy, meiosis, and gamete/zygote fertilization.
Müller pointed out that being diploid and exchanging genetic material would completely negate the negative effects of the accumulation of deleterious mutations, both by compensating for such mutations via their homologs and by eventually weeding them out by exposing them to selection following recombination.
This is very old stuff, folks. Once again, ID supporters are re-fighting the ancient battles that eventually culminated in the formulation of the “modern evolutionary synthesis”, while most evolutionary biologists have moved on during the subsequent half a century.
Comment by Allen_MacNeill — November 17, 2009 @ 6:05 pm
November 17th, 2009 at 6:16 pm
Allen, your "move along, folks – nothing to see here" attitude reflects exactly the myopic dogma that Woese was railing against.
Comment by chunkdz — November 17, 2009 @ 6:16 pm
November 17th, 2009 at 6:26 pm
There is more on genetic entropy. Biologists have uncovered data in recent years that could not even have been considered by analysts 75 years ago. I'll make that very clear in a follow-up entry.
Comment by Bradford — November 17, 2009 @ 6:26 pm
November 17th, 2009 at 7:11 pm
Mendel's Accountant
Comment by Mung — November 17, 2009 @ 7:11 pm
November 18th, 2009 at 2:21 am
Apologies for the delay in responding, I was mostly out of pocket for a coulple days.
Before I respond, I want to make clear I'm sensitive to those with hereditary diseases as I know people personally who suffer these issues. I don't wish to make light of their condition, but to appreciate the above statement, I will add a little satire:
The phrase "beneficial" when heterozygous is misleading. I have objected vigorously to characterizing "reproductively advantaged" as necessarily "beneficial".
I don't expect somone to be especially thrilled to discover they have certain heterozygous "benefical" traits.
The Darwinian view and the medical view of what is "beneficial" are occasionally at odds. Moalem called the situation, "Survival of the Sickest".
If the question is whether selection can create designs, this would suggest selection is blind, to use Alan Orr's words, selection can just as well lay waste to design. It behaves myopically, it doesn't have foresight.
Calling cystic fibrosis heterozygously "beneficial" somewhat distorts notions of what is functional and healthy.
Selection helps perpetuate a hereditary disease by preventing the nullizygous condition (the healthy condition) from being recognized. Selection thus helps to lock in the sickness and further population resources are consumed eliminating the homozygous form.
My understanding was their was the presumption that the high degree of polymoprhism was once believed to be due to selection. That would have led to presumption of high levels heterozygous advantage. It was concluded, that such a scenario would cause too much segregational load (getting rid of the homozygous forms). The Neutralists won the debate. Selection (an adapatationist story) was not responsible for the large degrees of polymorphsim.
So not only would heterozygosity impede progress toward the healthy form, it will cause segregational load.
Heterozygous advantage would not be a point against genetic deterioration, but rather for it since it is an example of seleciton helping infuse hereditary disease (such as sickle cell anemia).
Comment by Salvador T. Cordova — November 18, 2009 @ 2:21 am
November 18th, 2009 at 2:26 am
Zach,
What is the run if the population reproduces asexaully or we are dealing with essentially a haploid model (like the Y-chormosome).
What is the situation then.
The animation, if you didn't notice should be modeled as haploid.
Sal
Comment by Salvador T. Cordova — November 18, 2009 @ 2:26 am
November 18th, 2009 at 2:27 am
How can that be if they are presumed to have 1 new harmful, or are you pulling a renormalization trick?
Comment by Salvador T. Cordova — November 18, 2009 @ 2:27 am
November 18th, 2009 at 3:33 am
If the corresponding linkage blocks on all for chomosomes (2 for mom, 2 for dad), have at least one harmful each (and over time this is possible), since each linkage block is damaged and since all for blocks are damaged, the offspring will inherit them, and then this will be analogous to Muller's Ratchet in the asexual case, except to say it happens slower.
For selection to work, it must be able to see particulate entitities. Lumping entities together like this will foil the benefit of recombination, even though recombination helps.
I think the scenario I painted is what Sanford had in mind, although that is only a guess on my part.
Comment by Salvador T. Cordova — November 18, 2009 @ 3:33 am
November 18th, 2009 at 3:43 am
Here is another scenario, the advantaged linkage block with the bad mutation occupies all the forms, hence is fixed into the population (homozygous all the way, so to speak). This guarantees the bad mutation is dragged into every individual. Recombination doesn't flushout the problem since recombination does not have sufficient resolution due to linkage blocks.
Comment by Salvador T. Cordova — November 18, 2009 @ 3:43 am
November 18th, 2009 at 8:15 am
Alleen MacNeill,
Hopefully those evolutionary biologists moved on to find mutations that can collect to build useful things.
However they don't appear to have found any yet.
What's the hold-up?
Comment by ID guy — November 18, 2009 @ 8:15 am
November 18th, 2009 at 8:33 am
Everyone knows that gingerbread people include gingerbread men, gingerbread women and gingerbread babies.
You didn't provide a model. and there was a lot of discussion about human reproduction. Even bacteria often exchange genes. And I provided multiple examples and a detailed model that used recombination, yet you didn't make that point until now. Your argument was clearly provincial (human-centric).
If the mutation is a silent recessive, per the example, then the fitness of that individual is one (though the trait may show in their children).
In any case, one mutation per offspring does not inevitably lead to genomic meltdown in recombining genomes. It is clear that the parameters do matter.
Comment by Zachriel — November 18, 2009 @ 8:33 am
November 18th, 2009 at 9:15 am
Most likely because recombination is a design mechanism for preventing such a thing.
Again I refer to Dr Spetner's "Not By Chance".
Comment by ID guy — November 18, 2009 @ 9:15 am
November 18th, 2009 at 12:51 pm
BTW, in this week’s issue of the Proceedings of the National Academy of Sciences (see http://www.pnas.org/content/ea... ), Peter and Rosemary Grant report the origin of a new species of finch (according to the biological species concept) on Daphne Major in the Galapagos:
http://evolutionlist.blogspot....
Furthermore, the Grants’ observations provide empirical support for my hypothesis for the origin of species via first-degree inbreeding (see http://evolutionlist.blogspot.... ).
Happy Thanksgiving!
Comment by Allen_MacNeill — November 18, 2009 @ 12:51 pm
November 18th, 2009 at 1:48 pm
Allen,
This appears to be no more than variation within a Kind.
IOW it appears to support Linne's PoV that the Kind was not the species.
Rather the species evolved from the originally Created Kind.
Comment by ID guy — November 18, 2009 @ 1:48 pm
November 18th, 2009 at 2:16 pm
I wondered how long it would take for a creationist to move the goalposts…
So, what is a "kind", anyway? What taxonomic level does it correspond to (if any) — genus, family, order, class, phylum, kingdom, domain — and by what empirically verifiable criteria does one determine this?
Keeping in mind, of course, that some plants (e.g. orchids) readily hybridize across families (i.e. above the level of genera), and there is very solid evidence that whole genome fusion has produced new taxa at the domain level (i.e. the highest possible taxonomic grouping).
Comment by Allen_MacNeill — November 18, 2009 @ 2:16 pm
November 18th, 2009 at 2:21 pm
Methinks it is like a weasel.
Comment by Mung — November 18, 2009 @ 2:21 pm
November 18th, 2009 at 2:29 pm
Comment by Mung — November 18, 2009 @ 2:29 pm
November 18th, 2009 at 2:46 pm
And out come the labels and stereotypes…
The first casualty of culture war is critical thinking.
Comment by chunkdz — November 18, 2009 @ 2:46 pm
November 18th, 2009 at 2:52 pm
Comment by Mung — November 18, 2009 @ 2:52 pm
November 18th, 2009 at 2:56 pm
Mung:
Go here for my answer to your question and my take on the "problem" of species:
http://evolutionlist.blogspot....
Comment by Allen_MacNeill — November 18, 2009 @ 2:56 pm
November 18th, 2009 at 2:58 pm
chunkdz:
I didn't bring up creationist "kind", ID Guy did. Indeed, ID Guy even provided a reference to the creationist whose ideas were being cited. So, under these circumstances, should we all conclude that the undefined taxonomic category "kinds" is not a creationist trope?
Comment by Allen_MacNeill — November 18, 2009 @ 2:58 pm
November 18th, 2009 at 3:36 pm
I'm not interested in "he started it" arguments. I'm more concerned that you admittedly came into the conversation under the closed minded presumption that your opponents are deceitful.
You enter this conversation with such a defensive culture warrior posture that you reduce Carl Linnaeus to the label "creationist", and even referencing the ideas of the father of modern taxonomy is considered "creationist trope".
What hope does critical thinking have in such an atmosphere of labels, stereotypes, and warrior posturing?
Comment by chunkdz — November 18, 2009 @ 3:36 pm
November 18th, 2009 at 6:19 pm
Allen:
I didn't see an answer to my question, and it's not at all clear where you come down on the issue.
Let's say that species only exist in our minds. Is it your contention that our minds do not exist in the natural world?
Is there a reason that we are talking about the natural world rather than the real world, given that the claim is "there are really no such thing as species at all"? What if species exist independently of the material world?
Comment by Mung — November 18, 2009 @ 6:19 pm
November 18th, 2009 at 7:50 pm
Species is a term that has a lot of definitions, but in essence, we *observe* that populations of organisms exhibit traits that are consistent over generations (Ray 1686). From this, various utilitarian definitions and theoretical explanations have been developed.
Cladistic, biological, ecological, evolutionary, taxonomic, and many others. Keep in mind that species are not necessarily distinct categories, but may have chaotic boundaries. You might check out John Wilkins' tome, Species: A history of the idea.
Comment by Zachriel — November 18, 2009 @ 7:50 pm
November 18th, 2009 at 8:09 pm
Whatever you want it to be.
Comment by chunkdz — November 18, 2009 @ 8:09 pm
November 18th, 2009 at 8:17 pm
Mung: So what is a species anyways?
chunkdz: Whatever you want it to be.
Me: I want it to be real, not imaginary.
Comment by Mung — November 18, 2009 @ 8:17 pm
November 18th, 2009 at 8:18 pm
I'd say that a "kind" would be defined as a class of organisms that exist within evolutionary limits. IOW, "kinds" only evolve so far – a cat does not evolve into a dog. I remember reading an article some time ago that postulated that if you cross bred all of the "cat" species (lions, tigers, bobcats, house cats, etc.) you'd eventually reach the original cat "type" (or kind).
Experiment: breeding, induced mutation, etc. See how far evolution can take you in the real world – as opposed to the conjecture driven realm of the past.
Comment by Daniel Smith — November 18, 2009 @ 8:18 pm
November 18th, 2009 at 8:22 pm
From Allen's blog:
"discrete, objective entities that represent reproductively independent lineages or 'units of evolution'."
Comment by Mung — November 18, 2009 @ 8:22 pm
November 18th, 2009 at 8:39 pm
It's as real as any other idea.
Comment by chunkdz — November 18, 2009 @ 8:39 pm
November 18th, 2009 at 8:47 pm
Here is a short list of criteria that have been used to define species (there are more):
Morphology (i.e. appearance): Classically, this what has been used most often to define species. However, what if two individuals that have been classified as members of different species are capable of interbreeding? This has happened repeatedly, as naturalists have learned more about organisms in nature.
Reproductive Incompatibility: As has been noted previously, this is the most commonly used criterion today. However, there are organisms that appear virtually identical, yet are reproductively sterile if they mate. And, conversely, there are groups of organisms that are all reproductively compatible, yet include individuals with widely divergent morphologies.
Behavioral Isolation: Some biologists have proposed that species should be distinguished primarily by whether or not individuals recognize the mating displays and/or signals of other individuals; if they do, they are members of the same species. However, this definition is obviously very limited; how would one apply it to plants, or fungi, or bacteria, or any other organism that doesn't actively display and/or mate with another organism?
Niche Exclusitivity: In ecology, the niche of an organism is the totality of its interactions with the various components of its environment; a species can therefore be defined as a group of organisms inhabiting essentially the same niche. This concept is reinforced by Gause's Law: two species cannot simultaneously inhabit the same niche in the same ecosystem. However, if one uses this definition of species, one is using an essentially circular definition: a species is what inhabits a certain niche, which is defined by the species which inhabits it, etc. ad infinitum… Furthermore, it is very problematic to define the boundaries of a niche in nature; again, the usual way to do this is to inventory all of the interactions between the members of an already-defined species, and then call that the species' niche – circularity again.
Phylogeny (i.e. shared derived characteristics): This species concept is based on the same criteria as cladistic classification: the absence or presence of shared derived characteristics. However, phylogenetic lines can change without branching; does this mean that the organisms at widely separated parts (i.e. times) of such a line are members of the same species? Furthermore, what if two branches fuse together (i.e. hybridize); how should we classifying the resulting species?
Why are there so many different definitions of what constitutes a "species?" One reason is that each definition fits the "needs" of different groups of evolutionary biologists. For example, the phylogenetic species concept is preferred by systemitists and taxonomists; they generally reject the biological species concept as being unusable for systematic classification.
Again, there are almost as many definitions of what constitutes a species as there are different branches of biological science. However, most of these can be subsumed under one of four broad species definitions:
Biological Species Concept: species are defined by reproductive isolation (i.e. members of different species cannot successfully exchange genetic material). According to this definition, species have reproductive isolating mechanisms which prevent interbreeding. The problem with this definition is that it really only works with animals, and even then not with all of them (it seems to work best with birds). In particular, it is completely useless for defining species of asexually reproducing organisms (of which there are very, very many).
Phenetic Species Concept (also called "Numerical Taxonomy"): species are defined by a quantitative analysis of their characteristics, as determined via comparisons of empirically quantified characteristics. Formerly very popular, this technique relied on computer/numerical analysis of similarities and differences in empirical (i.e. observable) characteristics. However, there is no intrinsic mechanism that defines which characteristics are important (from an evolutionary standpoint) and which are accidental/irrelevant.
Phylogenetic Species Concept (i.e. cladism): Very similar in some ways to the evolutionary species concept, this concept is based on the presence or absence of shared derived characteristics. If two organisms share the "same" set of derived characteristics, they are considered to be members of the same species. However, cladism is not necessarily tied to any form of descent with modification; for example, it can just as easily be used to classify office furniture as living organisms. Furthermore, this concept is highly dependent on detailed information about the characteristics of organisms; these characteristics are quantified, and then the results are used to construct a cladogram, the "fit" of which is either accepted or rejected using statistical analysis. As in all such analyses, if there isn't enough data to perform such an analysis, an organism simply can't be classified. Recent attempts to clear up some of these difficulties using purely molecular genetic analyses have been confounded by the discovery that widely divergent organisms share very similar DNA sequences. Does this mean that they are closely related? Why or why not, especially if one's criterion is only the absence or presence of shared derived characteristics?
Species Recognition Concept: As described earlier, this concept depends on mating behavior – organisms are members of the same species if they recognize each other's mating displays and behaviors, and successfully mate. Again, this is only applicable to animals, and not even to many of them. And, there are many cases in which organisms from widely divergent species will attempt to mate with each other. Males, in particular, are not very "picky" about whom they will mate with, and will attempt to exchange genetic material with members of widely divergent species (or even kingdoms, in the case of orchid wasp-mimics).
The "Species Problem" is that none of these concepts really help in the understanding of species or speciation. As noted above, the definition of what constitutes a species varies from discipline to discipline in biology. However, if you ask the local people in a particular area what species of animals and plants they recognize, they will get very close to the same set of "species " as those recognized by professional systematists. So, are these "real" species?
Darwin (as usual) may have had a more useful definition of what constitutes a species than most modern systematists. According to Darwin,
In other words, the concept of a "species" is a useful fiction. As I pointed out in my blogpost, there is considerably less arbitrariness in what constitutes an individual organism, and it is individuals that live or die, reproduce or fail to do so. It is individuals, in other words, which are the focus of natural selection.
A population is also less ambiguous than a species, and is most precisely defined by phylogeny: a population is a group of organisms that have descended from one or more common ancestors. Sometimes a population is a species: if there is only one breeding population of a species in existence, that population is both a "population" (in the ecological sense of that term) and a "species" in the evolutionary and taxonomic sense of that term.
Personally, I'm not particularly interested in defining what a "species" is or isn't, as I am mostly interesting in the mechanisms by which evolution occurs, and the results of the operation of those mechanisms. I'm not a phylogeneticist, in other words, and find arguments over what constitutes a "species" to be similar to arguments over how many angels can dance on the head of a pin.
Comment by Allen_MacNeill — November 18, 2009 @ 8:47 pm
November 18th, 2009 at 9:03 pm
I am not a Creationist and I didn't move anything.
All I did was make an observation.
Your "soeciation" is nothing more than "variation within a Kind".
And perhaps you should take the time to learn about what your opposition is saying.
That way you can actually address their arguments.
I thought that is what science was for.
Science was required to make the determination of species was it not?
The Current Status of Baraminology-
Anything else?
Comment by ID guy — November 18, 2009 @ 9:03 pm
November 18th, 2009 at 9:04 pm
chunkdz makes a very important point: what constitutes "reality"? If you are a Platonist, ideas (i.e. those concepts that represent generalizations about any consistent pattern of phenomena) are just as "real" as rocks. However, if you are a Baconian, only those generalizations that can be formulated via induction from empirical phenomena are "real"; all other concepts are "imaginary".
Taken to its logical extreme, the Baconian viewpoint becomes nominalism, a metaphysical doctrine that asserts that all perceived generalizations are imaginary: there are no "rocks", there are only individual rocks, which bear no necessary relationship to each other.
Since the empirical sciences — physics, chemistry, geology, astronomy, biology, and (perhaps) psychology — are all ultimately founded on induction, all of the concepts in the empirical (i.e. "natural") sciences are closer to nominalism than to Platonic realism. This is why I tell my students that the empirical sciences aren't about "Truth", if by that term one means generalizations that are absolutely and necessarily "true". Here's part of the lecture I give on this subject in my evolution course:
Biology, like chemistry and physics, is an empirical science: it is based on observations of the real world. There are non-empirical sciences, such as mathematics and symbolic logic, which are formulated without direct observation of the real world. The theories of the empirical sciences, such as biology, are developed using a procedure called the scientific method.
Many people think that the scientific method is similar to magic, or is so difficult to understand and apply that only trained scientists can use it. Nothing could be further from the truth. The scientific method is basically just "common sense" consistently applied. In general, the scientific method consists of six or seven steps, beginning with observations of the real, natural world.
Step One: You observe the world around you, focusing on a particular object or process you find interesting:
• For example, if you have never eaten an apple before, and you encounter a green apple growing on an apple tree, you might be tempted to sample it. You do, and the apple tastes sour.
Step Two: You ask yourself a question about what you have observed:
You are curious as to whether all apples are sour, so you ask yourself the question, "Are all green apples sour?" You try another one, and it's sour too:
• Another one – it's sour as well. Hmmm…
Step Three: You formulate a hypothesis: that is, a tentative guess based on the pattern that you have observed so far:
• Based on the foregoing observations, a reasonable hypothesis about green apples would be:
"Green apples are sour"
There is a term used to describe the kind of reasoning that scientists use to formulate hypotheses: it is called inductive reasoning (or "induction"). Essentially, inductive reasoning is formulating a generalization based on a series of individual cases (i.e. arguing from the particular to the general).
Inductive reasoning is how virtually all human knowledge and understanding begins. Notice two things about any conclusions you might formulate using inductive reasoning:
• The validity of your generalization is only as good as the number of similar observations that you have used to formulate your generalization.
• Regardless of how many observations you may have made, you cannot be absolutely certain that the generalization that you have formulated is universally applicable. This is because you can only observe a small subset of all possible cases of whatever it is you are interested in. After all, no matter how many apples you bite, you might not (yet) have tasted a Granny Smith (a green apple that tastes sweet!)
Up to now, your reasoning processes have not really been any different than what everyone has always done. Even the ancient Greeks used inductive reasoning; what makes the scientific method different is what you do next.
Step Four: You formulate a prediction: that is, a guess about what will happen if you perform another observation in the light of your generalization.
• Your generalization was "Green apples are sour." Therefore, a logical prediction you might make about an apple that you have not yet tasted is:
"This is a green apple; therefore, it is sour."
There is also a term used to describe the kind of reasoning that scientists use to formulate predictions based on hypotheses: it is called deductive reasoning (or "deduction"). Essentially, deductive reasoning is formulating a prediction about a specific case based on a generalization (i.e. arguing from the general to the particular). Deductive reasoning is how many of us make judgements about the objects and processes we see around us.
Two caveats about any predictions you might formulate using deductive reasoning:
• The validity of your prediction is only as good as your generalization. If your generalization is based on a small number of individual observations (especially only one), then it is unlikely to be very useful in making predictions that will be supported by further observation.
• Regardless of how you have formulated your generalization (i.e. your hypothesis), if you do not then test it using one of the methods described below, you haven't really done any science. In fact, you haven't really done anything useful…yet.
Step Five: You test your prediction; that is, make further observations that could either confirm or deny the validity of your hypothesis. There are two somewhat different ways to do this:
• Make some more observations, similar to the ones that led you to formulate your hypothesis in the first place; this is called discovery science. In the case of your green apple hypothesis, this would consist of simply tasting another green apple (or two).
• Perform an experiment. This means performing two kinds of observations: an experimental test, where you manipulate the variable that you are testing, and a control test, where you do not manipulate the same variable. In the case of our green apples, there is no experimental or control test. However, in many tests of biological hypotheses, control tests are used to determine if the variable being manipulated actually affects the outcome.
Note that whichever way you test your prediction (by further observations or by experiment), you are once again using inductive reasoning. This means that all of the conditions listed above still apply:
• The validity of your conclusions is only as good as the number of similar observations that you have used to test your hypothesis.
• Regardless of how many observations you may have made, you cannot be absolutely certain that the hypothesis that you have tested and confirmed is universally applicable.
Step Six: You compare your test results with the prediction that you made using your hypothesis. If the results are pretty close to the ones predicted, then you have confirmed your hypothesis. However, if the results are significantly different from the ones you predicted, you take the next (and in many ways the most important) step:
Step Seven: You modify (or completely reformulate) your hypothesis and repeat all of the steps above.
• So, you try another green apple, and this time you discover it's sweet (it's a Granny Smith)! What do you do? You modify your hypothesis: "Most green apples are sour, except for Granny Smith apples." And then you keep on testing…
In science, a hypothesis that has been repeatedly tested and has not yet been shown to be invalid (i.e. all of it's predictions have worked out so far) is referred to as a theory. Notice that many non-scientists use the word "theory" to mean what a scientist means when s/he uses the word "hypothesis;" that is, a tentative guess about the way the world works, which has not yet been thoroughly tested.
When a scientist uses the word "theory," s/he is generally referring to what a non-scientist would call a scientific law. This difference in usage flows from the tendency of scientists to consider that virtually no scientific principle is ever completely confirmed; it's only as good as the experiments that have been done so far to test it. This means that what scientists refer to as "theories" generally have a great deal of evidence backing them up, more than alternative explanations.
In other words, scientific theories (such as the theory of evolution) are what most non-scientists would refer to as "scientific laws."
One of the most important implications of the foregoing is that nothing is really "true" in science, using the commonly accepted definition of "truth:" that is, always and absolutely "true." All scientific theories are open to revision, and even a cursory look at the history of science indicates that theories that were once considered "true" are now either highly modified or have been thrown out altogether.
So, what is science?
Science is our best current guess at how the universe works, until we find out otherwise.
Comment by Allen_MacNeill — November 18, 2009 @ 9:04 pm
November 18th, 2009 at 9:05 pm
You know if we take Allen's provided definitions of speciesas "gospel" then homosexuals should be considered a different species.
Just sayin'…
Comment by ID guy — November 18, 2009 @ 9:05 pm
November 18th, 2009 at 9:15 pm
It seems to me that "baraminology" is just a modified version of the morphological species concept, with a dash of reproductive isolationism thrown in. That is, a "baramin" is primarily defined by appearance (i.e. morphology) using methods that are partially borrowed from phenetics (i.e. numerical taxonomy). The baramins that are defined by this method are also at least partially defined by reproductive isolation, but only when morphology isn't sufficiently similar to conclude membership in the same baramin (or when successful interbreeding indicates that the boundaries of a baramin are reproductively "permeable").
Ultimately, however, the thing that most characterizes baraminology is that no amount of empirical observation can ever be accepted if it indicates that the baramins are not the ones either described in the Bible or insisted upon by fundamentalist Christian doctrine. For example, if it were possible for humans and bonobos to interbreed (to my knowledge this hasn't been attempted, although sometimes I wonder
), then this observation would nevertheless be twisted to indicate that humans and bonobos are in separate baramins, despite the empirical fact that orchids can successfully interbreed above the genus level (i.e. at the level of families, which is a greater "reproductive distance" than between bonobos and humans).
Comment by Allen_MacNeill — November 18, 2009 @ 9:15 pm
November 18th, 2009 at 9:22 pm
Actually, I know over a dozen lesbian couples in Ithaca alone, and every single one has children; some by artificial insemnation, others by adoption, and still others resulting from heterosexual relationships. I also know at least four gay couples in Ithaca that exhibit the same patterns. Would they therefore be unclassifiable as humans, despite the fact that they clearly have reproduced (although not necessarily consanguiniously)?
Furthermore, using this criterion then any person who has not reproduced (for whatever reason) cannot possibly be classified into a species. This would, of course, include all children under the age of puberty, all celibate individuals, and all people who either choose not to have children or who are physiologically unable to do so. Are toddlers, nuns, and guys who have had the mumps not classifiable as humans?
Last (but certainly not least) if one's criterion for species status were exclusively based on the ability to interbreed and produce fertile offspring, then all asexually reproducing organisms (all bacteria, many protists, many fungi, many plants, and even some animals) would be unclassifiable into species.
See the kinds of knots you get tied into when you assert that reproductive compatibility is the only valid criterion for membership in a species?
Comment by Allen_MacNeill — November 18, 2009 @ 9:22 pm
November 19th, 2009 at 1:14 am
Zachriel asked in another thread about the supposed agreement between the mutation rate inferred between the chimp human split and the direct measurement of mutation rates using solexa.
Let's assume for the sake of argument Zach's view is correct, what is the consequence?
If each new born human is adding 100 nucleotides of mutational differences from his parents, how can "deeply conserved" regions be maintained?
Assuming truncation selection, I don't think you can possibly kill off enough humans to maintain the regions in a "conserved" state for very long.
The conserved regions present an enigma for the idea of purifying selection in the wild:
Mystery DNA
If these regions are found to be non-functional, then how could selection maintain them since it can't see them?
If these regions are functional (as Sternberg hypothesized), then how could selection maintain them since policing 100 mutations per human would be impossible from a resources standpoint. Worse, if they are found functional, then how did they become functional in the first place without the help of selection?
I predict finding using Solexa and Illumina will continue to show that the mainstream evolutionary assumptions must necessarily result in irreconcilable contradictions regarding the ultra conserved regions.
Zach has claimed the 100 new mutations per new born are "neutral". Fine, then how are the deeply conserved regions maintained under such an assualt from mutations?
With enough mutations, the animation will represent the fundamental problem confronting the maintenance of ultra conserved regions.
It is clear with haploids, 1 harmful per new born will result in deterioration, I don't know yet what a good number is for diploids like humans. I guess the number is between 3 and 10. Muller thinks as little as 0.1.
This is worth more investigation.
Comment by Salvador T. Cordova — November 19, 2009 @ 1:14 am
November 19th, 2009 at 8:20 am
Ultimately however the one thing that characterizes the theory of evolution is no amount of evidence can ever make it go away.
I mean here it is 150 years after CD published "On The Origin of Species" and we still don't have any evidence for mutations that build useful structures and new body parts.
Do you have a reading problem Allen?
I never asserted that reproductive compatability is the only criteria.
However you are proving that your criteria are pretty worthless…
Comment by ID guy — November 19, 2009 @ 8:20 am
November 19th, 2009 at 8:58 am
Most of those mutations are neutral. Selection is sufficient to maintain deeply conserved regions of the genome. It depends—as you should now know—on a number of factors.
Orthodox theory predicts that conserved regions are functional.
Not all of them. Some are neutral. Some are nearly neutral. Some are not neutral at all. And conserved doesn't necessarily mean exact identity.
Your other concerns have ready answers. Conserved regions of the genome are predicted to have a function. That no function has been found for some regions may simply indicate human ignorance. Consider that aliens can (have?) lobotomize humans in laboratory situations and the humans will reproduce just fine. However, that doesn't mean that lobotomized humans "in the wild" are just as well-adapted. Or lack of known function could indicate a problem with evolutionary theory, but that just isn't clear at this point.
Comment by Zachriel — November 19, 2009 @ 8:58 am
November 19th, 2009 at 9:28 am
By design, of course.
Comment by ID guy — November 19, 2009 @ 9:28 am
November 19th, 2009 at 2:40 pm
I have to wonder if your students don't come away quite confused about both science and truth and the relationship between the two, and possibly even quite mistaken about the two to the point of repeating the mantra "science isn't about truth" without even understanding it.
Can science truly function as science of there are not some things which are absolutely and necessarily true?
Do you think it might be lost on some of them that you are referring only to some generalizations from observed phenomena?
And as such, your point about science and truth is really quite trivial, so why do you bother?
Comment by Mung — November 19, 2009 @ 2:40 pm
November 19th, 2009 at 7:28 pm
Weren't Aristotle's arguments based on observations of the real world? And take the triangle as an example, can we not observe triangles in the real world? So why limit mathematics to the purely non-empirical?
If mathematics is so darned non-empirical, how do you explain it's ability to describe the real world, and how do you explain our abilities for mathematics? Yes, even mathematics is a biologal, therefore empirical, problem.
Comment by Mung — November 19, 2009 @ 7:28 pm
November 19th, 2009 at 8:09 pm
Hey isn't that Sal's argument?
Just change a couple terms and…
"Consider that mutation can lobotomize humans yet the humans will reproduce just fine. However, that doesn't mean that lobotomized humans 'in the wild' are just as well-adapted."
IOW, reproductive fitness does not equal overall fitness.
You mean that non-coding parts of the genome might not actually be junk?
I doubt anything could convince you there's a problem with evolutionary theory Zach.
Comment by Daniel Smith — November 19, 2009 @ 8:09 pm
November 19th, 2009 at 8:33 pm
Selection pressures on humans in an alien lab are different from selection pressures on humans "in the wild." The question at issue is ultra-conserved regions of mammalian genomes. Being ultra-conserved the Theory of Evolution predicts they'll have a function for mammals in the wild. Deletion of these ultra-conserved regions in mice doesn't seem to affect their fitness. This could mean that test does not clearly relate to their fitness in the wild. Or it could mean a problem with evolutionary theory. It's too soon to tell.
That's been known for a long time.
Of course there is. We just talked about one such instance. Indeed, the Theory of Evolution has been modified in the light of new evidence many times.
Comment by Zachriel — November 19, 2009 @ 8:33 pm
November 19th, 2009 at 10:55 pm
Being ultra-conserved the Theory of Common Design predicts they'll have a function for mammals in the wild.
Comment by ID guy — November 19, 2009 @ 10:55 pm
November 20th, 2009 at 1:12 pm
Not really. That inference is made by presuming that function is visible to selection, therefore selection preserves "conserved" regions for hundreds of millions of years. How then do we knock out a region (like in those mice), the organism lives, is not affected, and yet the regions is still functional. That is clear evidence selection couldn't enforce conservation of these regions (exactly as Denton alluded to).
That is the problem with defining function in terms of immediate fitness and reproductive fitness.
Spare tires have function, but knocking them out of a car doesn't affect immediate operation of an otherwise undamaged car. Same with airbags.
That's why knockout is a clumsy way to identify design of contingency systems. Biology has large amounts of self-healing. Self-healing and contingency would imply lots of systems that can be knocked out without immediate effects.
That is a better view point! That is the better way that design can help biological investigation versus clumsy knockout. This is the hypothsized use of steganography: the designer created apparently "conserved" regions to help identify contingency designs which when knocked out have little effect on immediate fitness, but have effect on contingency function (like spare tires and airbags).
Comment by Salvador T. Cordova — November 20, 2009 @ 1:12 pm
November 20th, 2009 at 1:32 pm
The Theory of Evolution predicts that conserved regions are preserved by selection.
Yes, that's the question. I gave two possible answers above. 1) The experiment is not a valid test of function in the wild. 2) There is a problem with the theory that makes such a prediction.
Function isn't defined in terms of fitness. Evolutionary fitness concerns differential reproduction.
Yes, that's one reason why knock-out experiments are not considered conclusive. Redundancy can have evolutionary benefit.
Comment by Zachriel — November 20, 2009 @ 1:32 pm
November 20th, 2009 at 2:26 pm
Daniel Smith:
What do you mean by "overall fitness," and how do you define it, if not be reproductive success? Fitness is always context dependent.
Has anyone ever claimed that all non-coding parts of the genome are junk?
Here's a question that might help to clarify the issue of "junk" DNA. If the entire genome is functional, would one expect genome size to correlate with functional complexity? In other words, do the most complex organisms have the largest genomes?
Comment by Nick — November 20, 2009 @ 2:26 pm
November 20th, 2009 at 2:28 pm
No, no, the theory of common design predicts that they are functionless artistic flourishes that the designer finds particularly appealing.
Comment by Nick — November 20, 2009 @ 2:28 pm
November 20th, 2009 at 2:30 pm
Absolutely not.
Comment by Bradford — November 20, 2009 @ 2:30 pm
November 20th, 2009 at 2:31 pm
Bradford,
Can you clarify? If genome size is not correlated with functional complexity, then what design principles explain genome size?
In computer programs written by a single designer, does one normally find that the more complex programs are longer or shorter than the simple ones?
Comment by Nick — November 20, 2009 @ 2:31 pm
November 20th, 2009 at 2:36 pm
It does not have to be bigger to be smarter or more economical. That seems to be the lesson of ncRNA. The increased complexity is found in regulatory elements. But even when focused on protein coding genes we find splicing as evidence that the same can yield multiple transcripts. Smart and economical.
Comment by Bradford — November 20, 2009 @ 2:36 pm
November 20th, 2009 at 2:58 pm
So, if I understand you correctly, some genomes exhibit smarter or more efficient design than others? It strikes me as somewhat odd that, for example, pufferfish would exhibit much better design than zebrafish. Other than raw genome size, is there any other evidence that puffers have smarter design?
Comment by Nick — November 20, 2009 @ 2:58 pm
November 20th, 2009 at 3:28 pm
Nick, I think it presumptuous to assert that the DNA of one species is better designed than some other. There are too many unknowns. Up to now, most mapping projects have been focused on protein-coding sequences. Yet ncRNA regulatory circuits are key components of complex genetic phenomena in eukaryotes which could be the distinguishing marker of better design.
Comment by Bradford — November 20, 2009 @ 3:28 pm
November 20th, 2009 at 4:15 pm
Bradford,
OK, forget better. Let's focus on the words that you used: smart and economical. Given two eukaryotes of roughly equivalent complexity, if their genome size differs dramatically AND the entire genome is functional, then wouldn't it be a logical conclusion that one genome is more economical or "smarter" than the other?
Observation 1: The haploid zebrafish genome contains approximately 1.5 billion basepairs. The haploid pufferfish genome contains contains approximately 400 million basepairs.
Observation 2: Zebrafish and pufferfish exhibit roughly the same functional complexity. Do you agree or disagree? I mean, if we can't say that two teleost fishes are roughly equal in complexity, then I'd think that any design inference based on perceived complexity is problematic.
One way to explain these two observations is, as you suggest, that pufferfish have very smart and economical genome design. The obvious corrolary is that zebrafish are designed less smartly. That conclusion doesn't depend on any specific knowledge of ncRNA or other ways that genome function is regulated, but one would probably expect to find dramatic differences the way genes are regulated in puffers and zebrafish.
An alternative explanation is that the genomes of pufferfish and zebrafish differ in the amount of "cruft," regions of the genome with little or no function that have been lost in pufferfish. One would probably expect that gene regulation and ncRNA content would be very similar in the two species.
If the former explanation is correct, then it becomes an interesting question why puffers exhibit smarter design, but my money is on the latter.
Comment by Nick — November 20, 2009 @ 4:15 pm
November 20th, 2009 at 4:18 pm
It presumes selection can preserve functional regions. That is only true in some cases not all, maybe in less than 10% of the case based on Kimura's work and Sternberg's hypothesis.
90% or more of genome is not subject to selection (Kimura)
90% or more of genome is functional (Sternberg)
Ergo: majority of functions (say 90% * 90% = 81%) are not visible to selection. Thus to presume selection would preserve consderved regions is likely to be wrong about 81% (who knows the exact number) of the time.
But not enough to be of much visibility to selection.
Andreas Wagner points out drift mechanism are more potent than selection acting on contingency designs which means selection doesn't for practical reasons see them (except in the case where the redundancy is activated due to rare circumstances).
There is a cost issue associated with purification selection just like the cost issues of maintaning a car. Enough variation and insufficient population resources result in the inability to keep up the maintenance requirements.
Walter ReMine says 10 mutations with truncation selection (the best possible in theory) will result in the inability to keep up with maintenance costs in human diploid populations. Muller said, 0.1, so ReMine is actually being more conservative than Nobel Laureate Muller by a factor of 100!
Comment by Salvador T. Cordova — November 20, 2009 @ 4:18 pm
November 20th, 2009 at 4:30 pm
There is the basic theory of common design.
There is a more radical version of common design. Dembski's Steganography:
I've said before, ID could be advanced because of the profit motive. Searching and interpreting conserved regions could mean $$$. Same for junk DNA, where many treasures may reside. Many secondary designs there. Potential for $$$.
I've said the stalemate in ID/Darwinism debate might become broken once the ID viewpoint implies biotech making more money and more effectively curing diseases.
The line of investigation that the animation represents is along those lines. If there is genetic deterioration, the consequence is that selection did not create these secondary designs. If it did not create these secondary designs, it might remotely be possible that the designer created them for our benefit to understand the world. Recall, that was the premise of Privileged Planet. Thus, the secondary design might make medical advancement possible.
First, we must establish that there is secondary design independent of selection in the first place. The research into purifying selection (symbolized by the animation, and research done by ReMine and Sanford, not me) is the first step.
Don't underestimate the implicaitons of ENCODE. ID proponents and creationists will have more to say about this in the coming decade. The secondary designs are so obvious, it is only a matter of time before it become widely recognized.
These are exactly the sort of designs that have no selective purpose but do have purpose if the Intelligent Designer not only made a Privileged Planet fine tuned for scientific discovery, but a biotic reality even more optimized for biological discovery.
Comment by Salvador T. Cordova — November 20, 2009 @ 4:30 pm
November 20th, 2009 at 5:09 pm
I would not bet against you. Your fish point is analogous to your single designer computer programmer. A toddler scribbling some indecipherable nonsense in the margin of a paper could extend the analogy. But you are not arguing that the amount of DNA in a genome correlates to an organism's complexity and do acknowledge that regulatory elements can explain complexity correct? If so then genomic size is not helpful in assessing the quality of design except perhaps with regard to very similar organisms.
Very dissimilar organisms might be more analogous to two different computer languages.
Comment by Bradford — November 20, 2009 @ 5:09 pm
November 20th, 2009 at 5:57 pm
Bradford:
That's right. It's clear that genome size does not correlate with complexity, and I do acknowledge that regulatory elements play an important role in complexity. But that isn't really the point I was attempting to make.
As a followup to Daniel's comment at 8:09 pm about non-coding sequence and junk I was suggesting that despite all the new data on ncRNA and other conserved non-coding sequences, there is still reason to think that large chunks of the non-coding genome of vertebrates have little or no function. One way to get a handle on the issue is to compare genome size in similar species like pufferfish and zebrafish.
As you point out with your scribbling toddler analogy, this isn't an issue of design vs. non-design, because it is not necessary to assume that design is economical or smart. But either way, the zebrafish does seem to have a lot of "stuff" in its genome that the puffer doesn't need.
Comment by Nick — November 20, 2009 @ 5:57 pm
November 20th, 2009 at 6:08 pm
Some clarifications and corrections are in order. What I post is related to the discussion at UD and TT regarding the animation.
1. The animation was a haploid model. I could have been clearer about that. Under the haploid assumptions, the conclusion is correct if the assumptions are true.
2. The diploid model may have some analogs to the complication presented. Jistak mentioned at UD that there will be a level of mutation under which even the diploid model will go toward deterioration.
3. The rate of mutation in the diploid is more sensitive to modeling parameters than the haploid model on the assumption of 1 harmful per new born on average.
Nachman suggests that 3 new harmfuls per individual (U=3) would lead to meltdown without synergistic epistasis. ReMine says 10 new harmfuls per individual would be sufficient in any conceivable diploid model. The way to visiualize this is that the kids will inherit on average at least 1 harmful from mom and dad.
I re-iterate again, "harmful" in the functional or design sense is not always the same as "deleterious" in the population biology sense where everything is measured by immediate success toward reproduction.
The harm can be defined with respect to the deterioration of secondary designs and change in functional areas of the DNA, particularly those previously called junk.
Harm to these regions can be empirically determined by tracking SNP's using Solexa and Illumina technology.
Critics of my animation like Mung and Jistak etc. said there was no model of fecundity, selection coefficients, viability. I said that was moot.
That criticism is partially correct and partially incorrect. The model is moot if we are dealing strictly with haploid (the animation), but it is not moot if we are dealing with diploids and relatively low harmful rates.
By "low", I mean such rates where viability, fecundity, will result in potential arrest of genetic deterioration. As I said, even 10 might be enough independent of viability or seleciton coefficients assuming mainstream values of human fecundity.
But what I find troubling is that Nachman and Kondrashov don't provide numbers either to argue for the existence of "synergistic epistasis" in sufficient degree to arrest deterorioration. The knockout experiments on mice should raise serious doubts that sufficient amounts of "synergistic epistasis" really exist. I don't hear a lot of complaints about their lack of modeling parameters.
Doesn't matter in the long run for at least two reasons:
1. we may know the answer empirically over time as there is more DNA sequencing being done
2. there may be a point in the diploid case where most of the modeling parameter become moot. The number is 1 new harmful per individual in the haploid case, it may be 3-10 in the diploid case. If there is not synergistic epistasis and the mutations are slightly deleterious, then Nachman established it would be around 3. ReMine speculate 10 is the max number even if there is synergistic epistasis.
Comment by Salvador T. Cordova — November 20, 2009 @ 6:08 pm
November 20th, 2009 at 6:15 pm
First, let's resolve your previous misstatement.
Of course it's an orthodox prediction of the Theory of Evolution.
And that's why you point to something that would seemingly contradict the prediction.
What specific empirical research has Sternberg done to justify his claim?
That is simply incorrect. You should know by now that it depends! For instance, if the chance of a fatal catastrophe is high, then the value of redundancy may far outweigh the cost.
Comment by Zachriel — November 20, 2009 @ 6:15 pm
November 20th, 2009 at 6:21 pm
What are the good reasons?
1. broken function (I'll buy that)
2. Darwinism and evolutionary theories (bad reasons)
Even if common descent is true, it does not necessarily imply DNA is junk.
Proteins don't just need to be defined by genes. Proteins need to be regulated, controlled, and developed. I don't see how merely coding and producing proteins will make an organism.
Anything that is self healing and robust will have extraodinarily deep redundancy and fail safe systems. Thus, knockout experiments are horrible approaches to seeing these sorts of function since redundant systems are designed to allow function even when systems are knocked out and broken!!!!!
No need to argue much over these topics. The medical community will want to know the real answers about junk DNA. If there is function there that will translate into developing cures and $$$, the argument will be settled.
Sal
Comment by Salvador T. Cordova — November 20, 2009 @ 6:21 pm
November 20th, 2009 at 6:29 pm
Nick:
If the stuff are transposons you can infer what you wish from a design perspective unless you focus on something very specific. Maybe you're looking at a reservoir of "adaptation material" or perhaps truly metabolically wasteful junk. This is subjective without a context.
Comment by Bradford — November 20, 2009 @ 6:29 pm
November 20th, 2009 at 7:36 pm
Not exactly right Sal. I wasn't trying to get you to model those things in your animation. I understood that there was no underlying model, that it was simply an animation, and I made that point from almost my first question on the matter, where I asked about the underlying computer code.
My complaint about your model was that it makes ID supporters look stupid, a complaint I think you should take seriously, considering that I am one myself.
Comment by Mung — November 20, 2009 @ 7:36 pm
November 20th, 2009 at 9:17 pm
As for "junk" DNA, consider the hypothesis of front-loaded universal genomes.
Such a hypothesis predicts that "simple" organisms would have lots of vestigial "leftovers" from the original front-loaded genomes. In many cases we would expect to find these organisms with genomes encoding genes that would seem like "overkill" for the minimal functions they perform.
Take the Trichoplax for example:
This in an organism that…
and for which…
Comment by Daniel Smith — November 20, 2009 @ 9:17 pm
November 20th, 2009 at 9:52 pm
Nick,
The genomes of today are not the same as the designed genomes, as the existing genomes are an evolved state of those.
Comment by ID guy — November 20, 2009 @ 9:52 pm
November 21st, 2009 at 5:48 am
DAniel Smith:
I disagree on two points:
The theory of "front loading" offers no prediction that modern genomes contain more or less "vestigial leftovers."
The theory of "front loading" offers no prediction that modern genomes contain more or less "overkill."
Comment by Mung — November 21, 2009 @ 5:48 am
November 21st, 2009 at 1:29 pm
In the haploid case, which was the animation, the point was moot and thus effectively irrelevant. I repeated it several times and you still kept up with the "no model", when indeed there was one. A model does not require computer source code to be a model.
The model was simple. Where did I say there was source code needed to make the obvious inference.
In such case, you should have seriously considered pointing out the fact it was moot in the haploid case, and then pointed out the analogous diploid case had no less handwaving than the claims of "synergistic epistasis" and "positive epistasis" in Nachman and Kondrashov's writings, much less Dawkins WEASEL.
I don't take your complaints seriously. You seem to be an ID proponent with an axe to grind against me, and I thank you for being so open about the fact you find me repugnant (which you said specifically at UD).
I don't take your complaints seriously. You help perpetuate the "no model" strawman. I point to an exchange with jistak.
Yup, just as it is in the animation, no specifications for the other details and one can make a vaild inference. The model is simple, the inference obvious (except to those committed to distorting the obvious).
For Haploid U=1 (1 mutation per newborn) would lead to deterioration.
For diploid U=3 (3 mutations per newborn) might be enough if there is no synergistic or positive epistasis or weakly deleterious mutations. Peer reviewed papers have said as much (Muller, Kondrashov, Nachman, etc.). What this would mean is each newborn gets roughly 1 or more harmfuls from mom and dad and 3 new harmfuls on his own. This is a possibly worse scenario than what was depicted in the animation.
And to emphasize, heterozygous "advantage" would even make the diploid case worse since it would result in wasting a lot of individuals just to maintain the heterozygous forms at the expense of the homo or nullizygous forms. This would hardly constitute a refutation of genetic deterioration, but actually strengthen it. Recessive harmfuls would make the case worse since that make the mutation likely to just fester and spread in the population. Not modeling it was actually a generous assumption.
What the animation did demonstrate was the willingness not to even assess the meaning of what was right in front of ones eyes.
Comment by Salvador T. Cordova — November 21, 2009 @ 1:29 pm
November 21st, 2009 at 1:35 pm
Animation model.
Model: Haploids, 1 new harmful per newborn.
Model: Haploids, 1 new harmful per newborn.
Model: Haploids, 1 new harmful per newborn.
Model: Haploids, 1 new harmful per newborn.
There mung, do you get it yet.
Comment by Salvador T. Cordova — November 21, 2009 @ 1:35 pm
November 21st, 2009 at 1:41 pm
Analogous Diploid case:
Mentioned in the OP at UD.
U=3, 3 harmfuls per newborn
Result:
Kid gets about 1 or more harmfuls from each parent. Kid gets 3 additional.
Why is this. With 40 kids, 2 might have no mutations from parents carrying 3 mutations. If parents have fewer than 40 kids, then, it's likely all the kids have 1 mutation or more!
See the resemblance now to the haploid model.
Did I get the chance to make these clarifications? Maybe, but I was so distracted by the strawmen and derailments being put forward I didn't get the chance to add these. I make amends now.
But the data was right there in Nachman's paper.
Comprende?
Comment by Salvador T. Cordova — November 21, 2009 @ 1:41 pm
November 21st, 2009 at 1:46 pm
From Universal Genome in the Origin of Metazoa:
While he predicts that much of this information would be useless, evidence suggests that some of it is co-opted for simpler tasks – thus "overkill".
Comment by Daniel Smith — November 21, 2009 @ 1:46 pm
November 21st, 2009 at 6:05 pm
This is a long thread, and I haven't read every entry. Has anyone shown empirically that a high genetic load will inevitably lead to extinction? Sanford's book only shows inevitable fitness decline as mutational load increases in a simulation.
The reason I ask is, we know there is empirical data for eukaryotic, diploid organisms that contradicts this notion:
Estes S & M Lynch (2003). Rapid fitness recovery in mutationally degraded lines of Caenorhabditis elegans. Evolution 57(5): 1022-1030
From the abstract (my emphasis):
Comment by KC — November 21, 2009 @ 6:05 pm
November 21st, 2009 at 6:14 pm
I'm curious Zach, is the opposite true then as well?
Are non-conserved regions predicted to not have a function?
Comment by Daniel Smith — November 21, 2009 @ 6:14 pm
November 21st, 2009 at 6:35 pm
Clearly not. Some functional regions evolve.
Comment by Zachriel — November 21, 2009 @ 6:35 pm
November 21st, 2009 at 6:44 pm
Zeylab, Mizesko & Visser, Mutational Meltdown in Laboratory Yeast Populations, Evolution 2001. It requires very small populations to detect. The effect did not occur with wild-type yeast. Of course, we can always dial up the mutation rate so that the vast majority of offspring are not viable, though the interesting case is the one on the edge of extinction.
Wildlife biologists believe that once populations become very much reduced, extinction may be difficult to forestall. This is important with regards to safeguarding endangered species. However, as your study suggests, once populations increase, then overall fitness can return.
Comment by Zachriel — November 21, 2009 @ 6:44 pm
November 21st, 2009 at 7:38 pm
So let me get this straight:
Conserved regions are predicted to have function, but then so are non-conserved regions.
So how is conservation related to function?
Comment by Daniel Smith — November 21, 2009 @ 7:38 pm
November 21st, 2009 at 9:36 pm
Conserved regions are predicted to have function. But non-conserved regions can refer to non-functional regions (which evolve at the background rate of mutation) or to functional regions under selection (which are typically similar to precessors, but do not have complete identity).
Comment by Zachriel — November 21, 2009 @ 9:36 pm
November 22nd, 2009 at 2:32 am
You ought to. If you can't even make a case that an ID supporter can agree with you don't have much of a chance against the critics.
Did I say that I find you repugnant? If so, I apologize and retract the statement. If you want to send me a link to the comment I'll post a retraction there too.
Comment by Mung — November 22, 2009 @ 2:32 am
November 22nd, 2009 at 3:31 pm
KC,
I mentioned the Lynch and Estes study at UD. I felt it was too important to discount.
I think it is too hard to say if something is going extinct. My main objection to Lynch and Estes is that function does not necessarily mean the same thing as reproductive fitness. If there is a compensatory mutation, by defintion one has renormalized (redefined) what it means to be fit.
By that standard, microporidia a fit even though they've suffered so much reductive evolution and so badly disfigured that they were once thought to be primitive protozoa even though they were a very advanced species!
I respect the fact that arguments over fitness and function are very subjective.
The fundamental question is the degree of natural selection acting on individual nucleotides. I think this can be resolved without knowing selection coefficients or knowing if something is or is not functional. This can be achieved via sequence divergence studies in real-time (children, parents, grand parents, great grand parents, etc.) and using various mammals (like mice) as outgroups. The idea is to measure the amount of purifying selection that really may exist. I don't think it can be much.
I believe there is a relationship between reproductive excess and the maximum number of nucleotides that can be policed. I don't think that many nucleotides can be policed simultaneously. I think 3 mutations is close to the max based on Nachman's work. Walter ReMine informs me that simulations indicate 10 or more nucleotides would not be feasible to police lest the population experience meltdown. Thus one could have compensatory mutations provided you're just letting go lots of other nucleotides and not selecting against them. Thus a population could be reproductive "fit" but experience sequence divergence at the nucleotide level. This result is of course, completely unsurprising…..
I've advised Dr. Sanford's colleagues of the Estes and Lynch study, and I hope they will take it seriously as the objection you put forward bears serious consideration.
I do believe 10 harmfuls with any selection coefficient and even truncation selection cannot be purged from the population. Compensatory mutations can keep them reproducing, but like the case for microsporidia, this is a renormalization (redefinition) of what it means to be "fit". That's Ok. I'm merely trying to clarify what questions are really being asked and answered. One side is talking reproductive notions of "fitness", and the other side (the ID proponents) are talking about functional notions of "fitness".
I've suggested the question is not whether something is fit, but how much its nucleotides are subject to selection now or in the past. That is a question less subject to argumentation. I don't buy the notion that 100,000,000 years of selection pressure created ultra-conserved regions. The knockout experiments are proof the hypothesis of selection on these regions is probably wrong. We can know the answer empirically!
I have thus encouraged a line of inquiry toward sequence divergence rather than trying to predict reproductive fitness (which is too nebulous to investigate if we're dealing with variable selection co-efficients and nearly neutral mutations).
Sternberg notes the proteins seem to obey one molecular clock and the conserved regions another. I'm betting that there is really no clock as believed. The sequence divergences were the result of design and not clocked mutations. Furthermore the sequence identity (like in ultra conserved regions) are not the result of selection.
The animation, the discussion of Nachman, have bearing on these hypotheses.
Comment by Salvador T. Cordova — November 22, 2009 @ 3:31 pm
November 23rd, 2009 at 12:59 pm
Mung,
You used the word "repellent" on March 6, 2008 at 10:39 pm.
Mung finds Sal Repellent
Getting agreement from the critics is not my goal. Many of the behave like the Anthropogenic Global Warming lobby.
Getting agreement from you is not my goal.
Depiciting visually the impotence of selection is one of my goals. There will be future animations.
In the meantime you keep repeating the "no model" canard.
Zach hasn't modeled and reported the case I described.
He was quite willing to model other scenarious. He is silent on modeling what was clear in the animation and the actual scenario depicted in the animation.
He is invited to make a run on the parameters specified.
By the way, Mung, you're welcome to run a simulation under the model I just specified. The other parameters are of your choosing.
You keep taking cheap shots and making mischaracterizations of what I have written. So long as you persist doing so, I regard you like I regard KeithS — someone with an axe to grind against me personally.
Comment by Salvador T. Cordova — November 23, 2009 @ 12:59 pm
November 23rd, 2009 at 1:52 pm
It is predicted to have function based on the steganography hypothesis also. So which is closer to the truth, natural selection or steganography?
Well, we have reason to doubt selection.
1. How difficult is it for selection of police even 10 nucleotides mutations in diploids with mammalian reproductive capabilities? This is a corollary to Nachman's paradox. Selection may not have the population resources to police ultra-conserved regions or any other regions for that matter.
2. The knockout experiments on mice show serious empirical reasons to doubt selection's effectiveness.
3. If the functionality is a contigency (like a spare tire) it will not be visible to selection.
You keep repeating the error, and I merely repeating the correction to the mistaken notion you put forward.
Comment by Salvador T. Cordova — November 23, 2009 @ 1:52 pm
November 23rd, 2009 at 2:30 pm
It's under consideration, but it may not be worth the bother to produce models when you are not explicit in what needs to be modeled. You were exceedingly unclear in your first comment, immediately conflating your gingerbread people with humans, indeed saying a paper on human mutation was the inspiration for the animation. You only clarified your claim after a week of discussion, only after comments about equations for diploid selection, only after being provided an example of mendelian inheritance, and only after a working simulation was presented.
You only specified a single parameter. If there is a significant reversion and beneficial mutation rate, then the result could be evolutionarily stable.
Comment by Zachriel — November 23, 2009 @ 2:30 pm
November 23rd, 2009 at 2:48 pm
Salvador:
Yes, I did. What I did not say is that I find you repellant, or repugnant.
Wow, almost two years ago, lol. You sure manage to hold a grudge.
Gonna be hard for me to do when:
What I did was point out that:
Does not a model make. Have you since added more to it, or are you asserting that "Haploids: 1 new harmful per newborn. " is sufficient to constitute a model?
Has or has not?
I'll stop when you actually have one. I'm not an unreasonable person, after all. It's just that I prefer eating meat to trying to feast on air.
Comment by Mung — November 23, 2009 @ 2:48 pm
November 23rd, 2009 at 3:35 pm
How does "the art or practice of concealing a message" entail such a prediction?
Humans seem to do fine with many times that number. Indeed, they're overrunning their planet.
That contradicts your earlier stance that "knockout experiments are horrible approaches" to testing selection due to redundancy.
You should know by now that it depends! For instance, if the chance of a fatal catastrophe is high, then the value of redundancy may far outweigh the cost.
Comment by Zachriel — November 23, 2009 @ 3:35 pm
November 23rd, 2009 at 3:50 pm
Haploids: 1 harmful per newborn.
Other parameters you are free to choose. Is that not explicit enough for you? If not suggest some parameters before we make the run.
I look forward to your results. Thank you for taking the time to model things earlier.
But no thanks to your apparent withholding of information that might be supportive of my thesis.
Of course, I suspect you full well know the outcome of what the simulation will yield given any parameter you choose.
I've pointed out what I think the parameters of the corresponding analogous diploid case will be. But lets start with the haploid.
Am I clearer now? Suggest what you think is explicit enough for you for.
Haploids: 1 harmful per newborn
Corresponding Diploid Human: 3 harmfuls per newborn (Nachman), 10 harmfuls (ReMine)
The number of 3 to 10 has direct bearing on your supposition that selection can somehow conserve large regions of a 4 gigabase pair genome.
If it can't police 3 to 10 nucleotides per newborn, wouldn't you say this puts suspicion on the ability to police 4 gigabase pairs on an entire population?
You just claimed selection can police conserved regions. Well among humans we have 99.5% conservation between humans,that's still roughly 4 gigabase pairs. If you can't police 3 to 10 mutations per newborn, doesn't it make the story not exactly believable that selection can police 4 gigabase pairs in a population of thousands if not millions?
Doesn't that also accord with the OP, if selection can't police it now, what makes it believable that selection created the features in the first place.
The Nachman paradox is completely germane to the question of how much selection has specifically affected human evolution. I postulate, in the scheme of things, selection accounts for not much…..say less than 1% of the genome, if that. Thus the Blindwatchmaker is not the answer for the appearance of design and functionality.
I'm not saying anything that geneticists (like Masotoshi Nei of the National Academy of Science) haven't already suspected for quite some time. Selection is mostly irrelevent to the majority of molecular evolution, ergo, selection is irrelevent period.
Comment by Salvador T. Cordova — November 23, 2009 @ 3:50 pm
November 23rd, 2009 at 4:05 pm
If there is a significant reversion and beneficial mutation rate, then the result could be evolutionarily stable.
Not ergo.
Comment by Zachriel — November 23, 2009 @ 4:05 pm
November 23rd, 2009 at 4:29 pm
That's exactly the problem. How does this explain contingecies where there is not fatal catastrope. Example: knockout of ultra conserved regions in mice.
If you say, they ultra conserved regions are not functional, fine, then it is clear selection can't conserve them for 100 million years. If you say they are functional, then it is clear knocking them out doesn't result in fatal catastrophe. Worse yet, we then have an example of function that evolved that is also invisible to selection.
Either way, the counter argument "if the chance of a fatal catastrophe is high" is inapplicable.
Of course contingencies are valuable, but that is not the question. The question is whether selection can see it and preserve it.
Selection cannot estimate future probabilities, it cannot assess potential value, only immediate value.
It is myopic and bliind. It will have a propensity to lay waste to valuble contigencies as evidenced by microsporidia. It goes toward what is expedient, not necessarily what is good for the long term.
You're presuming if something is of value in the future or contingent circumstances, selection will necessarily see it and properly weigh the pros and cons. Wrong. Witness, cystic fibrosis, sickle cell anemia, blindess in cave fish, destruction of major function in microsporidia, etc.
And I already pointed out the case of ultra-conserved regions. Which ever way the data is reasonably interpreted will not be favorable to mainstream evolutionary theories.
Comment by Salvador T. Cordova — November 23, 2009 @ 4:29 pm
November 23rd, 2009 at 4:45 pm
That's not your claim. You made a universal claim, that redundancy is not much visible to selection. Yet that is not always the case. Redundancy can sometimes have a significant evolutionary benefit. It depends on cost and benefit.
Consider a situation where every other biological widget is defective, and the cost of redundancy is negligible. Children with a single widget will die 50% of the time. Children with a double widget will die only 25% of the time. But even if there is only a much smaller chance of the widget failing, it can still represent a selectable difference that has to be compared to the cost of redundancy.
Comment by Zachriel — November 23, 2009 @ 4:45 pm
November 23rd, 2009 at 6:08 pm
No.
If Widget A and Widget B are fully redundant, then if A is missing (as in knocked out like in the mice experiment), and B exists, the organisms survives at the same rate or close to it.
No. Such a scenario casts doubt on whether the systems can be classified redundant in the sense I was illustrating.
You've redefined my sense of redundant. That's fine. But it was not the sense I was using and thus you're refuting an argument I'm not really making.
You said:
you would have been more in line with the scenario I had in mind if you said
That is consistent with the redundancies I had in mind. You're criticising an argumement that I'm not making.
I gave an empirical illustration with the mice, microsporidia, blind cave fish of what I had in mind. In fact, in the case of blind cave fish, the redundancy (sensory organ in addition to touch or "smell" or hearing or taste) was selected against, not for.
Comment by Salvador T. Cordova — November 23, 2009 @ 6:08 pm
November 23rd, 2009 at 6:19 pm
That's right. If an organism has a redundant widget and one fails, it survives. If an organism has only a single widget and it fails, it dies. The odds of survival improve with redundancy. This is a selectable difference.
Redundancy, "serving as a duplicate for preventing failure of an entire system upon failure of a single component."
You might want to define what you mean.That's exactly what you mean.
Comment by Zachriel — November 23, 2009 @ 6:19 pm
November 23rd, 2009 at 6:24 pm
Significant reversion? As in surviving harmfuls magically mutating back to the functional state. Sure, you could blast the organism with radiation, and it could cause reversion but at the cost of creating more harmfuls. I don't know if this is a real fix.
I addressed the issue of renormalization and redefining "beneficial". Such a fix is dubious, especially with respect to being able to police genomes that are deeply conserved. Defining "beneficials" this way is analogous to "beneficials" like Sickle Cell anemia or blindess in cave fish or deletion of major parts of the genome in microsporidia or broken pumps in bacteria that develop anti-biotic resistance.
The issue was whether the harmfuls are purged, not weather one can argue that enough "beneficials" have compensated to keep the population levels maintained. The issue is with ultra conservation in the genome and whether selection can maintain it.
Thus the way to run the simulation to preclude this dubios "fix" is to run it without "beneficials" since that is not the question being explored. I'm not disputing the emergence of compensatory mutations, I'll even grant it as a given. The issue is if deterioration of existing function (or even simple ultra conservation) can be preserved.
Comment by Salvador T. Cordova — November 23, 2009 @ 6:24 pm
November 23rd, 2009 at 6:54 pm
Reversions are known to occur. You said to choose the values.
You previously said that "the beneficial-to-deleterious is irrelevant." Now, you say it has to be zero.
Beneficial and deleterious are defined in terms of fitness. Nearly neutral mutations modify existing complexes, perhaps slightly increasing or decrease binding, or the rate of enzymatic activity. Most nearly neutral mutations neither increase or decrease the overall complexity of an organism.
You seem to think that a deleterious mutation necessarily breaks some fundamental function, or that a beneficial mutation necessarily creates a new functional complex. That's not generally the case when the mutation is nearly neutral.
Reread Nick's comment.
Comment by Zachriel — November 23, 2009 @ 6:54 pm
November 23rd, 2009 at 6:56 pm
Who said the second widget fails? We didn't knock out the second widget in the mice did we? Obviously not, the thing lived.
You're presuming:
1. the likelihood of failure of widget A is high
2. the likelihood of failure of widget B is high
That is not consistent with the scenarios I outlined, such as the mice.
For selection to see things, you're having to assume high failure rates in both designs, that is clearly not what I'm describing.
this is not consistent with your mischaracterization:
The example of the airplane magnetos. Failure rate of one from mechanical issues is probably on the order of 0.1% or less. Thus failure of both would 0.1% x 0.1%. Such a small difference in an analogous case for organisms would be barely perceptible by selection.
In fact, for the blind cave fish the "contingency" desiging of sight, it wasn't enough, neither for the microsporidia genome. In such case, selection acted against a contingency design. Exactly the opposite of what you are suggesting, and consistent with my claim that selections eagerness toward expediency results in long term damage to contingency functions. Selection didn't like carrying what it perceived was extra baggage, junk DNA, junk function with no immediate benefit (like the ability to see in cavefish or the genome in microsporidia).
Your presumption that it is a selectable difference is flawed. You have numerous counter-examples.
I showed how your mischaracterization could be amended:
That is consistent with the magneto scenario I was describing.
Comment by Salvador T. Cordova — November 23, 2009 @ 6:56 pm
November 23rd, 2009 at 7:01 pm
I did say choose the values, and your need to choose of unrealistic scenarious that are indistinguishable from black magic underscores the difficulty I'm trying to illustrate.
I suppose if aspects of Darwinism are indefensible, sophistry is the last resort. Appealing to massive reversions (which is what would be needed) as a solution to the difficulty is sophistry. This is analogous to presuming all the hereditary diseases in the next generation will disappear spontaneously. Yeah right, reversion happens in sufficient quantity.
Comment by Salvador T. Cordova — November 23, 2009 @ 7:01 pm
November 23rd, 2009 at 7:15 pm
Then you shouldn't have said and continued to argue this point:
Redundancy can have a selectable benefit, for example, if the chance of a fatal catastrophe is high, then the value of redundancy may far outweigh the cost.
I provided an example in extremis so that you could see why your original statement was wrong. And you're wrong again. It's a simple cost-benefit analysis. If the cost of redundancy is low, then may only take a modest change in fitness for it to be an overall benefit.
Comment by Zachriel — November 23, 2009 @ 7:15 pm
November 23rd, 2009 at 7:33 pm
Comment by Mung — November 23, 2009 @ 7:33 pm
November 23rd, 2009 at 7:39 pm
And I already countered by saying redefining "beneficial" in terms of reproductive fitness is flawed.
Yup, and microsporidia is repleate with lots of "beneficial" mutations which essentially wiped out an advanced genome and turned it into a parasite. By the way, are phages considered to be parasites?
Compensatory beneficials don't at all help the case with deeply conserved regions where the question of purifying selection can actually be measured real time.
If anything, the scenario Nick provided makes deep conservation worse if that is indeed the case for mammals. Scrambling the genome like that and then having the relics favored would not bode well for the maintenance of deep conservation.
Another interpretation: Selection and mutation favor parasitic simplicity over mammalian complexity. Mammals, birds, etc. don't fare well under extreme mutagenic environments.
The study highlights the extreme inapplicability of extrapolating the frequency of beneficials in phages to humans.
Comment by Salvador T. Cordova — November 23, 2009 @ 7:39 pm
November 23rd, 2009 at 8:11 pm
Notwithstanding the objections I have already previously put forward:
1. Having only Widget A results in 50% deaths
2. Having A and B results in 25% deaths
In order to evolve and maintain the existence of widget B we have to have a kill rate 50% with Widget A only.
Say we have 10 such redundant pairs. We would need similar kill rates to establish Zach's scenario. But what would this demand of the population. Kill 50% off for one redundant pair, kill 50% off of the survivors for the next redundant pair, etc. The result is that after trying to evolve and maintain 10 such redundant pairs, one is left with one survivor for every 1024 births!
50% raised to the 10th power = 1/1024.
So to evolve and maintain even 10 such redundant pairs we have to kill off 1023 of 1024 individuals. Not too promising.
I don't think the cost of maintaining redundant systems has been explored, but it surely isn't cheap.
Zach might counter, "well the kill rates don't have to be that high". In such case, they become less visible to selection, the opposite of what he is trying to argue, and that begins to bend toward the other scenario I suggested. If Widget A is highly reliable, then Widget B doesn't get selected for if ever. This corresponds to this scenario:
But if we take Zach's scenario:
Even a small number of such redundant pairs would be prohibitively difficult to evolve and maintain in the first place from a cost perspective. Perhaps the scenario Zach provided isn't as promising as it seems on first glance when confronted with cost issues.
Comment by Salvador T. Cordova — November 23, 2009 @ 8:11 pm
November 23rd, 2009 at 8:30 pm
You seem to think that a deleterious mutation necessarily breaks some fundamental function, or that a beneficial mutation necessarily creates a new functional complex. That's not generally the case when the mutation is nearly neutral. The term "beneficial mutation" refers to those mutations that cause an increase in reproductive fitness.
I have no idea what you're trying to say.
Consider a deme of single widgets. Widgets have a failure rate of 50%, so half the young are stillborn. Each produces 100 children, or 50 viable children. Consider then a deme of double widgets. A quarter of the young are stillborn. Each produces 100 children, or 75 viable children. This is a very significant evolutionary benefit and it will take little time for the double widgets to predominate in the population.
In a more detailed model, we might then account for the cost of redundancy. Say double widgetism reduces fecundity by 10%. So, single widgets still produce 50 viable children, while the double widgets produce about 67 viable children. Still a significant advantage.
Because you don't understand this simple model, we have reason to doubt your analysis of the more complex models.
Comment by Zachriel — November 23, 2009 @ 8:30 pm
November 23rd, 2009 at 11:37 pm
There is more than one pair of redundant systems in an organism as complex as a mammal. Denton said almost every gene has some redundancy. I put forward the problem of even 10 such systems in an organism:
Widget A and B for function #1
Widget C and D for function #2
Widget E and F for function #3
Widget G and H for function #4
Widget I and J for function #5
Widget K and L for function #6
Widget M and N for function #7
Widget O and P for function #8
Widget Q and R for function #9
Widget S and T for function #10
If each these widget pair are all involved simultaneously in catastrophic issues such that the environment wipes out 50% of the survivors (after presumably selection uses up population resources for other problems), then there is a high cost of evolving and maintaining these systems simultaneously.
Widget pair A&B require 50% of the population to be wiped out. That leaves 50% of the population for by selection for other traits like say, Widget pair C&D. But Widget pair C&D require 50% of those survivors to be eliminated ( 50% * 50% ), etc. 10 such widget pairs would need 50% to the 10th power such individuals to be wiped out.
Thus the cost of evolving and maintaining just a mere 10 such widget pairs given your parameters is that 1023 of 1024 individuals must be wiped out (50% raised to the 10th power). You might argue that I should use 75% raised to the 10th power fine. In such case I'll merely point out even 20 such redundant systems using the 75% figure would still be a bad scenario for selection. And 20 is a gross underestimate for the number of redundancies in an organism.
If you think the calculation needs to be modified, feel free to give your version, but there is a population price to pay for saying the systems deal with catastrophic failure and also the highly austere environment you propose where a creature has a 50% chance of not reproducing even if Widget A is in place.
In any case Zach, I'm still intereted to hear about the 3-10 mutation diploid simulation under modestly realisitc parameters. This has complete bearing on the questions at hand.
My conjecture: selection can't police 4 gigabase pairs. 3-10 mutations per newborn would be too much for selection to maintain any sort of conservation much less conservation of 4 gigabase pairs.
This conjecture should not at all be surprising given 90% or more of molecular evolution is neutral according to the non-Darwinian (Masatoshi Nei, Kimura, others).
I find it astonishing that if 90% of molecular evolution is not subject to selection, that anyone still seriously believes the blindwatchmaker accounts for so much of evolution.
Comment by Salvador T. Cordova — November 23, 2009 @ 11:37 pm
November 24th, 2009 at 9:33 am
Of course you would. That would translate into a ~95% failure rate, so fecundity would have to be somewhat over 20 per female. This is not unusual for many species.
Before the evolution of species with lower reproductive rates, we would expect then to see increased reliability of individual widgets, or increased redundancy, or both. The relationship is very simple to model.
In any case, it was just an example to show you why your original statement was incorrect. Redundancy can be quite visible to selection. It would be useful to the discussion for you to explicitly recognize this.
Yes, it is astonishing. Of course, much of that molecular evolution is neutral or nearly so, meanderings across broad plateaus. Selection is still considered critical for adaptation, though, something we can directly observe.
Comment by Zachriel — November 24, 2009 @ 9:33 am
November 25th, 2009 at 12:31 pm
Not only can we observe it, we can write papers about how it doesn't seem to be the driving force behind evolution at all!
Comment by Mung — November 25, 2009 @ 12:31 pm
November 25th, 2009 at 1:03 pm
Zach,
Thank you for your previous computer sims. They were informative but I've suggested ones that I think are more relevant to the topics at hand.
1. Haploids: 1 new harmful per new born, parameters of your choosing, and I'd prefer some moderately realistic constraint
2. Diploids: 3 harmfuls, human fecundity, low selection
3. Diploids: 10 harmfuls, humand fecundity, truncation selection
I look forward to hearing of your scientific research into these questions as they have bearing on the question of deep conservation.
Happy Thanksgiving.
Comment by Salvador T. Cordova — November 25, 2009 @ 1:03 pm
November 25th, 2009 at 3:10 pm
Zachriel, I suggest you use a generation time of 100 million years.
Comment by Mung — November 25, 2009 @ 3:10 pm
November 25th, 2009 at 5:20 pm
Salvador T. Cordova,
You've held to the belief that there is no such thing as a beneficial mutation simply by redefining the term. You apparently think that a slightly deleterious mutation breaks a complex rather than just adjusting the rate or degree of interaction. There are certainly mutations that break complexes, and these are quickly weeded out of populations by selection. There are also mutations that can provide a significant advantage, and these tend to fix quickly.
But most mutations are neutral or nearly neutral. Being nearly neutral, they don't often break complex systems, but just alter them slightly. Sometimes, there is a great deal of play in particular interactions. On a fitness landscape that would be represented by a gently rolling plateau with little selective difference between local maxima. In such a landscape, not only are there nearly neutral deleterious mutations, but there are also nearly neutral beneficial mutations.
Gregor's Bookkeeper is an abstraction, an "all other factors being equal" simulation. There is no actual organism that is being modified. This is suitable to test many general claims, though.
Normally, when we say there is a deleterious mutation, we mean it has reduced reproductive potential. Conversely, when we say there is a beneficial mutation, we mean there is increased reproductive potential. We have to establish some sort of agreement as to what it means for a mutation to be deleterious and beneficial.
The first pair of breeders are due to mate next year. If all goes well, they'll be grandparents around stardate 100,002,010.
Comment by Zachriel — November 25, 2009 @ 5:20 pm
November 25th, 2009 at 11:07 pm
That should be more than enough time for them to become an entirely new species, so it's time to start the "mutational doomsday clock" (TM) all over again.
Comment by Mung — November 25, 2009 @ 11:07 pm
December 3rd, 2009 at 1:17 am
Which guarantees evolution will be at a virtual standstill. This really doesn't help Zach's case, in attempting to "fix" one thing in evolutionary theory, another breaks.
Comment by Salvador T. Cordova — December 3, 2009 @ 1:17 am
December 3rd, 2009 at 7:45 am
Zachriel wrote:
Yeah, like producing sterile offspring.
Comment by angryoldfatman — December 3, 2009 @ 7:45 am
December 3rd, 2009 at 8:09 am
Or even sooner. Twenty-five percent of human embyryos are miscarried before the sixth week, often without the mother even knowing about it.
Comment by Zachriel — December 3, 2009 @ 8:09 am
December 3rd, 2009 at 12:46 pm
In utero death does not necessarily imply the embryo had bad mutations. Healthy individual die from issues not realted to their genes. The baby could have died from issues with the mom.
By the way, did you ever run the 10 mutation dioploid scenario and are just not publishing it here at TT?
How about the 1 mutation haploid scenario?
Comment by Salvador T. Cordova — December 3, 2009 @ 12:46 pm
December 3rd, 2009 at 1:14 pm
Sure, but most first trimester miscarriages in humans are due to chromosomal abnormalities.
I've responded to most of your points analytically. As for a simulation, there is no way to do that without a clear understanding of what is being simulated. For instance, we normally define a deleterious mutation as one that leads to decreased chance of reproductive success, and a beneficial mutation as one that leads to increased chance of reproductive success.
Comment by Zachriel — December 3, 2009 @ 1:14 pm
December 3rd, 2009 at 4:12 pm
So the answer is that you didn't run the simulations?
Or is the answer you ran the simulations with various parameters, and you didn't want to report the results since it might not be in accord with the position you want to promote.
I've already, try shutting down the beneficials. This is reasonable since real beneficials (not as in fake beneficial like sickle cell anemia) appear 1 out of a million.
So the answer is that you didn't run the simulations?
Comment by Salvador T. Cordova — December 3, 2009 @ 4:12 pm
December 3rd, 2009 at 4:34 pm
Come on, Zach. Show how science is done.
Comment by chunkdz — December 3, 2009 @ 4:34 pm
December 3rd, 2009 at 4:38 pm
If you want to discuss a simulation you have to be willing to clearly the define the characteristics of such a simulation. For instance, we normally define a deleterious mutation as one that leads to a decreased chance of reproductive success. You seem to disagree with this widely accepted definition.
Perfeito et al., Adaptive Mutations in Bacteria: High Rate and Small Effects, Science 2008.
A rate of 10^-5 per genome per generation with a mean selective advantage of 1%.
Comment by Zachriel — December 3, 2009 @ 4:38 pm
December 3rd, 2009 at 5:09 pm
So Zach, did you or did you not run a simulation with 10 harmfuls?
If you ran a simulaiton, you were awfully quiet about reporting it here at TT.
So Zach, did you or did you not run a simulation with 10 harmfuls?
You don't have to answer that question, and you don't have to run the simulation. But a direct and non-misleading response would be appreciated by our readers.
Comment by Salvador T. Cordova — December 3, 2009 @ 5:09 pm
December 3rd, 2009 at 5:22 pm
Diploids:
To remove the effect of fake beneficials, set beneficial to harmful ratio to 0.
Try 4 offspring per couple, or 2 per individual.
Try 10 harmfuls per person.
Haploids:
To remove the effect of fake beneficials, set beneficial to harmful ratio to 0.
1 harmful per newborn
reversions, 0 (that is good enough for large genomes)
Comment by Salvador T. Cordova — December 3, 2009 @ 5:22 pm
December 3rd, 2009 at 5:23 pm
I provided details in my post prior to this one.
Comment by Salvador T. Cordova — December 3, 2009 @ 5:23 pm
December 3rd, 2009 at 5:45 pm
Yes, I have run simulations of evolving populations with an average of 10 deleterious mutations per genome (defined biologically as a reduction in reproductive potential). I have provided you the data on this blog, provided links to a bulletin board with screen shots, and provided you the open source software. But apparently never having bothered to look, you indicated that this wasn't quite what you had in mind with regards to the properties of the simulation. So I have repeatedly asked you to provide additional detail—which you have been unwilling or unable to do.
Gregor's Bookkeeper
Comment by Zachriel — December 3, 2009 @ 5:45 pm
December 3rd, 2009 at 11:39 pm
I provided details two posts above. Feel free to report the results with those parameters.
Comment by Salvador T. Cordova — December 3, 2009 @ 11:39 pm
December 4th, 2009 at 9:32 am
I appreciate you attempting to further the conversation, but again, there is still ambiguity. Gregor's Bookkeeper defines a deleterious mutation as one that reduces the relative reproductive fitness of an individual. If a mutation reduces fitness per this definition, then it is quite clear that a reverse mutation would increase fitness per the definition. So we know there must be such a thing as a beneficial mutation. When you say "fake beneficials," it indicates you reject this definition, so we are no further along.
Most mutations are nearly neutral in the simulation. That can be pictured as a wide and rolling plateau. Fitness can meander about in such a landscape, meaning fitness can increase or decrease, including leaving local maxima to explore other areas of the landscape.
(Also, without having the reprogram the model, mutations refers to average per individual. Gregor's treats each mutation as an independent event.)
Under the assumptions of the model, the average relative fitness is expected to decrease proportionally over time, slower with larger populations. And that is what we observe. If we increase the number of offspring, population, variances, or reduce the mutation rate, then the fitness decreases more slowly often appearing to stabilize. Again, this is the expected result.
The interesting experiment is to find the crossover points. When does fitness stabilize and under what conditions? Even a very small number of beneficial mutations changes the results dramatically.
Apparently, mice don't experience genomic meltdown with their high reproductive rate. So we're only quibbling over a few slowly reproducing mammals.
Comment by Zachriel — December 4, 2009 @ 9:32 am
December 4th, 2009 at 9:56 am
In large genomes, not 10 to 50 genes, reverse mutations will come at an enormous price, namely more damage before a reversion happens, sort of like starting out with 2,000,000,000 coins all heads. one then starts shaking the platform the coins are on. Reversion happens, but the configuration of coins is completely scrambled by then.
Comment by Salvador T. Cordova — December 4, 2009 @ 9:56 am
December 4th, 2009 at 10:21 am
Gregor's Bookkeeper uses the biological definition of a deleterious mutation as one that decreases the relative reproductive fitness of an individual, and a beneficial mutation is one that increases the relative reproductive fitness of an individual. Is that not an appropriate definition?
I keep providing cites to papers and illustrations about beneficial mutations, which you continue to ignore. if you don't think that a longer beak can be an evolutionary advantage depending on the environment, then I'm not sure how you are defining these terms or how you expect to simulate them.
Comment by Zachriel — December 4, 2009 @ 10:21 am