|
Author
|
Topic: The role of chance in biological evolution
|
peter borger
Member
Member # 722
|
posted 01. June 2006 10:10
quote:
Peer review is only the start of the process. It provides an opportunity for peers to read, understand, verify, extend or possibly refute the findings.
Peer review is only working when peer review is double blind.
Peer review is only working when science is not dogmatic.
peebee
IP: Logged
|
|
Zachriel
Member
Member # 1793
|
posted 01. June 2006 10:17
Peer review is hardly a perfect system, because peers are people with all their foibles.
But if your intention is to attempt to convince your peers, those educated in the details of often arcane scientific disciplines, then publication in journals is the appropriate and preferred method of communication.
IP: Logged
|
|
Scott
Member
Member # 1222
|
posted 01. June 2006 10:44
I'd like to get back to the topic of the role of chance in biological evolution.
Say a new mutation appears. What is the "chance" (i.e., probability) that:
1. It will make it into the next generation.
2. It will not make it into the next generation?
IP: Logged
|
|
John A. Davison
Member
Member # 1425
|
posted 01. June 2006 11:16
So would I Scott. My own answer to the question is the same as Berg's:
"Neither in the one nor in the other is there room for chance."
Nomogenesis, page 134
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 01. June 2006 12:19
Re: peer review.
In the crop circles case, if we assume that crop circles are somehow the product of man, then the likely generation of the intricate discussion in Levengood's work is that Levengood produced a work of creative science fiction. I would assume that the only real way to validate his results would be to go to specific sites that he documents, and validate that some of the phenomenon he describes are realistic. Do you think that the journals put that much energy into validating submitted works?
Sorry Scott, John. This thread has significant crop circle discussion. In a way the discussion seems to be winding down, so I don't want to create a separate thread for it.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 01. June 2006 12:48
Re: the role of chance in biological evolution
There's a new book out, "Genetic Entropy & the Mystery of the Genome" by Dr. John Sanford. I haven't read it, but there's a lot of chatter about it on Uncommon Descent.
Dr. Sanford seems to be suggesting that most random mutations are not neutral, but are slightly deleterious. He suggests that a slightly deleterious mutation will be treated just like a neutral mutation because it has insufficient signal to be heard above the noise.
One would certainly think that a significantly deleterious mutation would be filtered out by natural selection quite quickly. However, there are something over 5000 known human diseases, some of which are rather significant, which are known to be caused by specific point mutations. This would indicate that natural selection isn't all that effective even at filtering out deleterious mutations (especially mutations in recessive genes, I understand.)
However, a "significantly benefitial" mutation is all but impossible. It seems that neo-Darwinists have always said that the development of life is based upon many slightly beneficial mutations, not on a few significantly beneficial mutations.
Now comes the "slightly beneficial" mutation. Surely they happen sometimes, but they clearly happen much less often than slightly deleterious mutations do. If natural selection has trouble filtering out significantly deleterious mutations, it surely has more trouble filtering out slightly deleterious mutations. This, Dr. Stanford attributes to a lack of signal to noise ratio. (A term I used to try to explain this phenomenon to my brother, an electrical engineer, less than a year ago.)
If natural selection has trouble filtering out slightly deleterious mutations, by the same token it has trouble filtering in slightly beneficial mutations.
Think about it, each organism has in the order of 10,000 coding genes. Each gene has multiple alleles. Each of those alleles has been tested by natural selection to be at least "survivable". But for each gene, given a particular environment, one of those alleles will be the "best" and another will be the "worst", and a continuum inbetween. Now, one organism, let's call him bFast, got his hand dealt. In some genes he's got the "best" allele, but in other genes he's got the "worst" allele, but mostly his alleles are somewhere between the extremes. Let's say that he also has a lucky "slightly beneficial" mutation. Is that lucky "slightly beneficial" mutation going to be enough to give bFast enough advantage to outweigh the fact that he is competing with the guy next door who has a slightly better mix of alleles? In the vast majority of cases, it won't make one hair of a difference. In any case, some earthquake will come along and swallow up bFast and his slightly beneficial mutation.
So remains that there's five categories of mutations:
1 - Significantly deleterious. Natural selection, with some struggle, will remove the thing from the gene pool.
2 - Slightly deleterious. Natural selection will threat the thing just like it treats neutral mutations.
3 - Neutral mutations. Whether they end up in the genome or not is dictated by factors totally independant of natural selection. If my progeny survive, if I happened to pass on the mutation to them, then it survives.
4 - Slightly beneficial. Natural selection will threat the thing just like it treats neutral mutations.
5 - Significantly beneficial. It is reasonable to say that they don't happen, at least not by chance. Maybe once per million years in all of the organisms out there, but what difference does that make.
Let me only preface the above discussion with one addendum in support of NDE. When the environment changes, when the demands on an organism change, the ratio of deleterious to beneficial mutations also changes. But probably not all that much.
Does that help bring this topic around?
IP: Logged
|
|
Zachriel
Member
Member # 1793
|
posted 01. June 2006 14:00
Scott: "Say a new mutation appears. What is the "chance" (i.e., probability) that:
1. It will make it into the next generation.
2. It will not make it into the next generation?"
Consider the simplest case, a neutral mutation. Principles of population genetics (a field started with by Hardy-Weinberg a century ago), give us the answer. And the answer depends largely on population.
Bruce Fast: "1.2.3.4.5"
Neutral and near-neutral mutations are subject to genetic drift. In a large population, most neutral mutations will never be fixed, but as mutations are made continuously, some neutral mutations are always subject to fixation. In a small population, neutral mutations have a better chance of being fixed -- even if they are slightly detrimental.
The larger the population, the more likely a slightly non-neutral mutation will be subject to selection rather than drift. The smaller the population, the converse is true.
And then sexual selection is the big amplifier. You may or may not get the most preferred mate simply because of the shape of your nose (much less other factors such as physical or social grace).
IP: Logged
|
|
John A. Davison
Member
Member # 1425
|
posted 01. June 2006 14:54
It is perfectly clear that Pierre Grasse, Leo Berg, Otto Schindewolf, William Bateson, Richard B. Goldschmidt and myself are all dead wrong when we dismiss the particulate gene and random mutation as factors in organic evolution. If you are going to ignore us we will ignore you. That is that folks. Now you just carry on without me. Thank you very much. I have better things to do right now.
IP: Logged
|
|
Jack
Member
Member # 265
|
posted 01. June 2006 15:02
John A. Davison said: "It is unfortunate that can't also be done at ARN, Pharyngula, Panda's Thumb, FringeSciences and, most especially, Uncommon Descent where the PEH pleads loud and clear for Intelligent Design as do every one of my papers as well as the publications of my sources on which they are so firmly based."
Hi Dr. Davison,
Why are you not allowed to present your PEH on ARN or Uncommon Descent?
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 01. June 2006 15:45
Dr. Davison: quote:
It is perfectly clear that Pierre Grasse, Leo Berg, Otto Schindewolf, William Bateson, Richard B. Goldschmidt and myself are all dead wrong when we dismiss the particulate gene and random mutation as factors in organic evolution.
That's what my last post said, random mutation doesn't produce organic improvement. Albiet, some of us just can't look at a list of the names of significant players and say, "these great men said this, so it is decided." We/I must work through the facts. I would prefer if you joined in on the working through of the facts.
IP: Logged
|
|
John A. Davison
Member
Member # 1425
|
posted 02. June 2006 05:25
Bruce.
So you don't think I am working through the facts eh?
I discarded the Darwinian fairy tale 50 years ago if those the "facts" you are concerned about. I regard your comments as a cheap shot. Allelic mutation never had anything to do with a purely emergent and predetermined evolution. That is where I stand. It also has nothing to do with "working through tyhe facts" and certainly nothing to do with "joining in." "Joining in" is not my style.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 02. June 2006 12:34
Dr. Davison, the issue is not that you haven't worked through the facts, it's just that you seem to get frustrated that some of us haven't worked throught the facts yet.
My reality is that I have a job that has nothing to do with biology, I haven't had opportunity to spend years studying biology. I, and others, need to work through the facts. When we do, you seem to get frustrated. Further, as you clearly have worked through the facts, I would enjoy it if you spent more time reworked the facts with us, rather than pointing to experts saying effectively, "these men worked throught this stuff, they came to this conclusion."
IP: Logged
|
|
Scott
Member
Member # 1222
|
posted 03. June 2006 14:29
quote:
Consider the simplest case, a neutral mutation. Principles of population genetics (a field started with by Hardy-Weinberg a century ago), give us the answer. And the answer depends largely on population.
Well, what is the answer, lol?
Also, I am not asking (yet) about probability of fixation. Please understand that. Again:
Say a new mutation appears. What is the "chance" (i.e., probability) that:
1. It will make it into the next generation.
2. It will not make it into the next generation.
IP: Logged
|
|
Zachriel
Member
Member # 1793
|
posted 03. June 2006 17:26
Scott: "Also, I am not asking (yet) about probability of fixation."
Sorry, I read some ambiguity.
Scott: "Say a new mutation appears. What is the "chance" (i.e., probability) that:
1. It will make it into the next generation.
2. It will not make it into the next generation."
In a sexually reproducing population, there is generally a 50% chance that a given germ cell allele (recently mutated or not) from a parent will make it to a zygote of the next generation — unless it is extremely detrimental and causes immediate death.
This includes everything from beneficial mutations to neutral or near neutral mutations to somewhat detrimental mutations. Most mutations are neutral.
IP: Logged
|
|
Bruce Fast
Member
Member # 924
|
posted 03. June 2006 17:43
Zachriel, "Most mutations are neutral." I challenge this statement.
I understand that making a case from the molecular clock hypothesis is a bit dicey becaues the hypothesis, though it is recognized to have some value, certainly has its leaks. However, the usage I present is, I believe, well within the zone of validity with reguards to the hypothesis.
Now, Brown & friends published a few years back (I made a specific sitation on this forum within the past month) that they examined a segment of non-coding DNA, and analyzed it to detect the speed of the molecular clock. They found, in mammals, a 2% nucleotide point mutations divergence per million years. That translates to 1% per mil per branch. If one considers that this pace is a benchmark for how many mutation attempts there are, one notices that for most genes, the molecular clock ticks MUCH slower than this.
My simple conjecture is this, if a particular gene's molecular clock ticks at 0.1% mutations per mil (still a fast tick) then 90% of the mutations to hit that gene were rejected, 90% were deleterious. One also must recognize that the hystone H4 gene all but does not tick at all. So any nucleotide mutation which causes a different amino coding is deleterous.
If the above is true, therefore, the only way that your statement is "true" is if you count every mutation in non-coding "junk" DNA. However, within coding DNA, the vast majority of point mutations are deleterious.
(About junk DNA, I, along with many IDers doubt that junk is a very inapropriate term. However, I suspect that mutations within the non-coding DNA produce change in the organism, but change that is on scales that are not a priori beneficial or deleterious. I suspect that control, such as how long the femer will grow, will be discovered to often be determined in the non-coding DNA.)
IP: Logged
|
|
|
Printer-friendly view of this topic