The edifice of the Modern Synthesis has crumbled, apparently, beyond repair.
– Eugene Koonin (2009)
Does this make Eugene Koonin an evolution skeptic?
The summary of the state of affairs on the 150th anniversary of the Origin is somewhat shocking: in the post-genomic era, all major tenets of the Modern Synthesis are, if not outright overturned, replaced by a new and incomparably more complex vision of the key aspects of evolution. So, not to mince words, the Modern Synthesis is gone.
I’m still struggling to incorporate Alan Fox’s allegation that I am an evolution skeptic. I still don’t really know what it means to be an evolution skeptic. Eugene Koonin rather obviously rejects the view of evolution held by Alan Fox. Is Eugene Koonin an evolution skeptic?
Or is this just another example of Creationist quote mining. Maybe it’s both.
What say you, “skeptics”?
The Origin at 150: is a new evolutionary synthesis in sight?
How does blind and mindless processes explain HGT? Or is this just another science-free just-so story?
Yes. That’s all the reply you have to my comment?
John Harshman,
Both apoptosis and DNA repair reduces variation by repairing cells with mutations that may hamper or change their function if the cell cannot be repaired it is terminated. Thus cells with the original sequence are preserved reducing the DNA variation among the cells.
The process exists in the cell germ lines.
colewd,
Please, if you’re going to quote something, provide an actual citation so someone can look up your source. Also leave out the useless stuff like affiliations. I don’t know from this abstract what they’re talking about, and I suspect you also have no idea but are just picking an abstract with words you like.
John Harshman, \
doi: 10.1530/REP-10-0232
Apoptosis in the germ line
R John Aitken, Jock K Findlay1, Karla J Hutt1 and Jeff B Kerr2
This is one of many papers available that will show you that variation reducing mechanisms are in the germ line both male and female.
Too little variation is too cold. Too much variation is too hot. It’s fine tuning and fine-tuned turtles all the way down.
Natural Selection is an amazing fine-tuner.
colewd,
And yet they appear designed to increase variation with all that recombination stuff. Design – the universal buffer.
A complete citation includes the name of the journal, the year, the volume, and the page numbers. In the google age, that might seem unnecessary, but it just makes it easier to find. Have some consideration. A link would have been fine too.
OK, I found it. I see nothing to tell me that apoptosis has the function you claim. Mostly it’s a standard part of development. In the germ line it mostly regulates the number of sperm produced. There is an additional minor function of getting rid of sperm with badly damaged chromosomes. Nothing about limiting variation.
More importantly, the whole idea that the cell is engineered to get rid of variation is silly.
John Harshman,
Apoptosis is part of the cell cycle. Every time a cell divides it checks its DNA at a cell cycle check point. It then corrects it and re checks it. If it was unable to correct it apoptosis is triggered. So the cell is constantly removing changed DNA or variation of DNA is eliminated. If the cell cycle proteins that control these processes have mutations cancer can result. P53 is one of the famous proteins involved in the cycle so is beta catenin. The cell controls cell division by regulating the amount of beta catenin in the nucleus of the cell. Vitamin d regulates beta catenin.
To build a human you need 10^13 cell divisions or around 45 cycles. There is cell to cell variation inside a human as measured by changes in the DNA sequence but it is small. The repair or kill mechanisms must be very accurate to build a human with nominal variation. A human will incur around 10^16 cell division during his life time.
The variation that is seen in the sperms or eggs DNA may be nothing more than variation that you would expect between the 3X10^13 cells inside a human.
http://www.nature.com/nrd/journal/v8/n7/images/nrd2907-i2.jpg
Please support this claim. How does a cell “check its DNA”? Against what standard? Certainly nothing in the review you tried to cite says anything about that.
Yes, the repair mechanisms are very accurate, so that there’s only around 10^-9 mutations per site per cell generation, and only around 10^-8 mutations per site in the germ line per individual generation. And yet those mutations do occur, enough to account for evolutionary change between species.
Well, it’s no more than expected for the number of cell divisions that happen during gametogenesis. I don’t see you as having a point here. Mutations happen. The observed rate is due to mutation itself minus repair. So?
If you’re trying to pull a Sal Cordova maneuver, why not just cut and paste the picture itself? Note: in science, a picture is not worth a thousand words.
John Harshman,
https://dx.doi.org/10.1158/1535-7163.MCT-15-1002
This is a link to a paper the shows how apoptosis is part of the checkpoint process for repair, if repair fails. Many papers can be obtained on pubmed that support that apoptosis is part of the cell cycle and is triggered by damaged DNA and other problems like hypoxia.
This is supported but I don’t think it is accurate given the low variation between specie’s cells. I think it is low by about 10 orders of magnitude or sustaining complex multicellular life would not be possible.
10^13 cell divisions 3×10^9 nucleotides with 10^9 repair accuracy is a non starter.
Unfortunately, it’s paywalled. All I can see is the abstract, and nothing in the abstract suggests anything like what you are claiming.
You think what is low by about 10 orders of magnitude? I think you meant to say that the true mutation rate is lower by 10 orders of magnitude, i.e. about 10^-19 per site per cell generation. If that were so, then we would never see any mutations at all. You are very confused.
You are confused about what 10^13 cell divisions mean. They aren’t sequential. That’s the total number of mitosis events. But what counts for the number of mutations in a cell lineage is the number of divisions in that lineage, which will generally be no more than a few dozen, if that. Further, the mutation rate per generation can be empirically measured by a number of methods, all of which agree.
John Harshman,
The DNA is needs to be replicated 10^13 times each time it is replicated it has a probability of an error. So if the error rate is 10^9 and it has 10^13 possibilities to make an error the expected error rate would be 10^13 x 3×10^9/10^9. If the error rate is 10^-19 then we can build the animal.
Again, you are counting completely wrong. Your “error” rate is just the total of all mutations in all cell lines, a meaningless number. All the real mutation rate means is that a cell lineage 30 cell generations long is expected to accumulate around 30 mutations. The probability that any one of them will be seriously deleterious is very low. You can wish all you want for an error rate of 10^-19/site/cell generation, but the facts show that it isn’t true. Please accept that your ignorance on this subject is leading you seriously astray.
John Harshman,
The total number of mutations in all cell lines is relevant because it is not feasible for evolution to explore the entire genome. The mechanism will stop those delirious mutations through repair or apoptosis thus improving the repair mechanism orders of magnitude larger then the numbers you propose. If you have a repair of 10^-14 you will explore 10^11 mutations through the entire animal. This is not reasonable and at 10^19 you will have explored 160000 mutations through the animal. A more reasonable number but may not stand up to empirical data.
I don’t know what you mean by “explore the entire genome”; nor do you. The number of mutations “explored through the animal” means nothing. I like the bit about “delirious mutations”, though. Once again, the numbers I propose take into account repair mechanisms; we only see mutations that are not repaired. And apoptosis has nothing to do with it. You are spouting nonsense. Why won’t you believe that?
I think “delirious mutations” are those that lead to creationism.
Hey, I just thought of something. Is it possible that Bill believes in pangenesis, and that somatic cells throughout the body are sending gemmules to the gonads so that all the somatic mutations can be incorporated into the gametes? That would make sense of his numbers. If so, Darwin would be grateful.
And that’s so reasonable! Delerious mutations in brain cells can’t lead to atheism. Praise Darwin!
As usual you have no clue what this gibberish means. There aren’t 10^13 cell divisions (that would result in an incomprehensible number of cells, when are you people going to learn to understand exponents?). There are 10^13 cells. That is 10 trillion cells, which is a good estimate for the total number of cells the adult human body is made of. It didn’t take 10 trillion cell divisions to make 10 trillion cells, because they weren’t made sequentially. 1 becomes 2, 2 becomes 4, 4 becomes 8 and so on. It’s exponential growth.
10^9 isn’t a number of anything. It’s 10^-9, which is the pr. nucleotide pr. replication mutation rate, after repair. It means roughly one in a billion nucleotides will have mutated after a single cell-division of a human cell. With about 3 billion nucleotides, that’s 3 mutations pr replication. AFAIK the actual number is closer to 10^-10 pr nucleotide pr replication.
Now, explain why that’s a “non starter”?
This is reminiscent of discussion on the number of cell divisions ‘needed’ to get a chloroquine resistant Plasmodium strain. Exponentiation, and bottlenecking, seem to be something of a blind spot.
It would be pretty amazing if that were true. What an awesome design!
That was taken care of in a peer-reviewed paper. The same paper Dr Behe got his numbers from.
John Harshman,
The numbers you propose will not build a functional animal. It will result it cancer causing mutations to occur in the womb. What do you think will prevent this if not the repair mechanism tied to the cell cycle? All you need is a wrong mutation in a single APC gene and you have run away cell division. Your numbers result in 10^15 mutational events during embryo development.
What evidence do you have for your contention? So what if there are 10^15 mutations distributed among 10^13 cells? That’s only 100 per cell, about what we observe in every zygote. (And I haven’t actually looked at your calculations, which is probably a mistake considering your track record.) You are complaining that reality isn’t possible. I’m don’t “propose” numbers; I report what we actually see. What’s next? Bumblebees can’t fly?
It is true, at least for the first few cell divisions. Eventually the cells in the middle of the developing “ball” of cells can’t divide due to size constraints, so additional divisions happen in the outer layers, breaking the exponential growth curve.
Why would an exponential growth be a design?
Why?
You just claim this, but don’t back it up with anything other than more assertion.
Bill, doesn’t it some times bother you that you have to just make up stuff to “defend” your views?
What will prevent cancers from occurring at the embryonic development of every zygote? Well for starters, your numbers are wrong.
How likely are “run away cell division” mutations in APC genes? Try to start with correct numbers.
Bill, how the hell do you come up with these numbers? Please show your work.
Frankie,
Wrong. White’s paper to which you refer quoted a completely irrelevant figure: the total number of cell divisions occurring in the entirety of Plasmodium cells anywhere per occurrence of chloroquine resistance. It is much the same mistake as being committed now by colewd – counting the total number of replication events.
You too fail to understand the relevance of exponentiation and bottlenecking.
Seems to be moot now, since the cell division is not in fact exponential. You should not counter one false fact with another false fact, imo.
John Harshman,
The evidence I have is basic math and statistics that is sanitizing your claim to be false by several orders of magnitude. Your claim of 100 per cell is assuming no standard deviation.
I am saying your proposed version of reality is not possible. It does not pass a reasonable review. In order to build a complex animal you need extreme accuracy in the replication process perhaps as high as 10^-20 in order to avoid disease occurring during development. I will send you my excel spread sheet that I used for these estimates.
colewd,
The longest somatic cell lineage is perhaps a few dozen cell divisions in length. You don’t need 1 in 10^20 for that. And, indeed, measured rates, by numerous different means, agree on a range of 10^ -8 to 10^ -10 (the low end being the raw polymerase error rate, higher values when repair is factored in). The information s widely available. You don’t need Excel.
All exponential growths stop being exponential at some point, because physics. That doesn’t make it false that there was an exponential growth. It just means it stopped.
All linear growths also stop at some point, that doesn’t mean there wasn’t a linear growth. What else would you expect?
You don’t know any basic math or statistics. You barely comprehend exponents. Bill, stop making a fool of yourself. Swallow your pride and move on with your life.
ffs LOL, no.
Bill, just stop this. You’re a grown man, this isn’t a schoolyard.
Bill “I make shit up when I have nothing else” Cole is on a roll tonight.
Oh, you put it in an excel spreadsheet? You must be one smart fellow then. I retract all my objections.
Rumraket:
Except for the linear growth of Mung’s Pinocchio nose.
Or is it exponential?
Ouch keiths, that really hurts. It could take me years to get over that.
No, it’s assuming a mean. What effect do you suppose a standard deviation, by which I suppose you mean stochastic variation within some distribution, would have? (I notice your spreadsheet doesn’t feature any such thing). I think you demonstrate here that you know some words and you know how to get an exponential increase out of a spreadsheet, but you have no idea about biology, logic, or statistics.
What do you think your spreadsheet shows? The total number of mutations in all cells in the body is not an interesting number and is not relevant to anything at all. The total number of mutations in a cell lineage after 44 cell generations is, again, easy to see by sequencing the genomes of a parent and sibling, and it isn’t a tiny fraction of 1. And it’s also additive, not exponential; at least your spreadsheet manages to show that. Incidentally, your spreadsheet goes off the rails after generation 35 because you read 0.00102 as 0.0012, and then started adding 0.0003 per generation rather than 0.00003. However did you manage to come up with 129 mutations per cell in the final generation? Can’t you see that the actual number ought to be 43*0.00003=0.00129? You’re 5 orders of magnitude off based on your assumptions, though, oddly, your number is fairly close to the real value.
“A reasonable review” doesn’t mean that you let your prejudices determine what actual data ought to show. That’s backwards. You have to let the data test your hypotheses.
Incidentally, I find that the lowest mutation rate ever observed in any organism is about 2*10^-11 per site per cell generation. Here. How Bill will find an additional 9 orders of magnitude of fidelity is unclear.
John Harshman,
Thanks for the link, John. That picture might be worth a thousand words for me.
Who really cares about the mutation rate? What mutations can and cannot do are the real questions.
The mutation rate matters because without enough mutations you’ll not get enough variation for anything to evolve.
Mung,
Eggsactly!
Larry Moran recently had a blog post on the same subject: http://sandwalk.blogspot.dk/2016/12/learning-about-modern-evolutionary.html
John Harshman,
This is where we disagree. My argument is that every cell division in your case gives a likely probability that 3 mutations occur.
If these mutations are random then each time they can land in the genome with some probability of hitting a mutation that can create a catastrophic event like an APC mutation that does not allow binding to the beta catenin destruction mechanism therefore causing a tumor to form.
I will send you spread sheets with 3 scenarios of correction mechanisms
1. 10^-9
2.10^-14
3.10^-18
Number one says the embryo will experience 10^15 chances for a catastrophic mutation
Number two 5.5^10
Number three 5 million
Molecular perfection? Is this another fine tuning argument in the making!?
colewd,
E pur si muove. You’re still telling me that bumblebees can’t fly. But they can. I’ve seen them. Now in your case, note that the number of chances tells us nothing unless we know the probability of generating your catastrophe per mutation.
Hey, have you ever noticed that you only believe the scientific literature when it tells you something you want to hear? You totally ignored the bit about the lowest observed mutation rate in any organism.
No, according to rumrat it’s the RIGHT mutations that matter. Just take a look at voles. All of that evolution and the voles are still voles:
http://www.purdue.edu/uns/html4ever/2006/060914DeWoodyVole.html
No, they can fly, but only because there are tiny little pixies providing them with additional lift.