Evolution is mathematically impossible

[tas8831]

All legitimate science is founded on mathematics. Evolution is mathematically impossible.
maskofscience.com

Yeah yeah, saw your website. Good one, funny stuff. You and ReMine should get together.

If you don't understand binomial distribution, then just admit it. But don't pretend that you can just dismiss mathematical proof.

Do you understand the concept of GIGO?

Your 'equations' are super great, the best ever!

But the Garbage that you put in?

It is almost as if you have NO CLUE whatsoever about how genetics OR evolution operates.

I get it - you have an ego and a faith to prop up, but pretending to have made some grand defeat of evolution based on pretty fish and some dopey calculation about two identical brand new 600 base genes... Whatever...

"In reality, at least many tens of thousands of specific nucleotide substitutions would have been required."​

Reality - let's hear it, junior!

If you contend that I haven't considered other variables, what are they? How can I change the numbers to make even one tiny step of evolution possible?

You can stay up to date?

No Genome-Wide Protein Sequence Convergence for Echolocation
Zhengting Zou Jianzhi Zhang
Author Notes
Molecular Biology and Evolution, Volume 32, Issue 5, 1 May 2015, Pages 1237–1241, No Genome-Wide Protein Sequence Convergence for Echolocation
Published:
27 January 2015

Abstract
Toothed whales and two groups of bats independently acquired echolocation, the ability to locate and identify objects by reflected sound. Echolocation requires physiologically complex and coordinated vocal, auditory, and neural functions, but the molecular basis of the capacity for echolocation is not well understood. A recent study suggested that convergent amino acid substitutions widespread in the proteins of echolocators underlay the convergent origins of mammalian echolocation. Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages. The same is true for the group of 29 hearing-related proteins claimed to be enriched with molecular convergence. Reexamining the previous selection test reveals several flaws and invalidates the asserted evidence for adaptive convergence. Together, these findings indicate that the reported genomic signatures of convergence largely reflect the background level of sequence convergence unrelated to the origins of echolocation.


So sorry to burst your bubble. If you don't understand basic genetics and evolutionary theory, just admit it.
I await the removal of your website.

 

[pitabread]

All legitimate science is founded on mathematics. Evolution is mathematically impossible.

A mere assertion gets you nowhere.

If you don't understand binomial distribution, then just admit it. But don't pretend that you can just dismiss mathematical proof.

I fully understand binomial distribution (University statistics boi) and I also fully understand why you are abusing it in this instance.

What appears to be lacking is an understanding of biological evolution on your end.

If you contend that I haven't considered other variables, what are they? How can I change the numbers to make even one tiny step of evolution possible?

It's not a question of changing the numbers, it's that the fundamental premise is flawed.

You're trying to model evolution of something from scratch based on pure random chance. That's not applicable in this scenario. As has already been pointed out earlier in the thread, the actual starting point for the co-evolution of genes would be common ancestral genes and the number of changes required far fewer than what you have claimed.

You'd also need to account for natural selection, which is the opposite of pure random chance, but you don't have enough information to do so.

On top of that, you're effectively trying to model a very specific scenario after the fact. As was also pointed out earlier, evolution doesn't specifically target certain outcomes; it simply produces viable biological forms whatever those may be. To actually model the probability of evolution, you'd need to know the total probability space of all viable outcomes. But you don't have that information either.

Basically this whole exercise is just a case of GIGO: garbage in, garbage out.

By all means, though, keep claiming you've disproved biological evolution. I'm sure all of the world's biologists will love to hear it. /s

 

[tas8831]

Poor John the pathologist - didn't even notice the sound refutations in the first 2 replies...

Evolution is mathematically impossible

Evolution is mathematically impossible

Not sure which is sadder - that this guy actually thought he had made some kind of legitimate refutation of evolution, or that now that his masterpiece has been demolished, he will, almost certainly, double-down and continue to claim on other sites that he has this great disproof....

 

[sfs]

If you don't understand binomial distribution, then just admit it. But don't pretend that you can just dismiss mathematical proof. If you contend that I haven't considered other variables, what are they? How can I change the numbers to make even one tiny step of evolution possible?

You really should learn something about the field you're basing your argument on -- in this case, population genetics. Here's one of your numerous errors: "The probability of a mutation becoming fixed in the population is 1 in 500^5. To give evolution an advantage, I will assume that twice the number of mutations (1 in 250) will become fixed in the population." The probability of a neutral mutation being fixed in a population 1/2N, where N is the effective population size. The probability of a selectively advantageous mutation -- which is what we're talking about here -- is roughly 2 times the selective advantage (2% for a mutation that gives a 1% advantage). So right there, you're off by 10 or 11 orders of magnitude.

As I pointed out above, you're also calculating something that has no connection to the actual story. Only a couple of dozen mutations were fixed in the different lineages, not a mutation at every base pair.

This definitely goes in the "not even wrong" file.

 

[sfs]

You can stay up to date?

No Genome-Wide Protein Sequence Convergence for Echolocation

Ouch.

 

[John B. Andelin]

Lets see YOUR math. I don't mean the pap on your website, I mean legitimate, realistic math.


Wait - are you implying that they have the same genes?



Your claim is silly and not a legitimate argument. Nothing to refute.


"I have only calculated a tiny element of what would be required to create the complex genetic code for echolocation (600 nucleotides)."


Where on earth did you get that from?
Look, I was in a graduate program in a medical school in the mid 1990s. I know that the medical students there had but a 2 week crash course in basic genetics, and given your bio, I'm betting you had less than that. It seems pretty clear to me that you are not up to the task as far as genetics goes, and you sure are out in left field evolution-wise, too. Yours is the egotist's argument, not a legitimate 'challenge.'

Attacking my credentials does not constitute disproof of a mathematical challenge.

 

[John B. Andelin]

You really should learn something about the field you're basing your argument on -- in this case, population genetics. Here's one of your numerous errors: "The probability of a mutation becoming fixed in the population is 1 in 500^5. To give evolution an advantage, I will assume that twice the number of mutations (1 in 250) will become fixed in the population." The probability of a neutral mutation being fixed in a population 1/2N, where N is the effective population size. The probability of a selectively advantageous mutation -- which is what we're talking about here -- is roughly 2 times the selective advantage (2% for a mutation that gives a 1% advantage). So right there, you're off by 10 or 11 orders of magnitude.

As I pointed out above, you're also calculating something that has no connection to the actual story. Only a couple of dozen mutations were fixed in the different lineages, not a mutation at every base pair.

This definitely goes in the "not even wrong" file.

If all the mutations we're referring to are advantageous (a single nucleotide being "advantageous" is certainly unproven), then you've raised your odds to around 106^-401. I have only assumed that e

A mere assertion gets you nowhere.



I fully understand binomial distribution (University statistics boi) and I also fully understand why you are abusing it in this instance.

What appears to be lacking is an understanding of biological evolution on your end.



It's not a question of changing the numbers, it's that the fundamental premise is flawed.

You're trying to model evolution of something from scratch based on pure random chance. That's not applicable in this scenario. As has already been pointed out earlier in the thread, the actual starting point for the co-evolution of genes would be common ancestral genes and the number of changes required far fewer than what you have claimed.

You'd also need to account for natural selection, which is the opposite of pure random chance, but you don't have enough information to do so.

On top of that, you're effectively trying to model a very specific scenario after the fact. As was also pointed out earlier, evolution doesn't specifically target certain outcomes; it simply produces viable biological forms whatever those may be. To actually model the probability of evolution, you'd need to know the total probability space of all viable outcomes. But you don't have that information either.

Basically this whole exercise is just a case of GIGO: garbage in, garbage out.

By all means, though, keep claiming you've disproved biological evolution. I'm sure all of the world's biologists will love to hear it. /s

I have not assumed that each gene was starting out from scratch. I have only calculated the probability of one changed codon per converged gene. That is giving evolution a huge benefit of the doubt. Obviously, the creation of echolocation would require more than 600 changed nucleotides. And I have adjusted your contention that "advantageous" mutations have a 2% chance of being selected... that raises the odds to around 10^-401.
You state that evolution doesn't have an end "target".
It is impossible for random mutations to create nearly identical sequences, regardless of whether or not there's a "target". It's the same principle that tells us with mathematical certainty that no two snowflakes are exactly alike. There are simply too many possible alternative configurations. If you think evolution could create 200 convergent genes in different species, you need to factor in that there will be well over 10^400 other substitutions that natural selection will have to wade through.
Predictably, you try to justify evolution by assuming that all unknowns will fall on your side. I have allowed for unknowns and have given evolution a huge advantage. My challenge is to tell me what more concessions you need.
Your statement that all the world's biologists would love to see evolution disproven is not remotely true and is irrelevant to this discussion.

 

[John B. Andelin]

Ouch.

Simply posting a link to an article is pointless. You obviously selectively believe whatever falls on the side of evolution. The article I referenced (Genome-wide signatures of convergent evolution in echolocating mammals) is a peer-reviewed study which concluded that the degree of molecular convergence between bats and dolphins is striking. That convergence may not be seen in protein sequences, as indicated in the article you posted. Protein sequences represent only a tiny particle of echolocation. My calculations only consider 600 identical nucleotide sequences spread over 200 convergent genes.

 

[John B. Andelin]

A mere assertion gets you nowhere.



I fully understand binomial distribution (University statistics boi) and I also fully understand why you are abusing it in this instance.

What appears to be lacking is an understanding of biological evolution on your end.



It's not a question of changing the numbers, it's that the fundamental premise is flawed.

You're trying to model evolution of something from scratch based on pure random chance. That's not applicable in this scenario. As has already been pointed out earlier in the thread, the actual starting point for the co-evolution of genes would be common ancestral genes and the number of changes required far fewer than what you have claimed.

If you understand binomial distribution as you claim, then you will understand that it is impossible for random mutations to produce 600 specific nucleotide sequences in two separate lineages, when each of those lineages has a 10^401 probability of occurring. Your only possible refutation of this claim is to logically argue that the numbers I've presented such as population size, probability of mutation, probablity of fixation in the population, etc. are incorrect.

 

[pitabread]

Predictably, you try to justify evolution by assuming that all unknowns will fall on your side. I have allowed for unknowns and have given evolution a huge advantage. My challenge is to tell me what more concessions you need.

I didn't say all the unknowns were on "my side" (whatever that is supposed to mean). I said you are working with a scenario where you don't have all the available information and therefore cannot calculate any sort of meaningful probability.

You can keep running in circles with this, but it's still a case of GIGO.

Your statement that all the world's biologists would love to see evolution disproven is not remotely true and is irrelevant to this discussion.

I put a sarcasm tag at the end. You must've missed it.

 

[pitabread]

If you understand binomial distribution as you claim, then you will understand that it is impossible for random mutations to produce 600 specific nucleotide sequences in two separate lineages

Sure. Good thing evolution doesn't work purely by random mutations. And this is the problem: you're not actually modeling real-world evolution.

You can keep babbling on about improbability of your fictional scenario all you want, but it's not going to make it any more relevant.

 

[tas8831]

Attacking my credentials does not constitute disproof of a mathematical challenge.

Your lack of credentials helps explain why your 'challenges' are bogus and also why YOU think they are amazing.

I note that you have studiously ignored the refutations of your claims, like a good little creationist egotist.



No Genome-Wide Protein Sequence Convergence for Echolocation
Zhengting Zou Jianzhi Zhang
Author Notes
Molecular Biology and Evolution, Volume 32, Issue 5, 1 May 2015, Pages 1237–1241, No Genome-Wide Protein Sequence Convergence for Echolocation
Published:
27 January 2015

Abstract
Toothed whales and two groups of bats independently acquired echolocation, the ability to locate and identify objects by reflected sound. Echolocation requires physiologically complex and coordinated vocal, auditory, and neural functions, but the molecular basis of the capacity for echolocation is not well understood. A recent study suggested that convergent amino acid substitutions widespread in the proteins of echolocators underlay the convergent origins of mammalian echolocation. Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages. The same is true for the group of 29 hearing-related proteins claimed to be enriched with molecular convergence. Reexamining the previous selection test reveals several flaws and invalidates the asserted evidence for adaptive convergence. Together, these findings indicate that the reported genomic signatures of convergence largely reflect the background level of sequence convergence unrelated to the origins of echolocation.


So sorry to burst your bubble. If you don't understand basic genetics and evolutionary theory, just admit it.
I await the removal of your website.

 

[tas8831]

Simply posting a link to an article is pointless. You obviously selectively believe whatever falls on the side of evolution. The article I referenced (Genome-wide signatures of convergent evolution in echolocating mammals)

Hey, Superstar - this article was also peer-reviewed and published in a mainstream journal:

No Genome-Wide Protein Sequence Convergence for Echolocation
Zhengting Zou Jianzhi Zhang
Author Notes
Molecular Biology and Evolution, Volume 32, Issue 5, 1 May 2015, Pages 1237–1241, No Genome-Wide Protein Sequence Convergence for Echolocation
Published:
27 January 2015

Abstract
Toothed whales and two groups of bats independently acquired echolocation, the ability to locate and identify objects by reflected sound. Echolocation requires physiologically complex and coordinated vocal, auditory, and neural functions, but the molecular basis of the capacity for echolocation is not well understood. A recent study suggested that convergent amino acid substitutions widespread in the proteins of echolocators underlay the convergent origins of mammalian echolocation. Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages. The same is true for the group of 29 hearing-related proteins claimed to be enriched with molecular convergence. Reexamining the previous selection test reveals several flaws and invalidates the asserted evidence for adaptive convergence. Together, these findings indicate that the reported genomic signatures of convergence largely reflect the background level of sequence convergence unrelated to the origins of echolocation.


So sorry to burst your bubble. If you don't understand basic genetics and evolutionary theory, just admit it.
I await the removal of your website.​

and was in direct response to yours:

"Comparing 22 mammalian genome sequences, Parker et al. (2013) reported hundreds of convergently evolving proteins among echolocators and suggested that genome-wide molecular convergence underlay the origins of echolocation and associated phenotypes. However, protein convergence could also occur by chance (Zhang and Kumar 1997; Castoe et al. 2009). Here, we show that the reported genomic signatures of convergence largely reflect such chance events that are unrelated to the origins of echolocation."

Parker et al. (2013) is the paper YOU CITED.


Stop your whining and admit your errors. Be a man, not a snowflake.
The foundation of your bogus math has been overturned.
GET UP TO SPEED.

 

[tas8831]

If you understand binomial distribution as you claim, then you will understand that it is impossible for random mutations to produce 600 specific nucleotide sequences in two separate lineages, when each of those lineages has a 10^401 probability of occurring. Your only possible refutation of this claim is to logically argue that the numbers I've presented such as population size, probability of mutation, probablity of fixation in the population, etc. are incorrect.

You keep avoiding the fact that what you based your unrealistic scenario calculation on has been refuted:

No Genome-Wide Protein Sequence Convergence for Echolocation
Zhengting Zou Jianzhi Zhang
Author Notes
Molecular Biology and Evolution, Volume 32, Issue 5, 1 May 2015, Pages 1237–1241, No Genome-Wide Protein Sequence Convergence for Echolocation
Published:
27 January 2015

Abstract
Toothed whales and two groups of bats independently acquired echolocation, the ability to locate and identify objects by reflected sound. Echolocation requires physiologically complex and coordinated vocal, auditory, and neural functions, but the molecular basis of the capacity for echolocation is not well understood. A recent study suggested that convergent amino acid substitutions widespread in the proteins of echolocators underlay the convergent origins of mammalian echolocation. Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages. The same is true for the group of 29 hearing-related proteins claimed to be enriched with molecular convergence. Reexamining the previous selection test reveals several flaws and invalidates the asserted evidence for adaptive convergence. Together, these findings indicate that the reported genomic signatures of convergence largely reflect the background level of sequence convergence unrelated to the origins of echolocation.


So sorry to burst your bubble. If you don't understand basic genetics and evolutionary theory, just admit it.
I await the removal of your website.​

and was in direct response to yours:

"Comparing 22 mammalian genome sequences, Parker et al. (2013) reported hundreds of convergently evolving proteins among echolocators and suggested that genome-wide molecular convergence underlay the origins of echolocation and associated phenotypes. However, protein convergence could also occur by chance (Zhang and Kumar 1997; Castoe et al. 2009). Here, we show that the reported genomic signatures of convergence largely reflect such chance events that are unrelated to the origins of echolocation."

Parker et al. (2013) is the paper YOU CITED.


You also keep ignoring the first 2 replies in this thread that demolished your claims. But being refuted on multiple fronts has never stopped creationists with irrelevant backgrounds from driving on and on.

 

[sfs]

Simply posting a link to an article is pointless.

Note that my "ouch" referred to the demolition job done by the new paper on the old.

The article I referenced (Genome-wide signatures of convergent evolution in echolocating mammals) is a peer-reviewed study which concluded that the degree of molecular convergence between bats and dolphins is striking.

It concluded that the degree of molecular convergence at the protein level was striking.

That convergence may not be seen in protein sequences, as indicated in the article you posted.

In which case, the original paper was wrong and your calculation is pointless.

(a single nucleotide being "advantageous" is certainly unproven)

Say what?

My calculations only consider 600 identical nucleotide sequences spread over 200 convergent genes.

Ah, now I see what you're doing. In that case, your calculation is merely wrong. First, you assume that 200 specific changes, and only those changes, are possible targets for convergent evolution. You have no idea what the possible target space might be. Second, you assume that all three nucleotides have to change to produce a convergent amino acid. In reality, anywhere from one to three is required, depending on the starting and ending codons. Given neutral variation in the population, and the fact that the genes are homologous to begin with, and that the most probable convergent mutations are the ones we'll actually see (out of the unknown possible set -- see point 1), most of the time only one nucleotide change would be required. Suppose only one is required. Given your own estimate of the total number of mutations in dolphin, each site mutated 15,840 times, which means 5280 mutations to a particular base, assuming a uniform mutation matrix. (Note: in reality, more than one different mutation might lead to the amino acid.) If the probability of fixation really were 2%, then the probability of any given amino acid being fixed in one lineage would be 1 - .98^5280, or 1 - 4.7x10^-47. Using that value, the probability that 200 AAs would all fix in a given lineage is essentially 1.0, meaning that the same 200 would fix in multiple lineages as well. Using a more realistic guess of selective advantage -- 0.1% instead of 1% -- the probability of exactly 200 AA changes fixing in 2 lineages becomes 99.5%.

 

[xianghua]

Since we know that evolution has happened

but we dont know that.

 

[Speedwell]

but we dont know that.

We do. We know that life has changed and diversified since it began. We know that there are species alive now which have not always been here. We know that there were species once alive that are no longer. What part of that would you challenge? It's a fact.

No, we are not arguing about the fact that evolution has taken place. What we are talking about is the mechanism of evolution. The theory of evolution proposes a mechanism which is plausible and well-supported by evidence. You have given us nothing but the unsupported assertion that the mechanism proposed by the theory of evolution is inadequate to explain the fact of evolution. Why don't you come up with a mechanism of your own? I'm sure that would be interesting to talk about.

 

[HitchSlap]

but we dont know that.

You don't know.

 

[Warden_of_the_Storm]

but we dont know that.

But we do.

 

[John B. Andelin]

Note that my "ouch" referred to the demolition job done by the new paper on the old.

It concluded that the degree of molecular convergence at the protein level was striking.

In which case, the original paper was wrong and your calculation is pointless.

Say what?

Ah, now I see what you're doing. In that case, your calculation is merely wrong. First, you assume that 200 specific changes, and only those changes, are possible targets for convergent evolution. You have no idea what the possible target space might be. Second, you assume that all three nucleotides have to change to produce a convergent amino acid. In reality, anywhere from one to three is required, depending on the starting and ending codons. Given neutral variation in the population, and the fact that the genes are homologous to begin with, and that the most probable convergent mutations are the ones we'll actually see (out of the unknown possible set -- see point 1), most of the time only one nucleotide change would be required. Suppose only one is required. Given your own estimate of the total number of mutations in dolphin, each site mutated 15,840 times, which means 5280 mutations to a particular base, assuming a uniform mutation matrix. (Note: in reality, more than one different mutation might lead to the amino acid.) If the probability of fixation really were 2%, then the probability of any given amino acid being fixed in one lineage would be 1 - .98^5280, or 1 - 4.7x10^-47. Using that value, the probability that 200 AAs would all fix in a given lineage is essentially 1.0, meaning that the same 200 would fix in multiple lineages as well. Using a more realistic guess of selective advantage -- 0.1% instead of 1% -- the probability of exactly 200 AA changes fixing in 2 lineages becomes 99.5%.

I appreciate your effort in explaining this. I really do. You obviously have a lot of knowledge of proposed evolutionary processes. If I relied on a paper that has been refuted, then I stand corrected. I was under the impression that convergent nucleotide sequences had been documented.
There is obviously a lot more to echolocation than correct amino acid sequences. You mentioned that there is significant molecular homology between echolocating and non-echolocating species. In reference to the three genes that are involved in the auditory aspect of echolocation (a total of about 21,000 nucleotides...referenced in several papers) what do you suppose the nucleotide difference is, for example between echolocating and non-echolocating bats? If there is substantial molecular homology, then that means that few nucleotide sequences are responsible for the extremely complex adaptation of that one part of echolocation. If this is true, then surely specific nucleotide substitutions would have been required. That is not mathematically possible.
I think a core belief among proponents of NDT is that numerous pathways to complexity exist. If echolocation occurred in three separate lineages, then it is believed that evolution could have taken any one of a huge number of routes to reach those endpoints. If only one or a few pathways were possible, then convergence would be unthinkable. This leads me to another question... in human evolution... since there are about 3.2 billion nucleotide pairs, the probability of a point mutation replacing a specific nucleotide with a specific base is 1 x 10^-10. If a SV of .1 is accepted, this suggests a probability a specific mutation to occur and become fixed in the population (according to Ronald Fischer) to be about 1 x 10^-10 / 500 = 2 x 10^-13.
If you assume that humans evolved over 6 million years from a population of about 10,000 individuals, and assume reproduction every ten years, that would be 600,000 generations. That we means that a total of 600,000 x 150 mutations/generation x 10,000 = 900 billion mutations. If evolution of man required millions of mutations, how do you account for the fact that one specific mutation would require an estimated 2 trillion births to accomplish? Do you think that the gradual evolution of intelligence did not require a single specific nucleotide substitution out of the millions that occurred?