Poll: Does the Theory of Evolution have practical applications?

[pitabread]

The Kishony and Lenski experiments demonstrate DNA evolution with the mildest environmental constraints, only a single selection pressure acting. Real-world (non-laboratory) conditions have far more selection pressures acting and those pressures are acting simultaneously. If you want to slow down or stop the Kishony and Lenski experiments, add additional selection pressures. Additional selection condition makes for a more complex evolutionary trajectory. That imposes more instances of the multiplication rule simultaneously. The imposition of multiple simultaneous selection pressures is the condition that makes 3 drug combination therapy work for the treatment of HIV.

I agree that in nature you certainly have a more complex, dynamic environment. Consequently selection pressures will be more variable and fitness effects highly environmentally dependent. As well, you have non-selective evolutionary factors (e.g. constructive neutral evolution) as well.

If you want to argue that the evolution of feathers was comparable to extreme selective pressures like would occur in anti-viral treatments re: HIV, then by all means present evidence to that effect. Otherwise, it seems an unwarranted assumption.

And biologists are using the wrong models in the wrong way to trace molecular evolution.

As I already said, there are certainly known limitations with modeling evolution. In fact, I think if you were to comb the literature you'd probably already find literature discussing just this. I find it odd you seem to be suggesting that biologists are just blindly ignorant in that respect.

 

[FrumiousBandersnatch]

I don't know how many mutations it would take. I've seen papers where they are attempting to identify the control modules for the beta keratin gene and if I recall, so far they've identified 7 or 8. They don't indicate how many bases in these control modules but even if it only requires 1 or 2 mutations in each module, the variant population size for this evolutionary process is already in the range of 7-16 billion. It wouldn't surprise me as further research goes on that it takes many more genes and control modules to make feathers.

I'm not convinced that prokaryote evolution is a particularly good guide to eukaryote evolution, but I'm beginning to suspect that there may be more than just genetics involved. Some relatively simple multicellular organisms use non-genetic mechanisms as developmental pattern guides. How much of these persist in more complex organisms is yet to be established, but it wouldn't surprise me if they play a role.

Going from stem cell to full-grown differentiated adult takes a lot of precise instruction and control. It really takes a wild imagination to think this could happen by evolution once you understand the DNA evolution of bacteria to an antibiotic or to starvation stress.

But evolution and development are quite different processes - ontogeny may roughly recapitulate phylogeny, but the former nominally follows the patterns established by the latter.

You might find the work of Michael Levin interesting; he's shown it's possible to change the developmental patterns of planaria for generations by changing specific metastable patterns of ion-channel activity in the cell membrane. He's also shown that the cytoskeleton encodes part of the developmental pattern for the whole organism.

 

[Alan Kleinman]

I agree that in nature you certainly have a more complex, dynamic environment. Consequently selection pressures will be more variable and fitness effects highly environmentally dependent. As well, you have non-selective evolutionary factors (e.g. constructive neutral evolution) as well.

If you want to argue that the evolution of feathers was comparable to extreme selective pressures like would occur in anti-viral treatments re: HIV, then by all means present evidence to that effect. Otherwise, it seems an unwarranted assumption.



As I already said, there are certainly known limitations with modeling evolution. In fact, I think if you were to comb the literature you'd probably already find literature discussing just this. I find it odd you seem to be suggesting that biologists are just blindly ignorant in that respect.

The mathematics of DNA evolution (the evolutionary trajectory) does not depend on the intensity of selection. The probability of a particular beneficial mutation B occurring on some member of a population that already has beneficial mutation A depends only on the number of replications that those members with beneficial A are able to replicate. The mathematics of the evolution of HIV to antiviral selection pressures is no different than the mathematics for any other replicator to any selection pressure. And dynamic environments with changing selection pressures will only complicate the ability of any lineages ability to adapt to these selection conditions. The reason is that it takes time (generations) to have any chance to adapt to selection pressures. Lenski has done experiments using thermal stress on his bacteria rather than starvation. What he found that there were some variants that had greater fitness at temperatures below 37C and other variants with greater fitness above 37C. If Lensk were to take variants that were best adapted to 37C and put them at a lower temperature, that population would start accumulating mutations that would give improved fitness to that environmental temperature. But if Lenski were to change the environmental temperature to above 37C, a different set of mutations would have to be accumulated and a different evolutionary trajectory would have to be taken.

Whatever the reason, biologists are modeling biological evolution completely wrong. In particular, their Markov models are incorrect because they are based on the wrong assumptions and they are using incorrect transition matrices. And the strategy they use for these models for developing phylogenetic trees is not only wrong, it is silly.

 

[Speedwell]

The mathematics of DNA evolution (the evolutionary trajectory) does not depend on the intensity of selection. The probability of a particular beneficial mutation B occurring on some member of a population that already has beneficial mutation A depends only on the number of replications that those members with beneficial A are able to replicate. The mathematics of the evolution of HIV to antiviral selection pressures is no different than the mathematics for any other replicator to any selection pressure. And dynamic environments with changing selection pressures will only complicate the ability of any lineages ability to adapt to these selection conditions. The reason is that it takes time (generations) to have any chance to adapt to selection pressures. Lenski has done experiments using thermal stress on his bacteria rather than starvation. What he found that there were some variants that had greater fitness at temperatures below 37C and other variants with greater fitness above 37C. If Lensk were to take variants that were best adapted to 37C and put them at a lower temperature, that population would start accumulating mutations that would give improved fitness to that environmental temperature. But if Lenski were to change the environmental temperature to above 37C, a different set of mutations would have to be accumulated and a different evolutionary trajectory would have to be taken.

Yes. And?

 

[Alan Kleinman]

I'm not convinced that prokaryote evolution is a particularly good guide to eukaryote evolution, but I'm beginning to suspect that there may be more than just genetics involved. Some relatively simple multicellular organisms use non-genetic mechanisms as developmental pattern guides. How much of these persist in more complex organisms is yet to be established, but it wouldn't surprise me if they play a role.


But evolution and development are quite different processes - ontogeny may roughly recapitulate phylogeny, but the former nominally follows the patterns established by the latter.

You might find the work of Michael Levin interesting; he's shown it's possible to change the developmental patterns of planaria for generations by changing specific metastable patterns of ion-channel activity in the cell membrane. He's also shown that the cytoskeleton encodes part of the developmental pattern for the whole organism.

The fundamentals of DNA evolution for prokaryotes and eukaryotes are the same. The differences are, eukaryotes are usually diploid so that each member replications represents 2 genome replications, in addition, random recombination under the right circumstances can slightly accelerate DNA evolution but that is rare and it can only accelerate DNA evolution for a single evolutionary step. There's plenty of empirical evidence to support this claim. Consider the evolutionary effects of herbicides and pesticides on weeds and insects, especially when they are used in combination, weeds and insects are, of course, eukaryotes. These selection pressures have the same evolutionary impact on these eukaryotes as antibiotics have on bacteria.

And, of course, differentiation of a stem cell into an adult organism is not evolution. Even most biologists reject "ontogeny recapitulates phylogeny". My point is that the feedback control system for the differentiation of a stem cell into the adult form I suspect is in the non-protein-coding portion of the genome. Biologists like to call this junk DNA. The evolution of a feather-producing replicator from a non-feather-producing replicator requires more than the beta-keratin gene. It requires the correct control system to turn on and off these protein-producing genes. If these control systems don't work correctly, we call these birth-defects and cancers.

 

[Alan Kleinman]

Yes. And?

And these changing evolutionary trajectories would consume many billions more replications with the end result that variant would not be significantly genetically different than the original variant from the beginning of the experiment.

 

[FrumiousBandersnatch]

I agree that in nature you certainly have a more complex, dynamic environment. Consequently selection pressures will be more variable and fitness effects highly environmentally dependent. As well, you have non-selective evolutionary factors (e.g. constructive neutral evolution) as well.

If you want to argue that the evolution of feathers was comparable to extreme selective pressures like would occur in anti-viral treatments re: HIV, then by all means present evidence to that effect. Otherwise, it seems an unwarranted assumption.

As I already said, there are certainly known limitations with modeling evolution. In fact, I think if you were to comb the literature you'd probably already find literature discussing just this. I find it odd you seem to be suggesting that biologists are just blindly ignorant in that respect.

I think the 'retrospective probability' question raised earlier is also important. We simply don't know how many different ways something similar to an existing specialised trait (e.g. feathers) could have evolved, or what other traits might have developed if that particular trait hadn't.

 

[FrumiousBandersnatch]

The fundamentals of DNA evolution for prokaryotes and eukaryotes are the same. The differences are, eukaryotes are usually diploid so that each member replications represents 2 genome replications, in addition, random recombination under the right circumstances can slightly accelerate DNA evolution but that is rare and it can only accelerate DNA evolution for a single evolutionary step. There's plenty of empirical evidence to support this claim. Consider the evolutionary effects of herbicides and pesticides on weeds and insects, especially when they are used in combination, weeds and insects are, of course, eukaryotes. These selection pressures have the same evolutionary impact on these eukaryotes as antibiotics have on bacteria.

I wasn't so much referring to DNA changes but to the magnitude of effects possible from changes to gene or hormone regulation by a single mutation (a simple but striking example being the 'double muscling' of Belgian Blue and Piedmontese cattle).

My point is that the feedback control system for the differentiation of a stem cell into the adult form I suspect is in the non-protein-coding portion of the genome. Biologists like to call this junk DNA. The evolution of a feather-producing replicator from a non-feather-producing replicator requires more than the beta-keratin gene. It requires the correct control system to turn on and off these protein-producing genes. If these control systems don't work correctly, we call these birth-defects and cancers.

Yes; many of these developmental differentiation processes seem to be a self-locating & assembly via intercellular communication and using hormonal gradients and genetic control 'toolkits'; so rather than traits being explicitly coded, it's more of a distributed development schema. This makes it tricky to disentangle, but very flexible.

 

[Alan Kleinman]

I think the 'retrospective probability' question raised earlier is also important. We simply don't know how many different ways something similar to an existing specialised trait (e.g. feathers) could have evolved, or what other traits might have developed if that particular trait hadn't.

It doesn't matter, the mathematics is the same for any evolutionary trajectory. You should start learning how one of these evolutionary trajectories work. Then I'll show you how to do the math for multiple different evolutionary trajectories. It changes the probabilities very little. Here's a paper that demonstrates an empirical example of multiple different evolutionary paths:
Darwinian Evolution Can Follow Only Very Few Mutational Paths to Fitter Proteins
Unless every step in the evolutionary trajectory gives an improvement in fitness, the probability of a lineage taking that path is very low. The reason for that is that it takes a billion replications for each evolutionary step (and that's under the best of circumstances). If there is a single detrimental mutation in that trajectory, the evolutionary process stops there.

 

[Alan Kleinman]

I wasn't so much referring to DNA changes but to the magnitude of effects possible from changes to gene or hormone regulation by a single mutation (a simple but striking example being the 'double muscling' of Belgian Blue and Piedmontese cattle).

Yes; many of these developmental differentiation processes seem to be a self-locating & assembly via intercellular communication and using hormonal gradients and genetic control 'toolkits'; so rather than traits being explicitly coded, it's more of a distributed development schema. This makes it tricky to disentangle, but very flexible.

Sure, if you have an already existing control system, a single mutation in some part of that control system can cause the overproduction of some protein. It is not unusual for cancers to overproduce proteins beyond their normal levels. So, feather production requires the correct control system for the proper use of the proteins needed to produce feathers.

And I think you are right, it is a feedback control system that depends on some type of feedback from the surroundings which signals the genes which do the formation of the proteins to turn on and off. But I'm not sure it is so flexible because an alteration of the feedback control can cause 'double muscling' or some other type of defect. There probably also error checking systems such as seen with the DNA replicate system which can check to see if the given base is being replicated with the same base in the duplicate chromosome. Here's a question for you. How did the helicase and gyrase proteins evolve?

 

[FrumiousBandersnatch]

It doesn't matter, the mathematics is the same for any evolutionary trajectory.

That's the point; any particular trajectory may be incredibly unlikely, but some trajectory in the landscape of all possible trajectories will occur. The questions then are, what percentage of the possible trajectories are broadly neutral or beneficial? on average, how many steps constitute a beneficial or neutral trajectory? what is the minimum number of steps in such trajectories? what is the sample population size? and so-on.

...I'll show you how to do the math for multiple different evolutionary trajectories.

Did I forget to mention that the maths is no longer my competence?

 

[FrumiousBandersnatch]

... I'm not sure it is so flexible because an alteration of the feedback control can cause 'double muscling' or some other type of defect. There probably also error checking systems such as seen with the DNA replicate system which can check to see if the given base is being replicated with the same base in the duplicate chromosome.

By flexible, I meant useful in a variety of ways and for a variety of purposes.

Here's a question for you. How did the helicase and gyrase proteins evolve?

I don't know, it was some time ago IIRC they're found in bacteria & viruses, so perhaps they co-evolved with the development of RNA/DNA as heritable information carriers ¯\_(ツ)_/¯ Not my field, really.

 

[Alan Kleinman]

That's the point; any particular trajectory may be incredibly unlikely, but some trajectory in the landscape of all possible trajectories will occur. The questions then are, what percentage of the possible trajectories are broadly neutral or beneficial? on average, how many steps constitute a beneficial or neutral trajectory? what is the minimum number of steps in such trajectories? what is the sample population size? and so-on.


Did I forget to mention that the maths is no longer my competence?

These evolutionary trajectories only occur in your imagination. The evolutionary trajectory for a 500 amino acid protein (1500 bases) will be 1500 evolutionary steps just for that single gene and protein. For the ideal case of only a single selection pressure operating on the replicator at a time (and who knows what that selection pressure is), that will be about 1.5 trillion replications for some like lineage. Another 20-25000 genes to go and there you go, you have the protein-coding genes for humans. Now all you need are all the control modules for those genes. That's about 98% of the rest of the genome. If you have trouble with the math, that is 1.5e12*2e4*0.98e9=2.94e25 replications to achieve a human genome. Of course, that's based on the assumption that the evolutionary process is due to just a single selection pressure acting at a time. If there are multiple selection pressures acting simultaneously, the number of replications goes up exponentially.

 

[Alan Kleinman]

By flexible, I meant useful in a variety of ways and for a variety of purposes.

I don't know, it was some time ago IIRC they're found in bacteria & viruses, so perhaps they co-evolved with the development of RNA/DNA as heritable information carriers ¯\_(ツ)_/¯ Not my field, really.

But all those ways proteins are used are determined by something. Only about 2% of the human genome codes for proteins, most of the rest of the genome is what determines for what purpose those proteins are used.

Sure, bacteria and viruses have replicase systems. And these systems have dozens of different proteins including helicase and gyrase. What was the selection condition that caused the evolution of this system? It's really not fair for me to ask you this kind of question. You can't even explain the adaptive evolution of the simplest evolutionary experiments. It appears that is not a practical part of the ToE.

 

[FrumiousBandersnatch]

What was the selection condition that caused the evolution of this system?

It's really not fair for me to ask you this kind of question. You can't even explain the adaptive evolution of the simplest evolutionary experiments.

Well, you're the expert, you tell me - what was the selection condition? How did those proteins evolve?

 

[Alan Kleinman]

Well, you're the expert, you tell me - what was the selection condition? How did those proteins evolve?

The is no scientific explanation of how proteins evolve de novo. There is a scientific explanation of how adaptive mutations are accumulated in a lineage subject to a selection pressure verified by evolutionary experimental evidence. Sadly, biologists don't teach this to naive students to the detriment of people suffering from drug-resistant infections and failed cancer treatments.

 

[Frank Robert]

The Kishony and Lenski experiments are performed under the mildest selection conditions possible, a single selection condition only. Real environments have multiple selection conditions operating simultaneously. Multiple simultaneous selection pressures acting on a population impose multiple instances of the multiplication rule in the evolutionary process. That's why 99% of all species have gone extinct. If selection conditions in evolutionary experiments are made more complex, the experiments don't work. That's why Kishonyl can't get his experiment to work with two drugs.

Species go extinct for many reasons such as climate change, reduced food supply, an asteroid crashing into the earth, etc. Happily some are ancestral to younger species. I have two decedents of dinosaurs ancestors, an African Grey and a Cockatoo.

I think the reason why Kishonly isn't able to get positive results is what most scientists will admit: they do not know everything.

 

[FrumiousBandersnatch]

The is no scientific explanation of how proteins evolve de novo. There is a scientific explanation of how adaptive mutations are accumulated in a lineage subject to a selection pressure verified by evolutionary experimental evidence. Sadly, biologists don't teach this to naive students to the detriment of people suffering from drug-resistant infections and failed cancer treatments.

OK, so you can't explain it either. Meh.

You were right, it was 'really not fair' for you to ask me this kind of question - because you knew you couldn't explain it yourself - but you used it to get in an unnecessary dig about me not being able to do something I was never asked to do.

I now see where you're coming from.

 

[sesquiterpene]

These evolutionary trajectories only occur in your imagination. The evolutionary trajectory for a 500 amino acid protein (1500 bases) will be 1500 evolutionary steps just for that single gene and protein.

Is that for a protein being built up one AA at a time? Nobody thinks it happens that way.

If you have trouble with the math, that is 1.5e12*2e4*0.98e9=2.94e25 replications to achieve a human genome.

Is that the total populations of all our ancestors, summed over time? That might be achievable

 

[Alan Kleinman]

Species go extinct for many reasons such as climate change, reduced food supply, an asteroid crashing into the earth, etc. Happily some are ancestral to younger species. I have two decedents of dinosaurs ancestors, an African Grey and a Cockatoo.

I think the reason why Kishonly isn't able to get positive results is what most scientists will admit: they do not know everything.

What did you do, send in the DNA from your birds to ancestry.com? And what positive result didn't come out from the Kishony experiment? Biologists certainly don't admit that they don't understand the physics and mathematics of evolution. If they did understand that physics and mathematics, they could explain the results of the Kishony and Lenski experiments. Those experiments are totally predictable. You just won't find a biologist doing it.