Methinks it is sort-of like two weasels

[lee_merrill]

@Mercer

By combining genetics with biochemistry, as I have done throughout my career, we know that binding is routinely created, altered, or destroyed by changing single amino-acid residues.

Here is more of Behe’s argument (have you read it?), with references:

“The general results from Winter’s lab have been consistently confirmed: In order to get a particular protein to bind to any other one with modest strength, on average you have to wade through about ten to a hundred million binding sites.” (The Edge of Evolution, p. 132)

“So one way to get a new binding site would be to change just five or six amino acids in a coherent patch in the right way. This very rough estimation fits nicely with studies that have been done on protein structure. [13]” (The Edge of Evolution, p. 134)

“13.Even though protein-binding sites often involve a score of amino acids on each of the partners, experiments have shown that only a fraction of those are important for having the two proteins stick to each other. (For example, see Braden, B. C., and Poljak, R. J. 1995. Structural features of the reactions between antibodies and protein antigens. FASEB J. 9:9–16; Lo Conte, L., Chothia, C., and Janin, J. 1999. The atomic structure of protein-protein recognition sites. J. Mol. Biol. 285:2177–98; Ma, B., Elkayam, T., Wolfson, H., and Nussinov, R. 2003. Protein-Protein-protein interactions: structurally conserved residues distinguish between binding sites and exposed protein surfaces. Proc. Natl. Acad. Sci. USA 100:5772–77.) In terms of the swimming pool analogy, the five or six residues represent bumps and magnets that are aligned very nicely; if enough are aligned, then it doesn’t matter so much if other features aren’t aligned, as long as they don’t actively block the surfaces from coming together.” (The Edge of Evolution)

@Mercer

Anyone who has ever purified any protein (even undergraduates, @Eddie) is well aware that the greatest problem is that of the protein aggregating NONspecifically (binding to itself).

And Behe is interested in new beneficial interactions.

[John_Harshman]

@lee_merrill

But they do estimate about 2000 steps, correct?

Yes. And?

@lee_merrill

Not so much, if the orb is slowly changing, these steps would seem to be needed in sequence.

That’s because you aren’t thinking about multi-locus, quantitative characters. Consider height, and say it’s controlled by 10 genes, each with a short allele and a tall allele. And say that height depends on the number of tall alleles. The higher the frequency of any tall allele, the taller the population mean; in fact the entire height distrubution shifts depending on frequencies at each locus.

@lee_merrill

Because neural circuits are complex, on average each neuron connects to 10,000 other neurons, as I have heard.

Some are complex, some aren’t, and they can increase in complexity (more properly, acuity) through selection. I’d like to see documentation on that numcerical claim, which presumably applies to humans only.

@lee_merrill

I’ve been arguing all along here for design, in various ways. Behe’s edge of evolution is perhaps the clearest way to detect design.

That’s not an argument, though I can see how you might think so.

[lee_merrill]

@Rumraket

it’s very strange for Behe to think a new protein binding site is beyond the edge of evolution while simultaneously admitting it has evolved in HIV.

No, two new protein-protein binding sites, he places outside the edge of evolution.

@Rumraket

… and that a considerable fraction of natural proteins are something like one mutation away from developing a binding site.

But that would seem to gum up the works, if new binding sites were so profligate.

@Rumraket

If Behe was right and new protein-protein binding sites were prohibitively unlikely to evolve in the history of life on Earth, the adaptive immune system would be impossible, and so would the entire biotech industry that relies on immunoassays…

But the immune system is not evolving new binding sites, and the biotech industry relies on design.

@Rumraket

But what we find instead supports what I’ve already said. Most proteins have some weak background affinity for each other, so much so that if improvement in this binding ability is selectively beneficial, it is trivial to select for it…

“Suppose that thousands of objects from the shape space library are placed in a well-stirred swimming pool, and a prime goal for each is to find its one ideal mate. Reaching that goal faces a ‘Goldilocks’ problem. First, consider objects that bind indiscriminately. Suppose some shape-space objects weren’t very rigid—they had flexible octopus arms lined with magnets that allowed them to stick to many other objects in the swimming pool. Although they’d bind strongly, those stick-to-everything objects would gum up the works. The cell cannot tolerate objects that bind haphazardly. [6] They must be eliminated.” (The Edge of Evolution)

“6.A properly working protein will bind to just one or a few specific partners. However, because the interiors of most folded proteins are oily, if a protein accidentally unfolds (“denatures”), it might stick to almost everything in sight. Nonspecific aggregation is almost always detrimental to a cell (Bucciantini, M., Giannoni, E., Chiti, F., Baroni, F., Formigli, L., Zurdo, J., Taddei, N., Ramponi, G., Dobson, C. M., and Stefani, M. 2002. Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases. Nature 416:507–11). Cells take great care to avoid it, with systems that dispose of misfolded proteins. So such nonspecific aggregation is not a model for how functional, specific, protein-protein interactions could develop in the cell.” (The Edge of Evolution, p. 288)

[Rumraket (Mikkel R.)]

@lee_merrill

@Rumraket

it’s very strange for Behe to think a new protein binding site is beyond the edge of evolution while simultaneously admitting it has evolved in HIV.

No, two new protein-protein binding sites, he places outside the edge of evolution.

And for reasons already stated that is obviously false.

@lee_merrill

@Rumraket

… and that a considerable fraction of natural proteins are something like one mutation away from developing a binding site.

But that would seem to gum up the works, if new binding sites were so profligate.

Uhh yes, which is why things like protein aggregation is a huge problem for cells, and they have systems that detect, suppress, and eliminate aggregating proteins. It can also be shown through phylostratigraphy that selection reduces such randomly occurring binding sites over time, both in their number and strength over time. Scientists have shown that aggregation propensity negatively correlates with protein age. That is to say, proteins that can be traced further back in time are also less aggregation prone, and younger proteins are more prone to aggregation (because selection has had less time to suppress those inherent mutual-binding properties that proteins have).

@lee_merrill

@Rumraket

If Behe was right and new protein-protein binding sites were prohibitively unlikely to evolve in the history of life on Earth, the adaptive immune system would be impossible, and so would the entire biotech industry that relies on immunoassays…

But the immune system is not evolving new binding sites

Yes it is. That is literally what the immune system does. The hypervariable region of the antibody is where new binding sites develop when the immune system encounters novel antigens. That is a textbook example of a new protein-protein binding site. By any rational definition of new binding site, the ones that develop on antibodies that develop affinity towards novel antigens would qualify.

This is where the article I linked earlier becomes important for you to have read and understood. Here it is again:

Weak Chemical Interactions That Drive Protein Evolution: Crowding, Sticking,...

There is some low-level background affinity that all proteins will unavoidably exhibit towards each other, and to evolve a binding site is essentially just a matter of selecting for mutants that increase this affinity. That means any particular accessible spot on the surface of a protein can in principle develop into a binding site specific for another matching protein. When you understand that all proteins have this sort of inherent stickiness to them (and that it comes in degrees ranging from very weak/transient associations, all the way to basically inseparable), you can take that understanding to the immune system and see how even a very weakly interacting antibody protein can be improved towards a particular antigen, and thus develop a new (it’s different from what it was before and binds another target molecule, therefore new) and effective binding site.

@lee_merrill

and the biotech industry relies on design.

So you don’t know how new antibodies are developed? They literally inject novel compounds into animals, and then harvest the novel antibody molecules that evolve in response.

[Roy (34-year veteran)]

@lee_merrill

@Rumraket

So what justification do you have for thinking it is .1 in the entire history of life? Or whatever you think is the real number, and why.

Let’s assume each step in the evolution of the eye is 99% probable, …

Let’s not.

But even if we did, your maths is wrong:

This results in a 1 in 6.5 x 10^27 chance for the eye to evolve, even if every step is 99% probable! Now there have been about 10^40 total organisms in all life’s history, as I have heard, so for the eye to evolve twice (never mind 50 times), we would need more organisms than have been present on earth.

Having an eye evolve twice needs twice the number of organisms, since they would be evolving in parallel, not in series. Under your assumptions that’d require 1.3e28 organisms, far less that 1e40, and not 4.225e55 organisms.

Not even your invented numbers lead to the conclusion you’re after.

[T_aquaticus]

@colewd

You are assuming your conclusion. All you are observing is a pattern. It evolved is a hypothesis.

Evolution is supported by mountains of evidence. It’s a conclusion.

@colewd

A competent ID proponent would not make this claim.

Then there aren’t any competent ID proponents.

@colewd

Spliceosome data.pdf

What about it?

[Rumraket (Mikkel R.)]

@T_aquaticus

@colewd

You are assuming your conclusion. All you are observing is a pattern. It evolved is a hypothesis.

Evolution is supported by mountains of evidence. It’s a conclusion.

And incidentally we literally see evolution produce nested hierarchies in real time. Most apposite is the real-time tracking of the evolution of the coronavirus:
https://nextstrain.org/ncov/gisaid/global/6m

@T_aquaticus

@colewd

Spliceosome data.pdf

What about it?

He has linked that .pdf many times before accompanied with the same assertion. He refuses to elaborate and seem completely unable to explain how or why that figure is supposed to constitute evidence for design or a problem with evolution. Bill is a cargo-cult debating opponent. He thinks this is some sort of game where all he has to do is type responses that give the appearance of having some sort of fancy technical content. It’s like Sal Cordova who loved to do nothing but post huge tables full of abbreviations and numbers, and colored arrows cris-crossing all over the place.

See this? It’s a figure, it looks complicated, and it’s got numbers and colors on it. Me smart, me scientist. Why won’t plane land here?

[Mercer (John Mercer)]

[quote=“lee_merrill, post:228, topic:15043”]

@Mercer

@Dan_Eastwood

Behe calculates the rate at which resistance arises in a chloroquine- treated population . See the difference?
[/quote]

It’s worse than that! The intermediate hosts were not being treated with chloroquine, so the selection was intermittent. Behe pretends that it was constant.

No, “Nicholas White of Mahidol University in Thailand points out that if you multiply the number of parasites in a person who is very ill with malaria times the number of people who get malaria per year times the number of years since the introduction of chloroquine, then you can estimate that the odds of a parasite developing resistance to chloroquine is roughly one in a hundred billion billion.16 In shorthand scientific notation, that’s one in 10^20.” (The Edge of Evolution, p. 57)

Which has nothing to do with the mutations. Did you not read what I wrote? The selection was intermittent, not constant.

@lee_merrill

Here is the relevant quote from White’s paper: " Resistance to chloroquine in P. falciparum has arisen spontaneously less than ten times in the past fifty years (14). This suggests that the per-parasite probability of developing resistance de novo is on the order of 1 in 1020 parasite multiplications."

Do you see any mention of mutations in that, Lee?

[quote=“lee_merrill, post:228, topic:15043”]

@Mercer

@lee_merrill

One new binding site was produced, actually, which was beneficial.
[/quote]

Really? I count five, all of which are beneficial. How much would you bet that you are right and I am wrong?

I would need to see the details, the one instance I am aware of is the (disputed) Vpu gene.

Sorry, I don’t understand. Why are you quoting Behe as though you believe in his authority, but lack sufficient faith to bet money on his veracity?

And why are you acting like Evolution News is some sort of scientific publication?

@lee_merrill

From considerations of protein shape space, he deduces that arriving at a new, beneficial protein-protein binding site is also on the order of 1 in 10^20.

Behe has done no such thing. He claims to have done so, and you present his writing as Gospel truth, but are strangely unwilling to bet that he is correct.

@lee_merrill

Here is more of Behe’s argument (have you read it?),

Yes. It’s ridiculous. I have more experience publishing in the relevant fields than Behe does. How much do you have?

An example:

@lee_merrill

“13.Even though protein-binding sites often involve a score of amino acids on each of the partners, experiments have shown that only a fraction of those are important for having the two proteins stick to each other. (For example, see Braden, B. C., and Poljak, R. J. 1995. Structural features of the reactions between antibodies and protein antigens…"

This is utter nonsense because antibody binding sites are small, so they are not an example. Also, the larger the binding area, the more residues are involved.

Behe conveniently fails to mention that incredibly specific binding (nanomolar and even picomolar–do you even know what that means?) evolves in only two weeks, from a pool of 10^8 random sequences, when your body makes antibodies against antigens. That’s far beyond his alleged edge. How do you explain that?

@lee_merrill

@Mercer

Anyone who has ever purified any protein (even undergraduates, @Eddie) is well aware that the greatest problem is that of the protein aggregating NONspecifically (binding to itself).

And Behe is interested in new beneficial interactions.

Behe isn’t, or he’d be doing science. He’s interested in bamboozling people like you to make easy money.

[Mercer (John Mercer)]

@lee_merrill

@Mercer

If two are required to increase resistance, one can’t possibly increase resistance by itself.

No, two are required to provide basic resistance, and just one mutation does not provide this.

Then how much money are you willing to bet that no single mutation confers any resistance?

@lee_merrill

@Mercer

what exactly is a “trial of the malarial parasite”? That looks like word salad to me.

One parasite organism.

More word salad. Why does Behe omit most of the known sources of variation, like sexual reproduction and recombination?

[quote=“lee_merrill, post:228, topic:15043”]

@Mercer

@lee_merrill

He then computes the probability of a new beneficial protein-protein binding site developing, and finds it comparable to 1 in 10^20.
[/quote]

He claims that, but he doesn’t compute anything of the sort.

I must refer you to his argument in his book, where he examines protein shape space.

I am an expert in the field and familiar with the argument. Behe doesn’t compute anything of the sort; arguments are not computations. Is that not a clear statement?

[lee_merrill]

@John_Harshman

@lee_merrill

But they do estimate about 2000 steps, correct?

Yes. And?

So 2000 steps is very improbable, if the steps are independent, even if each step is 99% certain.

@John_Harshman

@lee_merrill

… if the orb is slowly changing, these steps would seem to be needed in sequence.

That’s because you aren’t thinking about multi-locus, quantitative characters. Consider height, and say it’s controlled by 10 genes, each with a short allele and a tall allele.

Yet there would still need to be a sequence of ever-increasing tall alleles, for the orb to get taller.

@John_Harshman

@lee_merrill

Because neural circuits are complex, on average each neuron connects to 10,000 other neurons, as I have heard.

Some are complex, some aren’t, and they can increase in complexity (more properly, acuity) through selection. I’d like to see documentation on that numcerical claim, which presumably applies to humans only.

Here Scientific American claims 1000 connections per neuron, on average. Here we see each neuron can be connected to up to 10,000 other neurons. And the worm C. elegans has about 16.6 chemical synapses on average per neuron, which is still complex.

@John_Harshman

@lee_merrill

I’ve been arguing all along here for design, in various ways. Behe’s edge of evolution is perhaps the clearest way to detect design.

That’s not an argument, though I can see how you might think so.

But this is not a refutation…

[T_aquaticus]

@lee_merrill

So 2000 steps is very improbable, if the steps are independent, even if each step is 99% certain.

I don’t see any calculations.

@lee_merrill

But this is not a refutation…

No argument, no refutation needed.

[lee_merrill]

@T_aquaticus

@lee_merrill

So 2000 steps is very improbable, if the steps are independent, even if each step is 99% certain.

I don’t see any calculations.

Here is my calculation:

@lee_merrill

Let’s assume each step in the evolution of the eye is 99% probable, now Nilsson and Pelger estimate about 2,000 steps to evolve an eye, without regard to the neural processing changes that would be needed, so let’s add another 4,000 steps for neural processing, which is more complex than simple changes to an orbital cavity.

This results in a 1 in 6.5 x 10^27 chance for the eye to evolve, even if every step is 99% probable! Now there have been about 10^40 total organisms in all life’s history, as I have heard, so for the eye to evolve twice (never mind 50 times), we would need more organisms than have been present on earth.

[lee_merrill]

@Mercer

@lee_merrill

Here is the relevant quote from White’s paper: " Resistance to chloroquine in P. falciparum has arisen spontaneously less than ten times in the past fifty years (14). This suggests that the per-parasite probability of developing resistance de novo is on the order of 1 in 10^20 parasite multiplications."

Do you see any mention of mutations in that, Lee?

Mutations are implied here, but I don’t know why we need to see the word itself in order to confirm Behe’s use of this calculation.

@Mercer

Sorry, I don’t understand. Why are you quoting Behe as though you believe in his authority, but lack sufficient faith to bet money on his veracity?

And why are you acting like Evolution News is some sort of scientific publication?

But this does not show me your five new binding sites.

@Mercer

@lee_merrill

“13.Even though protein-binding sites often involve a score of amino acids on each of the partners, experiments have shown that only a fraction of those are important for having the two proteins stick to each other. (For example, see Braden, B. C., and Poljak, R. J. 1995. Structural features of the reactions between antibodies and protein antigens…"

This is utter nonsense because antibody binding sites are small, so they are not an example.

But does this establish the point that only a fraction of the amino acids in each protein is relevant for binding?

@Mercer

Behe conveniently fails to mention that incredibly specific binding (nanomolar and even picomolar–do you even know what that means?) evolves in only two weeks, from a pool of 10^8 random sequences, when your body makes antibodies against antigens. That’s far beyond his alleged edge. How do you explain that?

But that’s not mutation-by-mistake and natural selection. Behe is interested in new protein-protein interactions in cells, etc., not in the immune system.

@Mercer

@lee_merrill

No, two are required to provide basic resistance, and just one mutation does not provide this.

Then how much money are you willing to bet that no single mutation confers any resistance?

Well, I’d like to see your evidence, since chloroquine resistance arises in about 1 in 10^20 trials, and atovaquone resistance (which requires one mutation) arises in about 1 in 10^12 trials.

[quote=“Mercer, post:241, topic:15043”]

@lee_merrill

@Mercer

what exactly is a “trial of the malarial parasite”? That looks like word salad to me.
[/quote]

One parasite organism.

More word salad. Why does Behe omit most of the known sources of variation, like sexual reproduction and recombination?

Those are not ignored, Behe looks at what evolution actually accomplished.

@Mercer

I am an expert in the field and familiar with the argument. Behe doesn’t compute anything of the sort; arguments are not computations.

And this is not a refutation of Behe…

[lee_merrill]

@Roy

Having an eye evolve twice needs twice the number of organisms, since they would be evolving in parallel, not in series. Under your assumptions that’d require 1.3e28 organisms, far less that 1e40, and not 4.225e55 organisms.

No, the probability of an event (probability p) occurring independently twice is p^2, whether the events occur in parallel or in series.

[lee_merrill]

[quote=“Rumraket, post:236, topic:15043”]

@lee_merrill

@Rumraket

it’s very strange for Behe to think a new protein binding site is beyond the edge of evolution while simultaneously admitting it has evolved in HIV.
[/quote]

No, two new protein-protein binding sites, he places outside the edge of evolution.

And for reasons already stated that is obviously false.

And I would need to see your reasoning, Behe calculates the probability of two new binding sites as about 1 in 10^40.

@Rumraket

@lee_merrill

But the immune system is not evolving new binding sites

Yes it is. That is literally what the immune system does.

But it does not involve mutation-by-mistake and natural selection, to produce a fresh, new antibody.

[John_Harshman]

@lee_merrill

So 2000 steps is very improbable, if the steps are independent, even if each step is 99% certain.

If.

@lee_merrill

Yet there would still need to be a sequence of ever-increasing tall alleles, for the orb to get taller.

No. There would just have to be a change of frequency of tall alleles (perhaps at several loci) in the population.

@lee_merrill

Here Scientific American claims 1000 connections per neuron, on average. Here we see each neuron can be connected to up to 10,000 other neurons. And the worm C. elegans has about 16.6 chemical synapses on average per neuron, which is still complex.

So, all verbiage aside, you have adjusted your claim by 3 orders of magnitude.

@lee_merrill

But this is not a refutation…

Nobody can refute your argument until you actually make an argument.

[RonSewell (RonS)]

@lee_merrill

I’d like to see your evidence, since chloroquine resistance arises in about 1 in 10^20 trials, and atovaquone resistance (which requires one mutation) arises in about 1 in 10^12 trials.

By trials, are you referring to mutation events, or to fixations? Please define, because that distinction matters as to the domain over which you can generalize that rate. This has already been raised. I do not know what you are talking about, and it may be that you do not either in that you seem to mangle mutation, emergence, and trial together.

[CrisprCAS9]

@lee_merrill

Let’s assume each step in the evolution of the eye is 99% probable

Let’s assume each is 99.9999% probable…

There, now there is a 99.4% chance of the eye evolving. Since every possible mutation happens in every single generation for all reasonably sized populations, the odds of each mutation happening is essentially 100%, so the odds are actually even better.

[Dan_Eastwood]

@lee_merrill

So 2000 steps is very improbable, if the steps are independent, even if each step is 99% certain.

Not so. It will take an expected 2022.2 trials (=2000/0.99) to complete 2000 steps sequentially (Negative Binomial expectation). Running parallel in a population we expect 1980 of the 2000 to occur on the very first trial. After that it depends on the speed of mixing. Optimally ~10 trials (1 + log base 2 of 2000 ), but biology won’t be quite so efficient.

[Mercer (John Mercer)]

@lee_merrill

@Mercer

Do you see any mention of mutations in that, Lee?

Mutations are implied here,

So is existing variation, recombination, fixation under intermittent selection, and sexual reproduction, all of which Behe ignores. He attributes the whole thing to new mutations and constant selection. It’s ludicrous.