[00:00:04] Speaker A: Id the future, a podcast about evolution and intelligent design.
Can evolution and design be syncretized? Welcome to id the future. Today, we're going to discuss that question with Dr. Brian Miller. Brian Miller has a phd in complex systems physics from Duke University. I'd call him a polymath because he's got major expertise in areas like protein evolution, the origin of life, all of which are relevant to our discussion that we're going to have today. So, Brian, thank you so much for being on the show with us today.
[00:00:38] Speaker B: Thank you for having me.
[00:00:40] Speaker A: So, to give some background, I'm sure that most folks would guess that if you want to try to synthesize design and evolution, it probably depends on what you mean by design and what you mean by evolution. And in some cases, perhaps, it is very easy to synchronize design and evolution. Well, in the last few years, a valiant attempt at trying to synthesize design and evolution has been attempted by a theologian at the University of Helsinki named Evie Rope Kajonin. It's a valiant attempt at synthesizing design and evolution. We would say it's ultimately unsuccessful, but it was a very good try and a very thoughtful try nonetheless. In 2021, Cajonin published a very thoughtful and serious book titled the Compatibility of Evolution and Design with Paul Gray McMillan and Springer Nature. And in the book, Kajonin basically argues that evolution and design can be harmonized. But he isn't just saying that evolution and design are compatible in terms of, say, change over time or common ancestry. He's contending at a very deep level that mainstream evolutionary biology is fully compatible with design arguments or design perceptions that are based upon biological phenomena. So, for example, when the wing of the Hummingbird shows evidence of design, even while it's product of natural selection, random mutation, other processes, he would say that you can explain this in terms of design and natural selection without the need for some kind of direct divine intervention or supervision per se. Well, some of the most valuable members of my team here at Discovery Institute, I'm thinking of Steve Dilly, Brian Miller, Emily Reeves, and I did a book club reading Kajonin's book in 2022. And then last year in 2023, we wrote a peer reviewed paper published in the journal Religions titled on the relationship between design and evolution, which is an analysis of Professor Kajonin's model. And we each contributed different parts of the paper. And because it covers so much ground, I'm doing a series of podcasts with each author to discuss their role in the paper. So today we have Dr. Brian Miller with us to discuss your contributions, Brian, to this paper. So, again, thank you for coming on the show with us.
[00:02:46] Speaker B: Thank you. I'm really looking forward to this conversation.
[00:02:49] Speaker A: Yeah, this is going to be very interesting. A little bit of a different take on why we're discussing the science compared to the usual. We've already had Steve Dilly on the show, who is a philosopher, to help us give sort of the philosophical framing of Kajonin's arguments. But I want to ask you, Brian, about your contributions to the paper on the science. So what was your portion of the response to rope Kajonin's argument?
[00:03:12] Speaker B: Well, one of the major questions about evolution and design deals with the issue of proteins, because proteins are the basic building block of life. It's what most of your cells are produced, are made from. And the question really goes down to the issue of both the rarity of proteins and the isolation. And we'll be talking about Doug Axe's work that really deals with the rarity. And what Kajonin argued was that the laws of nature were designed such that every protein we see in nature could have evolved through standard evolutionary processes. So what I did was I assessed his argument, explained his argument, and then ultimately critiqued his argument.
[00:03:56] Speaker A: Yeah. So his background, what has been that argument from design based upon Doug Ax's research on the origin of proteins?
[00:04:04] Speaker B: Well, yeah, and this really goes back to people like Bill Dempsky, because what design theorists have recognized is that if you want to show evidence of design, you have to accomplish two tasks. First, you have to show that some pattern you see in nature could not have emerged through natural processes or chance, because the probability is just simply too low. And two, you have to demonstrate that the pattern demonstrates purpose or meaning. Now, no one really disputes the fact that when you look at life, you see what at least appears to be purpose, things that seem to show that they have an end goal in mind. But the big question is, could they have come about through evolutionary processes? And that gets to the issue of how rare are these biological systems in what we might call possibility space? And the easiest place to answer that question is with proteins. And the reason for that is because proteins are basically chains of amino acids. So there's 20 amino acids, commonly used for most proteins. There's a few others in special cases, but we can just talk about the 20. And what happens is those amino acids are in sequences, much like the letters in a sentence. So, like in a sentence, if the letters are in the right order, you have a coherent message. If they're not, you have gibberish. In the same way, if amino acids are in the right order, then what will happen is that chain will fold into a very specific three dimensional shape, which will perform some useful function, like drive a reaction or produce the part of a molecular machine. And Doug Axe has done some of the pioneering research for the intelligent design movement on just how likely it is for proteins to emerge by chance. And he studied one part of what's called a betalactamase protein, which is an enzyme that catalyzes a very specific reaction. And what he showed was that if you had, let's say, a random sequence of amino acids, the chance of that random sequence folding into one part of this betalactamase enzyme would be like one chance in ten to the 77. That's sort of a lower estimate for the probability, and that's like a one with 77 zeros behind it, and that's just one part of the enzyme. So that means if you look at the entire enzyme, the chance of a random sequence of amino acids folding into a functional betalactamase enzyme is less likely than picking one marked atom in the entire visible universe. What that means that these proteins correspond to such rare sequences that they could not have emerged through an undirected process. That's the basic argument.
[00:06:50] Speaker A: So I'm sure that many of our listeners have heard about Doug Ax's work. But one of the fun things about writing this paper with you, Brian, was we got to get into some of the responses to access research and what some of the critics have said, and then we got to respond to the critics. And, of course, you really took the lead in helping us to do that. So what was the response to access research? And do the critics arguments hold up to scrutiny?
[00:07:12] Speaker B: Well, what you find is there was a lot of critics that would blog about criticisms of his paper, his famous journal, Molecular Biology paper in 2004, where he presented this research. And often the criticism simply reflected the fact that the critics didn't understand the biology, but others had more thoughtful criticisms. And if you look at those criticisms, what they would often do is they would say that, well, maybe this betalactamase looks really, really rare, but there may be lots of other proteins that are much easier to evolve. And what they would do is they would talk about research into what are called peptide or polypeptide chains. And those are simply chains of amino acids, which are a bit smaller. The chains are a bit shorter than most other proteins. And what you find is that if you have a chain of amino acids, it's very easy to find one, by chance that could stick to atp. It's like one chain in about a trillion. Or if you had a smaller chain, what would happen is about one chain in 100 million would embed into the cell membrane of a bacteria and prevent, let's say, an antibiotic from being taken up by the bacteria. So they would say that functions are actually not that hard to evolve, and they're both right and they're wrong, because they are correct that certain functions can evolve very easily, but they're mistaken if they think that's relevant to Doug Axe's research, because Doug Axe was studying a fairly modest size enzyme, which was fairly typical compared to other enzymes. And what he showed and argued is that, in general, enzymes or proteins that perform even modestly complex functions are exceptionally rare. While the critics were looking at these chains that performed incredibly simple functions, just sticking to something. So that would be like if I argued that I could train a dog to play Mozart on a piano at the level of a concert pianist in a year. And if I made that case, you would probably say that I'm very skeptical. And then I'd say, well, I can prove it, because I can train a dog to fetch a stick in a week. Now, obviously, that argument doesn't work, because fetching a stick is a much, much easier task than playing Mozart on a piano. But that's the exact same mistake that you see with people that criticize Doug Axe's research.
[00:09:33] Speaker A: So it's kind of like Brian. They're saying, look, if we can solve this really simple problem, then maybe we can solve this really hard problem. But you're saying that that's not necessarily true. There's sort of a logical gap there, a logical fallacy that the ability to solve a simple problem does not necessarily entail the ability to solve a very hard problem. And if all you ever talk about is a simple problem, okay, well, fine. Where's the solution to the hard problem?
[00:09:57] Speaker B: Yeah, that's exactly analogy. My analogy would be like this. It'd be like someone says, look, yes, catching a stick is different than playing Mozart on a piano. But if you just keep training a dog with these very simple tasks, eventually those simple tasks will build up into playing Mozart on a piano. And that's obviously not the case, because the two tasks of playing a piano and catching a stick are fundamentally and qualitatively different. And the same is true with an amino acid chain that performs a simple task like sticking to an ATP versus these very complex tasks performed by these proteins.
[00:10:36] Speaker A: Okay, so there's a lot of different proteins out there. Is the research that Doug Axe did, was that just an anomaly where he happened to find a single domain that was very, very rare? Or. His more recent literature supported Axe's conclusions that proteins are perhaps generally, in most cases, in many cases, too rare to evolve by undirected processes.
[00:10:56] Speaker B: Yeah, the more recent literature has really supported Axe's argument. In fact, even before published his paper, there were other studies that also showed the rarity of proteins.
Famous examples would be yaki or sour. Now, what happened since Axe's research is you have other very prominent researchers that have published very good research reaffirming that conclusion. So, for instance, Martin Novak is one of the leading researchers in mathematical biology. So he looks at mathematical models of evolution, and he has a very sophisticated paper on the timescales for evolutionary novelties to emerge. And what he did is he talked about what's called sequence space. And for those of you not familiar with sequence space, that simply represents sort of a graph or a matrix that correspond to all possible amino acid sequences. Or if you're talking about the genome, it would be nucleotide sequences in DNA. What he showed is that if you have a target in sequence space, then the time it takes to find that target is so great that you would never expect novelties to emerge by evolution unless you started very, very close to the target. So, for instance, in more biological terms, if you had an enzyme, you could evolve another enzyme, which was really, really similar if it only required a few amino acid changes. But you would never expect to evolve something that was fundamentally different through a random, undirected search. And another paper, which is a very nice paper by Tian and Best, was published in 2017. And what they did is they looked at ten smaller proteins, single domain proteins, and they actually estimated, through a technique very different from axis, it was more of a mathematical analysis comparing various sequences and protein families. And they found that of these ten domains, all of them were quite rare. And the ones that weren't very, very small were exceedingly rare. In fact, if you were to scale the length of the proteins, and you would scale their results to the size of Doug axis protein, what you would find is they found that the proteins were even more rare, more extremely uncommon when scaled by length, than actually Doug axis research. So all of this research has really affirmed that proteins often are exceptionally rare. So that is just not controversial among experts in the field.
[00:13:28] Speaker A: Okay, very interesting. So, obviously, we need to take it back to rope Kajonin's arguments here. He's arguing that you can see this rarity, you can see this design and yet have evolution producing it, and still have it be evidence for design. So we can leave aside sort of the philosophical nature of his argument for a moment. But on the science, what was Cajonin's argument related to proteins, and how did he support it?
[00:13:55] Speaker B: Well, to his credit, he affirmed the fact that many proteins are exceptionally rare. In fact, he recognized that they were so rare that you would never expect one to evolve by chance. But he followed, in the logic of another famous researcher named Andreas Wagner. And what he argued was that if you look at the distribution of proteins in sequence space, what you find is that the laws of physics were special, in that these functional proteins were so close to each other in sequence space that you could have one evolve into the other without any great difficulty. So analogy would be, if you imagine you were on a planet, and that planet only had, let's say, 1% of it could be traversed with solid land, the rest was like lava pits or water. But you found that all the land was perfectly aligned to create these narrow corridors where you could go from one side of the planet to the other through these very, very narrow paths. So cajonin is arguing, as Wagner, that the sequence space is very, very special, and that God actually designed the laws of physics in such a way that one protein could easily evolve into another. So that you see both design in the rarity of proteins, but also evolution in that the proteins could evolve because of these very special, what he referred to as fitness landscapes.
[00:15:22] Speaker A: So he thinks that these fitness landscapes that basically allow these proteins to evolve, have been fine tuned essentially by God or the designer, and that the fine tuning of these fitness landscapes, which allows you to get from one very rare sequence to another, that is sort of the telltale evidence that there is design hidden in the fabric of the laws of nature to make a very special sort of fitness landscape for these proteins to evolve. Is that basically right? And what's your response to that argument?
[00:15:53] Speaker B: Yes, that's exactly right. And my response to it is, if you look at the research that was produced by Andreas Wagner, or any other research that attempts to make this case, it's always highly, highly theoretical. What they do is they look at how different proteins are positioned in sequence base, and then they assume that you can transition from one protein to another in a continuous path. But it's always hypothetical, it's always theoretical. It's not based on hard data. In contrast, when you look at the actual experiments, you look at the hard data, what you see is that not only are proteins extremely rare, but they appear to be highly isolated in sequence space. So you do not see these continuous paths going from one protein to another. They're more like isolated islands.
[00:16:44] Speaker A: So what is your evidence that proteins are not just rare, but also isolated?
[00:16:49] Speaker B: Well, this is what's really exciting, because a lot of this research has really come out since Doug Axe has published his paper. And we'll talk a bit later about Dantofik, who has done some of the pioneering work. But what you really see in the hard data is that people have looked at proteins like betalactamase, like Hisa, like green fluorescent protein, and what they've done is they've applied series of mutations to these proteins, and they've allowed mutations to accumulate. So they just allow these mutations to add up. And then what they would do occasionally is look to see if the protein still worked. What you can then do is you can actually plot a very nice graph of the probability that a protein will work with a given number of mutations. And this is exactly what Dantoffic did with his work on betalactamase. And what you find with multiple proteins is, after you alter about 5% of the amino acids, there is almost no chance that the protein will work. So, in other words, any combination of five mutations that alter amino acids will make a non functional protein. And that strongly suggests that proteins are. That where the functional sequences are, are highly clustered around what you would call wild type sequences, or proteins, that you'd actually find in the wild. Now, let's compare this to what are called superfamilies. Now, a superfamily is one of the highest categories for proteins, and what they correspond to is proteins with the same basic structure. There may be a lot of variation in both what they do and their exact details, but they still show certain similarities in structure. And what you find is, if you look at two different proteins that are in different superfamilies, they will almost always have more than 80% sequence differences. So let's talk about those comparisons. So we see that proteins can change by about 5% before. It's very unlikely they'll function. But these superfamilies are isolated, where even the most similar proteins in different superfamilies are about 80% different. That's really striking. Now, another piece of evidence comes from what's called percolation theory. And this is a theory that's gone back for several decades. And what people have done is they've looked at latices or grids, and they ask the question that if you fill a certain percentage of these cells in these grids? What's the likelihood that you'll have continuous paths going from, let's say, one side of the grid to the other? So, in other words, if you imagine a grid where you have a certain percentage of black squares and the rest of the squares are white, how often do you see lines of neighboring squares forming paths that extend throughout the grid? That's the basic idea. In the context of sequence space, that would be like a series of functional sequences. That could lead from one protein to the other. So you have one protein, you have a mutation. That changes it by one amino acid, and it's still functional. You change it by another amino acid, it's still functional. So you have this continuous path of alterations. Where you go from one protein to another that might be quite different through a continuous series of functional intermediates that are only one mutation away? That's the idea. So percolation theory directly applies to sequence space. Well, what people have found is that if the probability of functional or filled squares is less than one divided the number of nearest neighbors to a given cell, then you will not see functional paths. Let me explain that in terms of sequence space and proteins. If you have a protein of some length l, then there are a certain number of sequences that could emerge with a single mutation. So if you look at a protein for any given amino acid, on average, you can change it to 7.5 other amino acids from a single mutation. That's an average.
In some amino acids, you could change to six other amino acids from mutations. Sometimes it's twelve, but on average, it's 7.5. So if you have a protein of length l and you multiplied l, let's say l is ten by 7.5, that's 75. That means that there's 75 other possible sequences that could occur from a single mutation. Now, according to percolation theory, if, in that particular case, the probability of a sequence being functional was less than one over 75, then you would not expect to see these continuous paths through sequence space. You couldn't evolve anywhere because there isn't enough of these sequences that are functional. Well, if you look at a typical protein you're talking about, the probability of a sequence being functional. Can't drop below somewhere between, like, one in 1001, in 10,000. But again, remember that if you look at Doug Axe's research, the probability of a sequence being functional is one over a one with 770. So this is such a massive difference in probabilities. That that strongly suggests that you're not going to find continuous functional sequences in sequence spaces. So let me use an analogy. If you imagined that you had a planet the size of our visible universe, and you had traversible land on that planet, a land that you could travel on, was the size with the area of, let's say, an atom, then the probability that you could drive from one side of the planet to the other with that little land is like the probability of there being an evolutionary path from one protein to another, given that research. So this is very compelling, because what it shows you is that protein rarity almost always will mean protein isolation, unless there's just incredible amounts of biasing of where functional sequences are in the sequence space.
[00:22:54] Speaker A: Well, thank you, Dr. Miller, for that comprehensive explanation, to understand why it's not just rarity, it's also the isolation of proteins that matter, and how we can actually demonstrate empirically, theoretically, all of those, that the rarity of proteins entails their isolation. And I think you've done some very interesting work here in developing these arguments, and I suspect we're going to be hearing more about this as time goes on. But you also reference the research of Dan Tophik. Who is he, and how has his research supported the evidence for rarity and isolation of protein sequences?
[00:23:27] Speaker B: Well, Dantoffic was always considered a leader in the field, and unfortunately, he had a very tragic accident a few years ago. So he's passed away, and the entire field has really mourned his loss, because he was such a pioneer and a wonderful researcher.
But just to really appreciate what he contributed is that he showed that proteins demonstrate negative epistasis. And what that means is that as mutations accumulate in proteins, the likelihood of the next mutation being tolerated drops, and it can drop very quickly. So that means that proteins, as they accumulate mutations, the probability that they remain functional, drops what's called hyper exponentially, which is very, very fast. And that's why you're not able to change proteins very much before they stop functioning. Now, a second piece of research that was incredibly important is he did a lot of pioneering work on altering proteins to allow them to perform some new function at a much higher level. And what he showed, though, is very significant. He showed that he could alter proteins to do something interesting, but only if you did not alter the structure. So the basic three dimensional structure had to remain intact. And all you did is you tweaked it a little bit, particularly in the active site, to enhance a certain function at a much greater level. So, again, what this shows you is it's the opposite of what rope argued, is that in fact, the biosing of functional sequences in sequence space does not favor the evolution of new proteins, but it actually prevents the evolution of new proteins if they have completely different structures. And finally, Dantoffic has lectured about how different protein superfamilies exist in what he described as separate galaxies. They really have almost no connection either in sequence or structure. So again, his arguments and his research really reinforced this argument that proteins are both rare and isolated.
[00:25:31] Speaker A: It's always good to get support from other researchers in the field. I mean, and Dan Tophick was an evolutionist. He was not a proponent of intelligent design. As you mentioned, Brian, it was very tragic when he passed away a few years ago. But, I mean, he certainly agreed with, despite the fact that he was an evolutionist, he agreed with this fundamental claim that protein sequences in many cases, are very rare.
[00:25:52] Speaker B: Yeah. And he was a very honest researcher, so he assumed on faith and he would make statements about this, that this is essentially what he assumed had to be true, that proteins could evolve one into another or through some other natural process. But he acknowledged that there was no explanation to date for how these superfamilies came into being. He described it as.
He said that the origin of a new protein family looked like almost a miracle. That's a quote from him. So he was very honest about what he saw, even though his philosophy didn't like what he saw.
[00:26:27] Speaker A: Okay, well, thank you very much, Dr. Miller, for helping us to understand this question of whether protein sequences are both rare in sequence based and also isolated when it comes to being able to traverse from one sequence to another by a blind process like natural selection, random mutation. So, given all this, how would you summarize the evidence for rarity and isolation as it relates to the claims of ropekajonin in his book and this question of design and evolution?
[00:26:55] Speaker B: Yeah, certainly. So there are several independent pieces of evidence that reinforce the fact that proteins are both too rare and too isolated to have evolved from an undirected process. At least the first representatives of superfamilies unfolds. Now, the first is that many, many proteins are exceptionally rare. So any protein that performs even a modestly complex function will almost certainly be too rare to have evolved through an undirected search. The second piece of evidence is that proteins in different superfamilies are highly, highly separated in sequence space. Again, Dan Toffic talked about them being as part of different galaxies. The third piece of evidence is that the empirical data shows that proteins can tolerate only a very limited number of mutations, and the mutations they can tolerate do not change the structure. So the problem is, to evolve a new structure that's fundamentally different, you have to have mutations that will change the structure dramatically. And the empirical evidence shows you that that's not plausible. And finally, percolation theory puts hard numbers to these more qualitative arguments. And the hard numbers are that when the probability of a protein being functional drops below a certain threshold, you will not expect there to be continuous paths that would evolve one protein in another. And that value in the disparity between the actual rarity of proteins and what the percentage of functional sequences would have to be is so great that there is no realistic chance that you're going to see evolutionary accounts for the origin of radically new proteins.
[00:28:41] Speaker A: So it sounds like this claim that Robekea Jonan has made, which, again, he's trying to take a very thoughtful approach. So we give him major credit for trying to take this question so seriously of design and evolution. But this fundamental idea that there are these narrow, very tenuous pathways through sequence space to connect one protein sequence to another, that could be a way that the design could have been implemented in the world. But we just don't see evidence in our universe that that's the way things work. We don't see evidence for these pathways connecting these very rare and actually isolated protein sequences. Is that really what it comes down to here?
[00:29:21] Speaker B: Yeah, you're exactly right. Sirope has presented a very thoughtful argument of how, in principle, a universe could have been created to allow proteins to evolve through natural processes. But that's not our universe. Our universe seems to be created in such a way that proteins are more like platonic forms. They're entities that can change by a small amount, but that amount is very, very small compared the distance between fundamentally different proteins.
[00:29:52] Speaker A: Okay, well, thank you very much, Dr. Brian Miller, for sharing with us today about this fascinating question of whether or not protein sequences can evolve and how it relates to this paper that you co authored, responding to this book by theologian rope Kajonin on the question of whether or not design and evolution are compatible. So thank you so much for speaking with us.
[00:30:14] Speaker B: Thank you. It's been a pleasure.
[00:30:16] Speaker A: I'm Casey Luskin with ide the future. Staying tuned for more on whether or not design and evolution are compatible. Thanks for listening.
Visit
[email protected] and intelligentdesign.org. This program is copyright Discovery Institute and recorded by its center for Science and Culture.
