Is There an Edge to Evolution? Part 6: The Cathedral of Life

Bookmark and Share

December 4, 2010 Tags: Design

Today's entry was written by David Ussery and Darrel Falk. You can read more about what we believe here.

Is There an Edge to Evolution? Part 6: The Cathedral of Life

An Evaluation of Behe’s Edge of Evolution, Chapter 9 – The Cathedral and The Spandrels

This series of posts has been going through Michael Behe’s book, The Edge of Evolution, chapter by chapter. This penultimate chapter focuses on the findings of one of the most fascinating new topics in biology today, evolutionary developmental biology (evo-devo). In essence this is a field that couples two sub-disciplines, evolutionary biology and developmental biology using the tools of molecular biology. Chapter 9 is moving on to "higher levels of biological organization", and Behe readily admits that things are now a bit less well-defined, and "the arguments in this chapter will necessarily be more tentative and speculative than for previous chapters" because now the subject will be dealing with more complicated things - plants and animals, and "much less is known about what it takes to build an animal than to build a protein machine" (pages 172-173).

As often happens in science when one examines a phenomenon through a different window, many new and often surprising insights come into view. In 1940, for example, few people studying genetics imagined that DNA would be the genetic material; most everyone thought it would be proteins. However, soon afterwards the tools of microbiology began to reshape how biologists viewed the genetic material, and that in turn opened the window for Watson and Crick to see the gene’s true molecular nature. With that, the now-famous double helix came into view for the first time.

Examining the surprises that appear when one looks at a phenomenon from a new vantage point is what makes science so engaging. Scientists love surprises. In this chapter, Behe focuses on one of the most exciting scientific discoveries of the past thirty years, and implies that because evolutionary biologists were surprised, that evolutionary theory had reached the edge of its scientific limits. Let’s examine the basis of the surprise and then explore whether Behe is justified in concluding that the scientific surprises discussed in Chapter Nine correspond to a cliff-edge. Is Behe correct in concluding that going beyond that edge, one enters into territory that can only be explored by inserting a [supernatural] Intelligent Designer into the scientific “equations?” Is Behe’s edge simply a window of opportunity to see where mainstream biological tools will take us, or is it a blank wall? Behe believes it is a blank wall. Why?

In an earlier post our colleague, David Kerk, described the tinman gene, the gene required for making a heart. It is one of the many conserved “master” genes whose functions are now understood, through the new perspectives afforded by evo-devo. These genes serve as genetic switches that have the capability of activating particular developmental programs. A given switch (i.e. a master gene) is often structured quite similarly throughout the animal world even when comparing widely disparate species like flies and frogs. This high degree of conservation shocked evolutionary biologists. It was startling, for example, to realize that the same gene that served as a switch to turn on eye development in flies was found in humans, because if you think about it, the eyes of flies are a lot different than human eyes! Indeed, the mouse master gene for making eyes has been transplanted into fruit flies where it still works. Fly cells respond to the mouse switch by making eyes—fly eyes, not mouse-like eyes—but eye tissue nonetheless. Biologists didn’t expect genes to be conserved through the greater than 550 million years since mice and flies had a common ancestor. However, even though it was a surprise, it is extremely consistent with evolutionary theory. Despite the surprise, the finding is completely consistent with natural selection and common descent. Master genes are conserved through the parade of life. Like the hour hand on a ticking clock, they change, but only at a crawl.

Actually, the surprise comes from just how beautifully consistent the view is from this vantage point. Scientists were expecting consistency, but certainly not in such an eye-popping, mind-boggling manner.

Behe chooses to view things differently. This is evidence, he says on page 190, that:

... the best minds in science have been misled. They justifiably expected randomness and simplicity…

These scientists were NOT expecting randomness and they were most certainly NOT expecting simplicity. What they were expecting was greater complexity—not the degree of simplicity they found. The same genes are being used to build insects as what are used to build mammals. What could be simpler than that? So from this perspective, it is difficult to even begin to grasp Behe’s point about expected simplicity.

Let’s go back though to his statement regarding the notion that the scientists’ “expected randomness.” Why would he tell a general audience that? Natural selection is the very converse of a random process with an unanticipated outcome. They knew it would be non-random—natural selection is by definition non-random. What surprised them—what shocked them actually—was just how foundationally simple and non-random evolutionary mechanisms turn out to be. Evo-devo is not inconsistent with the core of evolutionary theory. Quite the opposite actually—natural selection is by definition a non-random process.

It is important to be fair to Behe here. He has stated clearly that the data as a whole are consistent with common descent. This is not in question for him. Indeed, it would probably have been good for him to emphasize in this chapter that these data are beautifully consistent with his own premise—common descent. One can track the lineage of the “genetic toolkits.” The toolkits get modified slightly and one can trace their modifications as one examines the tree of life. But there is a tree—one tree—Mike agrees with this! Indeed his entire approach to intelligent design is grounded in common descent. So in that regard Behe is in total alignment with mainstream biology. In that regard BioLogos and Behe are truly at one. We wish he would say that more often. There is a sense in which Mike Behe is more closely aligned with BioLogos than with many of his colleagues at the Discovery Institute including Bill Dembski and Stephen Meyer, who, although they waffle on occasion, have come out against common descent. Neither Bill nor Steve are biologists. It would be great if they would listen to their own biochemist. If they would, then perhaps Mike Behe’s statement on page 191 would take us to a whole new day:

Let’s acknowledge that genetics has yielded yet more terrific (and totally unanticipated) evidence for common descent.

Do you hear that, members of the ID Movement? Perhaps the single most important figure in the ID movement over the past fifteen years has called for an acknowledgement that common descent has occurred. Implied in this statement is evidence for common descent all the way from single cells to human beings. If the leaders and followers who do not have credentials in biology and biochemistry would get on board with their expert who does, then half of the concerns with the ID movement would be over.

Behe goes on from there to demonstrate the complexity of the genetic circuitry needed to build various cell types. Vertebrates, for example have B lymphocytes to help fight off infections; invertebrates, he says, do not. The genetic circuitry to build any cell type is exceedingly complex. Organisms are placed into classification groupings, based on somewhat subjective human ideas. Vertebrates are member of the phylum, Chordata. Invertebrates are members of other phyla. Behe proposes that the differences between phyla are so large, that they require the invention of whole new cell types. Since new cell types require new protein interactions and since he believes he has already shown that new substantive protein interactions won’t occur without intervention, new phyla as he sees it cannot arise without intelligence.

Let’s be clear, there is an Intelligence behind all of life. So, even here we don’t disagree. The question is why Behe wants to draw a line (an edge) between presence of God and absence of God in life’s history—presence of intelligence and absence of intelligence. Perhaps it is because of the necessary “absence of intelligence” to serve as an experimental control for “presence of intelligence?” If so, this sounds as though his theology is flying free. It is not grounded in Scripture. The Bible asserts that “by him all things were created…He is before all things and in him all things hold together.” (Colossians 1:16, 17). It also says, “Through him all things were made, without him nothing has been made that has been made” (John 1:3).

Further, one could build a case that he has now floated free of his scientific roots as well. Based on the data available so far, Behe may be correct that we cannot successfully trace the step-by-step lineage of new particular cell types in certain phyla. Behe’s assertion that for scientific reasons, however, we must now insert an Outside Architect is deeply flawed. The only scientific evidence he lays out to support the scientific hypothesis of the need for this architect harbors back to the same sort of calculations on the probability of new protein/protein interactions. We have already demonstrated that those calculations are off by many orders of magnitude.

What are those calculations that show no new protein/protein interactions have occurred? What is the data he analyzes? On page 200 Behe suggests that out of a billion rats subjected to warfarin in the past 50 years, we might have expected “many new regulatory regions; none seemed to have helped against warfarin.” Did anyone check these billion rats to see if some had undergone changes in regulatory regions? It seems that this is really a premature conclusion to put forward to the public without vetting it before the scientific community first. From there he goes on to fruit flies that have been studied in the lab for 100 years. During this time “no new, helpful, developmental-control programs have appeared.” Is there some reason why we might have expected some new “helpful” program in flies? What sort of “new help” would Behe have envisaged for fruit fly development? How would it have been detected? Was anyone actually looking for such a thing?

In the chapter, Behe then goes on to report that the malaria parasite has evolved no new reported “cell forms or regulatory systems” in a hundred billion billion chances. How does he know this? It is true that no one reported new regulatory systems. But was anyone looking for them? For all we know the parasite might have been evolving and even changing elements of its regulatory system. A careful analysis might even have been able to show this.

Based on analyses like these, Behe ends his chapter by discussing spandrels, the space between the arches that hold up a great cathedral. The arches, he says are clearly designed by a great architect. The artwork that decorates the spandrels were added after the fact—after the architect had left the scene. Now moving towards a metaphor, he states that science, his science, has now shown that the major classification groups of animals are like the arches of a great cathedral—they have been designed by God, the Greatest Architect. Darwinian evolution comes in and decorates the spandrels with all sorts of species and maybe genera and families--but the existence of phyla requires an Architect. This is Professor Behe’s cathedral and although one has to give him credit for being creative, this is based on his claim that rats that don’t evolve new systems (for which no one was carefully looking, to be honest). It is based upon fruit flies that don’t seem to be developing new and better body plans than they already have, and it is based on billions of billions of malaria parasites that are not being analyzed for changes at the molecular level. Surely ID is now floating free of scientific data. A theology based on a God whose Presence in creation comes and goes is equally problematic. Is not ID also floating free of Scripture?

It doesn’t have to be this way. Professor Behe, since he accepts common descent, is already half way home towards accommodating the scientific community. As imperfect human beings, we are all wrong on occasion. As mentioned early on in the chapter, "the arguments are more tentative and speculative" here. But there's also a danger that perhaps the arguments have strayed far from solid science as well as sound theology. It doesn’t have to be this way.


David Ussery is an associate professor of comparative microbial genomics at the Center for Biological Sequence Analysis at the Technical University of Denmark and on the faculty at the University in Oslo, Norway. Ussery is the co-author of Computing for Comparative Microbial Genomics and has authored or co-authored 130 articles for science and professional journals. He is also a frequent public speaker on the topic of bacterial genomics.
Darrel Falk is former president of The BioLogos Foundation. He transitioned into Christian higher education 25 years ago and has given numerous talks about the relationship between science and faith at many universities and seminaries. He is the author of Coming to Peace with Science.

< Previous post in series


View the archived discussion of this post

This article is now closed for new comments. The archived comments are shown below.

Loading...
Page 2 of 5   « 1 2 3 4 5 »
Alan Fox - #43112

December 9th 2010

In my nine years of discussing ID on the internet, I’ve never read about tail or leg genes.

That surprises me, Bilbo. One of my several bannings from Bill Dembski’s former blog, Uncommon Descent, was for attempting to point out to Dave Springer (his “blog czar”) that there is no blueprint in the genome, but rather organisms grow via cell division and differentiation under the control of genes acting as growth factors. I also seem to recall the evolution of bat wings has cropped up here at Biologos..


Alan Fox - #43113

December 9th 2010

It’s a 27% identity over the entire 525 amino acids of the Volvox protein with an E value of 1e-25.

As Dr. Ussery points out “How can proteins NOT be ‘borrowed from a previous system’ - unless they are created de novo by God or something”.


Alan Fox - #43120

December 9th 2010

@ Dr. Ussery

From Dr. Ussery’ previous thread:

chunkdz - #42198
December 1st 2010

Dave Ussery wrote: “I wouldn’t be surprised if many (most?) of the eukaryotic proteins have homologs in bacteria - this just makes sense…”

Just curious - is this a prediction that flows from Modern Evolutionary Theory, or is it a novel prediction based upon your own research?

chunkdz has mentioned a few times he is waiting patiently for an answer to his question so this is a heads-up. Hope you find time to respond and thanks for your contributions to Biologos.


John - #43123

December 9th 2010

Ashe:
“I think what happened was that the architects of the Modern Synthesis expected selection to have a great amount of freedom in terms of creating new forms/functions. But what we have found is that selection is quite constrained by what already exists.”

This is just insane. The point that you’re missing is that the variety of forms and functions we see does not correlate with the variety of proteins. So the functions of those proteins through evolution are surprisingly UNconstrained.


Bilbo:
“How do you know that?”

Because no one’s found anything like that.

“He didn’t say that scenario was the norm.”

Doesn’t Behe say that biology has many examples of his edge?

“He said that if such a scenario occurred, it was not the result of random mutations.”

That’s just silly. Evolutionary theory wouldn’t predict that mutations alone would do it *IF* it occurred either. It obviously would require a whopping dose of selection.

“In my nine years of discussing ID on the internet, I’ve never read about tail or leg genes.”

——
Bilbo Says: 
March 10th, 2009 at 10:39 pm
I might be mistaken, but I think the “toolkit” for development of the eye appears in most of the phyla.
——


Ashe - #43161

December 9th 2010

John:

This is just insane. The point that you’re missing is that the variety of forms and functions we see does not correlate with the variety of proteins. So the functions of those proteins through evolution are surprisingly UNconstrained.

Not insane at all. It’s a point that was made by actual evolutionary biologists, Gerhart and Kirschner. Why would you expect the variety of proteins to correlate with forms and functions? You can just recombine them in different ways/clusters.


Ashe - #43166

December 9th 2010

Alan:

As Dr. Ussery points out “How can proteins NOT be ‘borrowed from a previous system’ - unless they are created de novo by God or something”.

You yourself pointed out how you can get functional, folded proteins from random sequences.


Alan Fox - #43185

December 9th 2010

You yourself pointed out how you can get functional, folded proteins from random sequences.

Well, I drew attention to Szostak’s work. In doing so, my point was that the argument that the chance of finding a functional protein is 1 in (n^20) depends on the unwarranted assumption that all other possible sequences have no function. Szostak demonstrates this assumption is false.


Mike Gene - #43186

December 9th 2010

John,

Thanks for the suggestion.  No, beta catenin does not top the list, as there are 10 other hits with slightly smaller E values.  They all seem to be either hypothetical proteins or uncharacterized proteins.  Are you somehow suggesting this means human beta catenin and the Volvox protein are not homologous?

Interestingly enough, if I had started with beta catenin from Macaca fascicularis, then yes, I would retrieve the Volvox protein at the top of the list and the Macaca beta catenin would top the list if its sequence was probed with the Volvox protein.

The same holds true with the Volvox protein and Aardvark from slime molds.  Use the Aardvark sequence to probe Volvox sequence and the protein-in-question tops the list.  Use the Volvox protein to probe the Dictylostium genome and Aardvark tops the list.


Mike Gene - #43187

December 9th 2010

So anyway, here is what I did.  I took the 10 hypothetical human proteins mentioned above, human and Macaca beta catenin, Aardvark, and the Volvox protein and aligned them all with ClustalW.  Then, I used PHYLIP to make a tree (keeping in mind that I am a novice).  I made a tree using the default settings and a tree excluding positions with gaps and correcting for multiple substitutions.  Both trees clearly show the Volvox protein nesting with the beta catenins from human, Macaca, and Aardvark to the exclusion of the 10 hypothetical human proteins.

It’s looking better than I thought.


Bilbo - #43194

December 9th 2010

@John - #43123

Because no one’s found anything like that.

You mean no one has found protein complexes where the individual proteins do not interact with proteins from other complexes?

That’s just silly. Evolutionary theory wouldn’t predict that mutations alone would do it *IF* it occurred either. It obviously would require a whopping dose of selection.

My scenario would require a whopping does of luck before selection kicks in.  The selection wouldn’t need to be “whopping” at all, though. 

Bilbo:  “In my nine years of discussing ID on the internet, I’ve never read about tail or leg genes.”

——
Bilbo Says:
March 10th, 2009 at 10:39 pm
I might be mistaken, but I think the “toolkit” for development of the eye appears in most of the phyla.

I fail to see the relevance of the quotation, though I am flattered that you follow everything I say so closely.  I’m curious.  Where did I write that?


John - #43202

December 9th 2010

Mike wrote:
“Thanks for the suggestion.”

You’re welcome.

“No, beta catenin does not top the list,…Are you somehow suggesting this means human beta catenin and the Volvox protein are not homologous?”

I suspect that they are simply members of the ARM family. If you’re going to claim special status for beta-catenin, you’d have to check the homology after deleting the ARM repeats (BLAST IDs them for you) and reducing the gap penalty.

“Interestingly enough, if I had started with beta catenin from Macaca fascicularis, then yes, I would retrieve the Volvox protein at the top of the list and the Macaca beta catenin would top the list if its sequence was probed with the Volvox protein.”

Sorry, that’s not very interesting.


Ashe - #43212

December 10th 2010

Alan:

Well, I drew attention to Szostak’s work.

You drew attention to the fact that proteins need NOT be ‘borrowed from a previous system’ , or even from the homolog.


Ashe - #43213

December 10th 2010

Since even a protozoa, choanoflagelata has a cadherin gene, I can believe
that Volvox has beta catenin.


Alan Fox - #43229

December 10th 2010

Ashe claims:

You drew attention to the fact that proteins need NOT be ‘borrowed from a previous system

I am not sure how you gleaned that from my comments but you are incorrect in that assumption. Proteins are coded in DNA so to be part of any organism’s biochemistry, the gene for that protein must exist in the genotype of that organism. There are routes other than mutation (HGT, gene duplication and subsequent drift, viral infection for example) but otherwise proteins, like whole organisms will evolve through viable (or for proteins not immediately lethal) intermediates. It is perfectly plausible that metazoan proteins have deep homology with very ancient organisms such as prokaryotes. Given the biochemical similarities across the domain of life it is inevitable.

Szostak demonstrates that there may be many unexploited functional proteins and intermediates (now untraceable) that must have existed if evolution is correct as a theory may be vastly more common than a i in a gadzillion. Do you think the Israelites wandered for fourty years in the desert before getting to Canaan? I want a record of every footstep before I accept the hypothesis!


Ashe - #43251

December 10th 2010

Alan, not really sure what the relevance of your comment is. If you can get a functional, folded protein from a pool of random sequences, as many papers demonstrate, then there is no actual need for deep homology. 

It is perfectly plausible that metazoan proteins have deep homology with very ancient organisms such as prokaryotes.Given the biochemical similarities across the domain of life it is inevitable.

The Nematostella genome shows that about 20% new genes have been added to the eukaryotic repetoir since the last eukaryotic common ancestor.

What contemporary data now shows is that prokaryotes gave us DNA, RNA, protein synthesis, the bi-layer membrane, etc and that basic machinery hasn’t changed much in the last few billion years. Then there was the completely new invention of compartmentalization, with proteolysis, signaling systems,etc.  And then that remained unchanged. That continues all the way up to the Cambrian. So what you have is patterns of completely new inventions followed by a long period of stasis.


Mike Gene - #43292

December 10th 2010

Hi John,

Thanks for the additional suggestion.  I had actually thought of that before, but didn’t follow through because it’s mostly the Arm repeats that are responsible for beta catenins sequence conservation.  Nevertheless, since you raise it, I took the N-terminal 130 amino acids and the C-terminal 115 amino acids from human beta catenin, as the 12 Arm repeats are sandwiched between.  I used them to BLAST the Volvox protein and there is nothing there.  But as a control, I also used them to BLAST Aardvark and there is nothing there either. Finally, I used them to BLAST a sponge beta-catenin.  The C-terminus has nothing, but there’s a 50 amino acid span with low 30% identity in the N-terminus that gets picked up.  Bottom line here is that given Volvox and humans last shared a common ancestor a billion plus years ago, and the sequence outside the repeats is not strongly conserved, I’m not sure a negative result here means all that much.


Mike Gene - #43293

December 10th 2010

However, I do think your concern is indeed legitimate, as you may be right.  But I would maintain the Volvox protein can be scored as putative homolog of beta-catenin (more than “just a protein with Arm repeats”) and this scoring is supported by some circumstantial evidence (after all, I would further argue that beta-catenin and alpha-importins are also homologs, yet there the evidence is even more qualitative).  However, right now, this is mostly a testable hypothesis.  What is needed are functional assays with this protein to seal the deal.


John - #43328

December 10th 2010

Bilbo:
“You mean no one has found protein complexes where the individual proteins do not interact with proteins from other complexes?”

No. I mean that no one has found protein complexes in which the ancestors of all three don’t interact.

Since evolution is generally constrained to finding local optima while ID has no such constraint, can you see a testable prediction for very large complexes that will differentiate between the two hypotheses?

“My scenario would require a whopping does of luck before selection kicks in.  The selection wouldn’t need to be “whopping” at all, though.”

But you have no evidence for your scenario.


John - #43330

December 10th 2010

Bilbo:  “In my nine years of discussing ID on the internet, I’ve never read about tail or leg genes.”

——
Bilbo Says:
March 10th, 2009 at 10:39 pm
I might be mistaken, but I think the “toolkit” for development of the eye appears in most of the phyla.
——

“I fail to see the relevance of the quotation,…”

Really? Then you won’t hesitate to answer a simple question: what proportion of genes with roles in eye morphogenesis (morphogenesis, not merely physiology) play roles ONLY in eye morphogenesis?

“... though I am flattered that you follow everything I say so closely.”

I don’t. Google does.

“I’m curious.  Where did I write that?”

On TT, of course.


John - #43334

December 10th 2010

Mike:
“Thanks for the additional suggestion.  I had actually thought of that before, but didn’t follow through because it’s mostly the Arm repeats that are responsible for beta catenins sequence conservation.”

You’re welcome, and MET predicts that. That’s precisely why BLAST tells you up front where the conserved domains are, so that you can discard them!

“Nevertheless, since you raise it, I took the N-terminal 130 amino acids and the C-terminal 115 amino acids from human beta catenin, as the 12 Arm repeats are sandwiched between.  I used them to BLAST the Volvox protein and there is nothing there.  But as a control, I also used them to BLAST Aardvark and there is nothing there either.”

So there’s no homology to beta-catenin, just homology in the sense that Volvox has a fellow ARM domain-containing protein.

“Finally, I used them to BLAST a sponge beta-catenin.  The C-terminus has nothing, but there’s a 50 amino acid span with low 30% identity in the N-terminus that gets picked up.”

Only 30% in a region that small? Sorry, that won’t wash.


Page 2 of 5   « 1 2 3 4 5 »