When I first saw that the new book by Steve Meyer, Darwin’s Doubt, centered on the Cambrian Explosion, I was loathe to read it. I had been led to believe over the years that everything that could be said about the Cambrian Explosion has already been said. I was quite happy to believe that the only real discontinuities in the story of life occurred at the origin of life and at the origin of human consciousness.
I should have known better; science marches onward, and old arguments get reexamined as new data arises. Steve Meyer’s book is a wonderful, comprehensive case that the origin of the major types of animals, namely the phyla, is just as strikingly discontinuous as the origin of life. As such, it represents a solid second volume complementary to his previous work, Signature in the Cell, which focused on the origin of life.
I had come to think that discussing the Cambrian Explosion was misguided because of two arguments: 1) that the explosion was merely an artifact of the fact that organisms before that time did not have hard bones or shells, and 2) that the explosion was short on the geological time scale, but was really quite long on the biological time scale. Meyer disposes of both of these arguments quite handily. On the first, modern science shows that soft-bodied organisms are well preserved in the strata before, during, and after the Cambrian. Also, many of the body types which appear in the Cambrian can’t even be imagined without their hard parts to give them structure. An earlier, boneless version could not have had the same body plan at all. On the second objection, Meyer shows that the geological time scale has gotten more compressed over the years, not less; best estimates now are 5-10 million years, which is quite short geologically. Meyer then spends a good number of chapters establishing what the natural time scale is for evolution.
From a physicist’s perspective, I am used to thinking of time as a relative thing (for electrons in solids, a few trillionths of a second can be a long time, while for stars in clusters, a few million years can be a short time.) What makes something a short time or a long time is the natural time scale of the system– much less than the natural time scale is short, and much longer than the natural time scale is long. A fairly convincing case has been made in the literature of molecular clocks that the natural time scale for evolution of the degree seen in the Cambrian is a billion years, not 5 million years. Even that billion-year time scale may be an underestimate, if one looks at the microscopic details of protein folding. Thus the intrinsic biological time scale is not less than the geological time scale, and the Cambrian Explosion does indeed occur in a fantastically short time. Meyer cites many evolutionists who acknowledge this problem; the Cambrian problem has not gone away for those who are really in the know, no matter what popularizers may say.
This is a solid scientific review, not a polemic diatribe. It also comes at a good time. Like Signature in the Cell, it comes after 10-20 years of debate on intelligent design. Thus Meyer can summarize the back and forth of the debate in a nice story-like approach. The story is not one of gaps in our knowledge constantly being filled, but the paradox of the Cambrian becoming sharper and sharper. Again, when evolutionists talk to each other instead of to the public, they are remarkably candid about this, and Meyer well documents this with many quotes.
After posing the problem, Meyer discusses some of the non-orthodox, semi-Darwinian proposals floated in the last few decades, such as Gould’s punctuated equilibrium and epigenetic neo-Lamarkianism. All of these are built on a surprising amount of hand-waving, invoking new terms but brushing over the actual physical mechanisms. One section I was quite happy about was the section on “self-organization”, promoted by Kaufmann, Prigogene, and others. This area has had a strong following in the physics world for three decades, but I have always thought it was sterile, for the reasons that Meyer cites. Essentially, getting “order” from natural self-organizing process and getting “information” are two totally different things. “Order” is easy– all you need is a natural length scale to arise in a system and “spontaneous symmetry breaking” will lead to orderly patterns on this length scale. This is true of atomic crystals at low temperature and rows of clouds in the sky. But the very nature of information, whether in DNA or human writing, precludes natural forces from generating it. DNA can hold information precisely because there is no natural force demanding the nucleic acids be in one location or another. All information requires this type of “contingency”, that is, openness to many possible choices; a system which is driven to one required state holds no information. (Something I was not aware of before reading this book: there is another, equally information-rich, code in biological systems, known as the “sugar code”, which is written on the outside of cells to govern their interactions. Like the DNA code, there is no force driving the locations to hold one piece of information instead of another.)
And this is also the problem with identifying where the information came from. Many anti-ID critics demand that ID proponents identify the physical process by which the information came into being. But by its very nature, information is fungible–it can be exchanged into many different forms. Any system with many physical possibilities and no force driving the system to any of them can hold the same information. Thus the demands of the anti-ID critics are like a person who would demand that you deduce from reading a novel whether it was first written with pen and ink, or with a typewriter, or with a modern computer processor. While one can easily identify information when one has it, the very fact that information can remain the same while being embodied in any number of different media, makes it impossible to deduce a physical cause for it.
A few small things that I would have liked to see Meyer address: 1) in his discussion of the molecular clock data, he points out the variation in the numbers over a wide range, but doesn’t discuss at all the scientific concept of “uncertainty”. Having different numbers for the same measurement vary by a factor of ten or more does not mean the numbers are meaningless, unless the claimed uncertainty is much less than the scatter. 2) He mentions that the molecular clock data don’t work at all for histones, but doesn’t mention that the reason histones are highly conserved is because they are an integral part of the reproduction system– one change there and you die. A proper molecular clock calibration would be a “weighted average” in which each gene is weighted by the likelihood that a change will kill the organism. Apparently this has not been done in the literature yet in any quantitative way.
One of the fascinating side stories, which I have heard in ID circles for years but have not before seen documented as Meyer does, is the problem of making consistent genetic trees. I have often heard evolutionists, such as Francis Collins, make the argument for universal common descent by showing two genes in different species that have remarkable similarity but key differences, such as a fusion of two genes or a viral insertion. The argument basically goes: species 1 has the pattern A-B-C-D-E-F, while species 2 has the pattern A-B-C-X-D-E-F. What is the likelihood that these would be so similar in two unrelated species? Is this not clearly an insertion of X going from 1 to 2, or a deletion of X going from 2 to 1? Sounds good as far as it goes, but the problem comes when you try to do it for many more than two species. Let’s write this relationship as 1> 2. Suppose now that you look at four organisms, and find the relationships 1>2, 2>3, 3>4, and 4>1 in four separate genes. Can you make a consistent tree from that? What if I further tell you that 1 is a plant, 2 is an insect, 3 is an animal, and 4 is a worm? Now, this is a fictional example, but are you willing to bet the farm that no such relationship can exist in nature? It turns out that relationships like this are all over the place. To explain it, some evolutionists invoke “convergent genetic evolution”, which means that that same gene (same sequence of DNA) arose two times, independently. I could sort of buy convergent structural evolution (e.g. placental wolves and marsupial wolves that look nearly identical but have very different DNA), but convergent gene sequences? It defies the imagination. I once met a German scientist who told me he lost his faith in Darwinism after realizing he could not make self-consistent genetic trees (but he is not willing to come out of the closet out of fear for his career). In general, although I don’t think there are a lot of theological stakes in the question of universal common descent, I am surprised at how weak the case for it is.
Meyer ends with general thoughts on ID, similar to his arguments at the end of Signature in the Cell. His experience, like mine, is that some people literally can’t “see” God as an explanation, because they have defined God-explanations as non-explanations. Meyer doesn’t go into detail about the jump from knowing what human intelligence can do, to invoking non-human (presumably divine) intelligence as a similar causal agent, but the case can be easily made. I addressed this in an essay in PSCF.
Overall I don’t expect this to change the views of diehard atheist evolutionists, but I would hope that my theistic evolutionist friends will give this book a close reading. A caution: this is a tome that took me two weeks to go through in evening reading, and I am familiar with the field. Like the classic tome Goedel, Escher, Bach, it simply can’t be gone through quickly. I was struck that the week it was released, within one day of shipping, there were already hostile reviews up on Amazon. Simply impossible that they could have read this book in one night.
I am about 2/3 through the book. The first part seems to be a pretty good review of the Cambrian explosion. Then you get to the evolution part and one has to start swallowing camels. I will go back and document these wait a minute statements and arguments he makes where he passes off a “heavy ball falls faster than a lighter ball” proposition which the readers accepts as so eminently reasonable and then he goes on to make his case. Let me give one example. Surely the Cambrian explosion involved the introduction of new protein folds. Now, Meyers does not know that. He does not know how many folds were already present at the time. Eukaryotes had been extant for 2 billion years.and as far as we can tell, there is no such thing as a primitive eukaryote and therein lies most of the complexity. Furthermore, studies on folds show that the same basic fold can be used in a variety of unrelated proteins and unrelated protein functions. So the argument that many new folds are needed at the Cambrian explosion is without foundation. But it sounds good.
To add a bit more, I think Meyers could be so much more compelling to examine some wider data than just what Doug Axe says. I do not know Doug and have nothing against him. Furthermore, I could not care less if the Darwinian paradigm falls apart on the Cambrian explosion. But there is what is known as the Ig superfamily of proteins which contain, as the name suggests, the Ig fold. Members of the Ig superfamily are involved in homotypic adhesion (the foundation for making tissues), receptors, signaling proteins (tyrosine kinases) and of course antibodies and T cell receptors. Now here are two points that are interesting. First, one fold is used for many different and relevant functions, and one in particular that lies at the heart of the Cambrian explosion – mulicellularity – which involves homotypic adhesion. Secondly, what is the sequence variability of the Ig fold in members of the superfamily. If it is as constrained as Meyer portends, then we should see it in the sequence data.
So Meyer argues that exon shuffling cannot work because of interactions between side chains. I am not sure that he understands that domains are autonomously folding regions. But, lets take the example of green fluorescence protein – essentially just a beta-barrel fold. Since I cannot mix and match domains, therefore, I cannot fuse green fluorescent protein to another protein to make a functional chimera because of side chain interactions. Hmm – this is done routinely by biologists (including those in my lab) all over the world and both the fluorescent protein and its fusion partner, whatever protein it is in the cell that one is studying, continue to work just fine. Occasionally it doesn’t work because the GFP gets in the way but in a great majority of cases it does.
Previously I said that Meyer’s assertion that folds are the minimal selectable unit of information was just outrageous. Furthermore, his argument is that building new folds in protein sequence space is vastly improbable and since the Cambrian explosion requires many new folds, therefore, his argument prevails – that there is no natural or mechanistic explanation for mechanism of the Cambrian explosion. I would like to examine this argument.
One of the hallmarks of the Cambrian explosion is multicellular organisms. More than that these organisms are composed of differentiated tissues, So, lets ask, “What does it take to make a tissue”? Fundamentally it takes a group of cells that stick to themselves better than they stick to other cells. This has been shown to depend on homotypic cell surface adhesion proteins. Homotypic adhesion proteins are proteins that bind to themselves and they are mediated largely by two families, the CAMs and the Cadherins. In other words, there are tissue specific CAMs and cadherins where each family specifies different homotypic adhesion proteins. So, for example we have N-cadherins, E- cadherins, P cadherins, T cadherins, and so forth, each mediating different tissue-specific homotypic binding. Now, these are all within a single family of proteins. A family of proteins is a group of different proteins that are characteristically contain a common fold. So, we conclude that the different proteins that mediate different tissue specific interactions all share the same fold.
Thank you for your fantastic comments, gandaulf!
Gandaulf, I find your comments interesting, but you don’t addres much key evidence Meyer discusses. You write, “I think Meyers could be so much more compelling to examine some wider data than just what Doug Axe says”. But in my reading of Darwin’s Doubt, I saw that Meyer discussed a lot of data that goes far beyond Doug Axe’s work.
For example, When analysing the combinatorial inflation problem, he discusses work by Michael Behe, David Snoke, Ann Gauger, Ralph Seelke, Rick Durrett, Deena Schmidt, John Mayndard Smith, Robert Sauer, and others. So I don’t think your criticism is fair.
Indeed, one paper Meyer cites in Chapter 2 co-authored by Ann Gauger (and Axe), looked at two enzymes that were structurally VERY SIMILAR and yet found that many simultaneous changes would be required to convert one into the other–more than could be accomplished by Darwinian evolution. Here’s what Meyer wrote on this point:
“Having carefully examined the structural similarities between members of a large class of structurally similar enzymes, they knew that Kbl2 and BioF2 were about as close in sequence and structure as any two knowjn proteins that performed different functions. Thus, if it turned out that converting one protein function into the other required many coordinated mutations— more than could be expected to ocur in a reasonable time— then the outcome of their experiment would have devastating implications for standard accounts of prtien evolution. If proteins that perform two diff erent functions have to be even more similar than Kbl2 and BioF2 in order for mutational changes to convert the function of one to the other, then for all practical purposes co- option would not work. There simply aren’t many known jumps that small. Axe and Gauger first identified those amino-acid sites that were most likely, if mutated, to cause a change from Kbl2 function to BioF2 function. They then systematically mutated those sites individually and in groups involving various amino-acid combinations. Th eir results were unambiguous. They found that they could not induce, with either one or a smal number of amino acids, the change in function they sought. In fact, they found that they could not get Kbl2 to perform the function of BioF2, even if they mutated larger numbers of anino acids in concert— that is, even if they made many more coordinated mutatoins than could plausibly occur by chance in all of evolutionary history. Although their attempts to convert Kbl2 to perform the function of BioF2 failed, their experiment did not. It allowed them to estabhlish expermentally for the first time that the co-option hypothesis of protein evolution lacks credibility — simply too many coordinated mutations would be required to convert one protein function to another, even in teh case of extremely similar proteins. That implied that generating new genes and proteins would require multiple coordinated mutations, and thus, the waiting times that Behe and Snoke had calculated do present a problem for neo-Darwinian theory.”
Meyer then explains that the conversion would require more simultaneous changes than could arise over the history of the earth. He writes: “Indeed, Axe and Gauger’s experiments showed that the smallest realistically conceivble step exceeded what was plausible given the time available to the evolutionary process. In their words, ‘evolutionary innovations requiring that many changes … would be extraordinarily rare, becoming probable only on timescales muchlonger than the age of life on earth.'”
In sum, here’s what’s noteworhty about that paper by Gauger and Axe:
In that experimental research, they WERE NOT tying to generate a new fold and yet the conversion of proteins could not be accomplished without MANY simultaneous mutations. So this directly challenges your claim that if a new fold isn’t required, and we’re talking about (as you noted) “A family of proteins … that are characteristically contain a common fold” then the required changes can’t always arise via blind evolutionary processes!
So I think it is NOT established that even proteins within the same family could arise by Darwinian evolution.
////In fact, they found that they could not get Kbl2 to perform the function of BioF2, even if they mutated larger numbers of anino acids in concert///
.
This flawed idea has already been debunked a while back. Kbl2 and BioF2 are two extant proteins. Evolution doesn’t claim that you can convert two extant proteins into one another. That will be like claiming that the number of changes required for me to evolve from my cousin is improbably high. Thus Gauger & Axe were “testing” a prediction evolutionary theory never made in the first place!
In addition they fail to consider neutral mutations. If multiple changes are required to convert one ancestral protein to a descendant protein, all those changes doesn’t have to happen in one fell-swoop event and every single change doesn’t have to alter the function. Some of those mutations can be neutral that causes no apparent change in function, yet serves as a backdrop on which more mutations can become relevant. This was elegantly shown in the following paper:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3116920/?tool=pubmed
If Stephen Meyer is still citing the flawed work of Gauger & Axe to support his already outlandish claims, then that’s another reason to throw his book into the dustbin.
Borny, I find it highly suspicious that you seem very eager to throw books away (“in the dustbin.”) Actually, Axe and Gauger have been refuting this fallacious criticism for years, and today, it’s only made by people who aren’t paying attention to what ID proponents are saying:
Ann Gauger explains why this study is intended to be a much stronger “disproof of concept,” not a direct test of a historical transition (a more limited argument) in an ID the Future interview at:
http://intelligentdesign.podomatic.com/entry/2011-05-16T17_01_43-07_00
Ann Gauger and Douglas Axe also respond to this objection here:
http://www.biologicinstitute.org/post/27516628899/on-enzymes-and-teleology
In other words, this isn’t a flaw in the paper, but it makes it far more applicable to questions of protein evolution.
Also, the basic question of the paper WAS to test neutral evolution–that’s why they tried to induce MULTIPLE MUTATIONS before any functional change would occur. Axe’s 2010 paper in BIO-Complexity tested the neutral model and established limits for how much neutral change could be fixed in a large population of continuously evolving bacteria. Axe and Gauger’s 2011 paper showed that this simple protein-protein conversion could NOT be accomplished within those limits. This poses a major problem for protein evolution.
No amount of beating-around-the-bush will save Axe & Gauger’s flawed work. Why do you think their paper was published in their own non-journal Biocomplexity? If the work had any merit it would have found a real mainstream scientific journal.
Axe says:
///We also knew that in order for a Darwinian process to generate the mechanistically and chemically diverse families of enzymes that are present in modern organisms, something like the functional conversion of one of these enzyme to the other must be possible.///
This is like saying that the diversity of life on earth was derived by converting one existing life form to another. For eg: humans were derived from chimps. It is not difficult to see why this is a flawed idea. Extant proteins, like extant life, evolved from a common ancestor, not from each other. Each lineage evolves separately from its cousins after splitting from their common ancestor. Therefore converting one extant form to another will be difficult. Tell me, how easy will it be to convert a chimp into a human? Both species are very closely related, yet it will be a monumental task trying to convert one species to the other. But this difficulty doesn’t mean common descent is wrong or evolution didn’t happen.
Moreover, as the following paper shows, functional divergence of duplicate genes also results from structural divergence by processes involving insertions & deletions and gain & loss of exons and introns. It is not just about amino acid substitutions.
pnas.org/content/109/4/1187.short
If Axe & Gauger really wants to do proper science and not something for the sake of bashing evolution, then they should listen to their critics and decipher the ancestral sequence of their proteins. Constructing their phylogeny will show what mutational paths each protein took. That’s what the authors have done in the paper I cited in my first post above.
I’m not all that informed on these issues, so, correct me if I’m wrong. If your beef with him is not that he did not reverse engineer a protein to make it like an ancestor protein, I think you are misinformed as to the purpose of his argument. He is not saying that proteins cannot be reverse engineered, he is saying that the total amount of possible proteins sequences/folds is much greater than the amount functional protein folds, thus, as mutations do not have the benefit of knowing which mutation will be beneficial, it would seem his logic holds. His experiment may have been misleading, but, if it weren’t, it seems that his point would hold.
I’m also surprised that you took Axe to mean what you claim he meant as you quoted him above. He specifically said “something like.” He said this right after saying these proteins were obviously not in an ancestor/progeny relationship.
So my comments went public at the http://www.evolutionnews.org site. I thought that the “society” was a forum where we argued among ourselves. Be that as it may, it is fine. I will respond here to one of Axe’s remarks on ORFans. ORFans are “Open Reading Frames” that are unique to one or a few closely related organisms. Open reading frames are one requirement in determining that a DNA sequence is a gene. Molecular biologists usually require more evidence than just an open reading frame to call something a gene. Showing that it is expressed is normally required.
As more and more genomes were sequenced more ORFans were found. Originally it was estimated that 30% of ORFs were ORFans in each organism whose genome was sequenced. This number has come down. Currently the estimate for prokaryotes is closer to 15%. For purposes of my reply, it does not matter what the exact percentage is.
In reply, I would say Axe’s argument seriously backfires. ORFans are present in all organisms; prokaryotes and eukaryotes as well as in bacteriophages and animal viruses. The consistent percentage of ORFans in genomes argues both that it is not a remarkable feature of the Cambrian explosion and that they are most likely produced by mundane processes that operate in all genomes. In one particularly informative example, Toll-Riera et al. (2009. Mol. Biol. Evol. 26, 603–612) identified 270 primate-specific ORFans. Of these, 70% contained a transposable element. In other cases, where a particular gene appeared to be ORFan the same organism had a paralogue that did show homology with other organism suggesting that the ORFan in question underwent rapid divergence. It is likely that a number of different mechanisms contribute to ORFan formation and while a lot remains unknown, we are gaining in our understanding. The point being that contrary to Meyer and Axe, ORFans represent genes formed by normal genetic mechanisms. Meyer’s argument is that forming new genes is next to impossible.
I would also like to comment on Axe’s statement that few genes are represented in all kingdoms. I think that is true. However, it is not true when one restricts the range to the metazoans. So, I am not sure of his point. Weren’t we talking about the Cambrian explosion?
Final note, I have a number of grants to review before the end of the month and editorial duties on top of my normal lab engagement (did someone say this was summer vacation?) and my engagement will necessarily be sparse here.
Just FYI for all members, “posts” (such as this one) and comments on them are completely public. We also have a “forum” which is closed to members only. Anyone is welcome to start a private discussion as a new thread on the “forum”. Forum discussions are also unmoderated, as opposed to public comments in which a moderator has to approve them before they go public. Though, of course, lack of decorum in the private forum could get a person kicked off, at least temporarily.
Re: ORFans, it seems to me there is a bit of a heads-I-win-tails-you-lose argument in some cases. What I recall is evolutionists arguing that homology proves that there is no design, because clearly if God created the species, he would have created each species with an new entire genome de novo. But when we find that a large fraction of the genetic code is apparently de novo, it is argued that this proves that de novo code is ubiquitous and therefore a standard product of evolution, and can’t be used as evidence of design.
What Martie appears to be arguing, though, is that the ORFans in many cases are not totally disconnected to other genes. So they are not truly de novo?
No they aren’t anything supernatural. Existing genes duplicate and the extra copy can mutate to become an ORFan. Even an existing non-coding sequence can mutate to become an ORFan.
///if God created the species, he would have created each species with an new entire genome de novo.////
.
De novo as in “making it from scratch”. The so-called de novo genes we find are not made from out-of-the-blue, but from pre-existing genes/DNA by processes we observe and understand. In other words, we don’t need to invoke a God to account for them.
Being an ORFan simply means that at the time a genome was sequenced there were no obvious homologous representatives in the protein databank. As time goes on, and with more detailed scrutiny, some of these have been found to have homologies and they are moved out of the ORFan category. But many remain. It is rather easy to see how ORFans can be generated when the genome is understood as a dynamic assembly that is modified by both new gene formation and by gene loss. As the paper I cited notes, there are a variety of genetic mechanisms that can create orphan genes – such as gene duplication, frame shift fixation, creation of overlapping genes, horizontal gene transfer and exaptation of transposable elements. Loss of genes can break the connections between organisms that might have once shared these genes.