It is much easier for a mouse to get a set of genes which enable it to resist Bacillus typhimurium than a set which enable it to resist cats.
-- J.B.S. Haldane
The dead are dancing with the dead.
-- Oscar Wilde
Stephen Meyer's book Darwin's Doubt makes three claims: That the Cambrian explosion was real; that it remains unexplained; and that these facts sanction, or support, an inference to intelligent design. Writing at Panda's Thumb, Nick Matzke has denied the first, rejected the second, and ignored the third ("Meyer's Hopeless Monster, Part II"). What a man ignores is his business. There is little point in demanding that Matzke assess an inference to which he is indifferent. If the Cambrian explosion was not real, Matzke is surely right to reject claims for its explanation. The Cambrian explosion, Matzke is persuaded, was a tedious, long-drawn Darwinian affair, one lasting for thirty million years and taking forever. This position has been warmly endorsed by Donald Prothero, an encomium, one might think, as impressive as one wrung from the lips of a Kardashian. Yet even if the Cambrian explosion was not real as an explosion, it remains real as an event. Cambrian creatures are in their body plan, nature, way of life, and order of complexity unlike their predecessors in the Ediacaran ooze. Having seen nothing new in the Cambrian era, it is hardly surprising that Matzke is persuaded that there is nothing new to see. "[T]here's no evidence," he writes, "that new protein domains were required in the Cambrian -- I'd be surprised if any protein domains are known that are both unique to and required for the existence of Animalia."
Writing in the Proceedings of the National Academy of Sciences, Susumo Ohno came to a different conclusion. He is disposed to see more than Matzke, perhaps because his threshold of astonishment is lower. It could not, of course, be higher. "It now appears," Ohno writes, "that this Cambrian explosion, during which nearly all the extant animal phyla have emerged, was of an astonishingly short duration, lasting only 6-10 million years." Thereafter Ohno draws the conclusion drawn in Darwin's Doubt. New proteins did not originate in the Cambrian by means of an incremental nibble. " ... [I]t is more likely that all the animals involved in the Cambrian explosion were endowed with nearly the identical genome, with enormous morphological diversities displayed by multitudes of animal phyla being due to differential usages of the identical set of genes." What Ohno may have meant by the curious and suggestive word endowed, he does not say.1
The Cambrian genome was distinctive, Ohno argues, in five respects: It contained (i) a gene for lysyl oxidase, a protein that in the presence of molecular oxygen crosslinked collagen triple helices to produce ligaments and tendons; (ii) genes for hemoglobin; iii) genes for glass (silicified) skeletons; (iv) the Pax-6 gene for eye formation; and (v) a series of Hox genes for the anterior-posterior (cranio-caudal) body plan. Ohno's argument is offered on the level of molecular genetics, and not protein chemistry, but if Cambrian animals required a gene for lysyl oxidase, it is because they required lysyl oxidase as a protein. By parity of reasoning, if they incorporated a new and distinctive genome, presumably they required a new and distinctive suite of proteins. The idea of a purely decorative Cambrian genome is not a contribution to biology.
In an article published in Science, Chothia et al. added circumstances to what is largely a circumstantial case.2 While 429 families of protein domains are common to all eukaryotes, they argue, 136 are unique to animals. These proteins presumably arose after the last common animal ancestor. How otherwise could they be unique? If they arose after the last common animal ancestor made his last stand, the victim, no doubt, of passive smoking, they must have arisen throughout the Cambrian era. By what mechanism of arousal? Chothia et al. offer only the vaguest of speculations. The evidence is suggestive: It is not conclusive. Suggestive evidence is better than none at all. Beyond appealing to surprise as a factor in his deliberations, Matzke has offered no evidence at all. He is in this regard serene.
In arguing against a Darwinian explanation for the emergence of complex new structures, whether in the Cambrian era or any other, Darwin's Doubt appeals to arguments of long standing and continued controversy. Skeptics about Darwin's theory have for almost all of the 20th century appealed to the same triplet to express their skepticism: The complexity of biological structures, the random nature of the search required to find them, and the limited time available to conduct the search. Popular accounts of Darwinian theory have been devoted to mountain climbing metaphors or mummeries, the progression by cumulative selection toward some cleanly defined optima.3 Discussions have been pre-theoretical; indeed, they have been pre-scientific. They have barely counted as discussions.4 No one need argue that complexity could be better defined because complexity has never properly been defined at all. But for more than thirty years, it has become clear enough that hill climbing is an irrelevance in evolutionary thought. The exquisite complexity of various biological structures cannot be explained by means of a strategy of no greater intellectual depth than Twenty Questions.5 The path to virtually any complex structure must proceed by some scheme of deferred gratification.
In this regard, Matzke is of the Old School; his allegiances are to the Old Breed, biologists determined to face the facts by vigorously ignoring them. "The multiple-required-mutations stuff," he writes, "...is basically just Behe's refuted ‘irreducible complexity' argument disguised as an argument about sequence evolution, and is only relevant if it can be shown that 2 or more neutral mutations ever were required for anything relevant to the Cambrian Explosion, but, as is typical in DI literature, this is just blithely assumed rather than argued for."
Michael Lynch is the last man on earth likely to offer support to any theory of intelligent design. He is about as blithe as a canine incisor. His support is not needed. It quite suffices that he denies what Matzke affirms. "... [A] broad subset of adaptations," he writes
cannot be accommodated by the sequential model, most notably those in which multiple mutations must be acquired to confer a benefit. Such traits, here referred to as complex adaptations, include the origin of new protein functions involving multiresidue interactions, the emergence of multimeric enzymes, the assembly of molecular machines, the colonization and refinement of introns, and the establishment of interactions between transcription factors and their binding sites, etc. The routes by which such evolutionary novelties can be procured include sojourns through one or more deleterious intermediate state...(emphasis added).6
Is this "basically just Behe's argument disguised as an argument about sequence evolution?" It is. But with the caveat, of course, that Behe's argument, if it has been widely rejected by the Old Breed, has not been persuasively refuted by any of them -- or by anyone else. "It is indeed true," Jerry Coyne remarked in The New Republic, "that natural selection cannot build any feature in which intermediate steps do not confer a net benefit on the organism."7 Having recognized a challenge to right thinking in principle, Coyne is, of course, prepared to deny its significance in fact. Irreducible complexity? There is no such thing, Coyne remarks, and he has looked, too. In this, he is very much like a man wandering in a bakery and wondering petulantly where the bread might be. Under your nose is in both cases the instructive answer. Definitions are one thing; the real world, another; and irreducible complexity is not in doubt as a fact of life. A number of biologists like Michael Lynch quite understand this; they have pitched their tents in their enemies' camp. They appear to be quite at home.
Whether the Cambrian explosion was ignited by a short fuse or one that was long, the fact remains that having sputtered, the fuse in the end fired, the ensuing explosion producing new Bauplans or Baupläne, new phyla, new creatures entirely, such as Hallucigenia, all bristling porcupine quills and drooping snout, the director ancestor, as it happens, of both Stephen Jay Gould and Simon Conway Morris, new skeletons, new tissues, and new nerves, new brains and new eyes, the first to control the second and the second to inform the first, and to accommodate all this newness, new genes and the proteins they express, the staff and the stuff of life on every man's account. Old Breeders see this as a part of the flow, the endless incoming and outgoing tide of variation and incremental change, one protein folding in onto itself and giving way to another. This thesis, Darwin's Doubt rejects. The argument that it makes depends upon and amplifies research conducted by Doug Axe and Ann Gauger.8 Their conclusions are similar to those reached by Mike Behe in The Edge of Evolution, but careful historians of biology will notice that these arguments, when taken collectively, have a distinctive history of their own, similar arguments having been made long ago and then forgotten.9
Axe and Gauger begin by drawing a distinction to which critics such as Nick Matzke are demonstrably insensible. It is, that distinction, of the essence. "Functional innovations throughout the history of enzymes," Axe and Gauger write, "may be divided into two categories based on the degree to which they depend upon structural innovation." Big Time innovations are those contingent on a "fundamentally new structure" and thus upon a new protein fold. Small Time innovations are matters involving trimming or tightening or tinkering with an existing fold. It is the difference between cutting a new pattern and embroidering an old one. Axe and Gauger examined two proteins: KBl2 and BioF2. Although structurally similar, they perform different functions. They are in their active identities distinct. How difficult would it be to change one protein so that it acquires the function of the other? To the extent that KBl2 and BioF2 are generic proteins, off the shelf and so off the cuff, the answer demanded by Darwinian theory is unequivocal: It should be pretty easy. Darwin's theory of evolution is above any other consideration a theory of continuous transformation and if small steps can do wonders over the course of evolution, why not in the laboratory? One good question deserves another. How difficult would it be selectively to breed a cat that barks in the night? The inescapable answer to both questions in the world in which facts are plain is that it would be exceptionally difficult.10 Difficult enough, indeed, to prove unlikely in nature and impossible in the laboratory. No matter the extent to which they tugged or pulled at KBl2, Axe and Gauger never achieved a variant capable of executing the functions of BioF2, the original protein remaining obdurate in its attachment to itself. The observation that whatever the selective pressure, a cat remains a cat, is as pertinent in protein chemistry as in zoology. If they were never able to do what they proposed, they were able to judge what it would require. A "successful functional conversion," they write, "would in this case require seven or more nucleotide substitutions."
They drew the obvious conclusion: The functional conversion of proteins is not possible under orthodox Darwinian scenarios. The job in question is too difficult for the time at hand.
Having read with irritation, or studied with indifference, Axe and Gauger's work, Matzke has remained adamant in his animadversions. If new proteins were needed in the Cambrian era, they were easily derived. Cambrian organisms had only to look around. Things were much easier than Darwin's Doubt suggests. Stephen Meyer has made far too much of far too little. Ah, the Old Breed. In this case, Matzke is not alone. He worships among a crowd of Old Believers, Jerry Coyne among them, genuflecting spastically. Axe and Gauger, they are persuaded, have surrendered to a pointless pessimism about what evolution might accomplish. If belief engenders belief, the facts are entirely less forthcoming. In 2003, Long et al. published an interesting paper entitled "The origin of new genes."1 It is this paper to which Matzke appeals, often with satisfaction, always with assurance. Long et al. argued that the genes Sdic, Sphinx, Jingwei, RNASE1, and AFGP, among others, have originated, or evolved, within the last few million years, and this by what is essentially the old fond familiar process of typographic variation within the genome: Cut, Snip, Fiddle, Transpose, Shuffle, Replace. It is easy to see why Old Believers should find this paper rewarding. It looks so simple.
Those unforthcoming facts: In every example cited by Long et al., nothing like a new protein structure is in evidence; Long et al. remain throughout within the ambit of Small Time innovations, and from this ambit, they never depart:
A recently evolved chimeric gene encoding a protein in the species Drosophila melanogaster. Considering the order and sequence similarity of the surrounding genes, AnnX and Cdic, it is both easy and invigorating to imagine a process of duplication, deletion, fusion, and sequence rearrangements transforming AnnX and Cdic into Sdic.12 Matzke says as much: He has lovely colored pictures to pass around. His analysis has been corrupted throughout by a systematic equivocation between genetic novelties and novelties in protein chemistry. The first is no good guide to the second. Cdic is a gene encoding a dynein intermediate polypeptide chain, one expressed in the cytoplasm. But Sdic also encodes a dynein-specific intermediate folding chain, one expressed in the testis. Save for the fact that it is truncated at the N-terminal, Sdic has the same structure as Cdic. It is expressed in sperm because the gene has acquired a new promoter from its fusion with AnnX.
The argument at issue concerns enzyme specificity and novel protein folds, not promoters and expression patterns.
In a stimulating paper entitled "Functional evidence that a recently evolved Drosophila sperm-specific gene boosts sperm competition," Shu-Dan Yeh et al. reported on work in which they deleted the region of the chromosome containing Sdic copies in order to test the effect of Sdic on fitness.13 They subjected male flies with and without Sdic to mating rounds with females. The results were anything but provocative, least of all to the flies, and not at all to Shu-Dan Yeh et al. There was no discernable effect on the fly phenotype. Under benign conditions, those individuals without Sdic left as many progeny as those with Sdic. In multiple rounds of mating, when a female was first subjected to a mutant male without Sdic, and then one with, the presence of the gene gave no statistically significant advantage in sperm displacement. Repeating the experiment in reverse did lead to an advantage, but the relevance of this result to any larger issue of principle remained obscure. This is where it remains today.
The emergence of Sdic thus suggests a weak form of microevolution, so weak as to be embraced with equanimity by those proposing to champion, and those prepared to deny, the power of Darwinian theory. A protein dispensable to function, uninteresting in its chemistry, and with no control over morphology, has evolved by stochastic means. Whoever thought to deny it?
The lysyl oxidase family is otherwise.
A chimeric gene in Drosophila as well. Something of a hodgepodge, it was apparently scavenged from neighboring gene segments, a retroposed sequence of the ATP synthase F-chain gene from chromosome 2 obligingly inserting itself into the 102F region of chromosome 4. And the result? Nothing at all, as it happened. The issue is the subject of research conducted by Hongzheng Dai et al.14 Their analysis followed the characteristic Darwinian trajectory in which having prominently puffed up a claim, its authors were then obliged unobtrusively to puff it down. The puffing up: "[C]himeric genes often evolve rapidly," they write, "suggesting that they undergo adaptive evolution and may therefore be involved in novel phenotypes," (emphasis added). The puffing down: "... [A]lthough [Sphinx] is derived, in part, from a protein-coding gene, it is most likely a noncoding RNA (ncRNA) because its parental-inherited coding regions are disrupted by several nonsense mutations."
A gene too impotent to produce a protein is hardly what Matzke requires to rebut Darwin's Doubt.15
He must do better.
He could not do worse.
Well known for being well known, Jingwei is a chimeric gene made from the Alcohol Dehydrogenase (Adh) gene and another called yande. The protein expressed by Jingwei is a member of the broad and noble class of enzymes that degrade alcohol (friends to humanity, if nothing else). Its parent gene, Adh, serves the same function. In a paper to which Long is a contributor, Jiaming Zhang et al. remark that
Drosophila ADH belongs to the short-chain dehydrogenase/reductase (SDR) family ... SDRs share a common protein fold, consisting of a central β-sheet surrounded by α-helices and a typical nicotinamide coenzyme binding βαβαβ subdomain, with a characteristic Gly-Xaa-Gly-Xaa-Xaa-Gly motif ... Asp-37 confers specificity toward NAD binding, whereas the active site is characterized by a Ser-Tyr-Lys catalytic triad .... These and other conserved SDR features are preserved in JGW ..."
The familiar deflationary wheeze now follows: "We predict that JGW will retain NAD-specific dehydrogenase activity" (emphasis added).16
No new folds. No new structures. No new nothing.
But, of course, something slightly different. When compared to AdH, it is plain enough that Jingwei is both more effective and more specific in degrading long-chain alcohols. These modest improvements in specificity did not simply emerge all at once, like Venus emerging from a clamshell in Botticelli's famous painting: Adh encodes a polypeptide that can metabolize a diverse range of alcohols as well. Jingwei thus augments the functions of an ancestral AdH enzyme. "Substrate specificity of JGW," Jiaming Zhang et al. write, "was further characterized in a survey of 34 alcohols that included representatives from all major classes found in nature ... Like ADH, JGW shows activities toward a broad range of alcohols. However, compared with ADH, JGW also shows a systematic preference for long-chain primary alcohols and increased specificity... including farnesol and geraniol ... These results confirm that JGW has evolved altered specificity after diverging from its parental genes, Adh and ynd."17
Improvements? Yes. Structural innovations? No. Relevance? None.
A gene destined for the short-term glory peculiar to genetic publicity campaigns. "The origination of new genes," Zheng et al write, "was previously thought to be a rare event at the level of the genome ... However, it does not take many sequence changes to evolve a new function (emphasis added). [W]ith only 3% sequence changes from its paralogues, RNASE1B has developed a new optimal pH that is essential for the newly evolved digestive function in the leaf-eating monkey" (emphasis added).
This is interesting, exciting and false.
a) The new and improved optimal pH promoted by RNASE1B does not involve a new function, let alone a new protein fold. The fold remains the Ribonuclease fold. Colobines are old world monkeys that that eat leaves and employ symbiotic bacteria in their foregut to digest cellulose. The bacteria are themselves digested in the small intestine. To efficiently recycle nitrogen from RNA in quickly growing intestinal bacteria, it is better that the expression levels of RNASE1B be higher in the small intestine than in the foregut. And better that the optimal pH for its enzyme be lower, since the intestinal environment itself has a pH of between 6 and 7. The adaptation involved in lowering the optimal pH from 7.4 to 6.7 is accomplished principally by three forward mutations, something Zheng et al demonstrated by reconstructing and then expressing the ancestral sequence in bacteria. Do the mutations altering RNASE confer a tangible benefit? They do not. There is no increased catalytic activity for the enzymes operating at the lowered optimal pH as opposed to the raised optimal pH. The mutations that lowered the optimal pH caused the enzyme to lose other features. Nothing in Zheng et al. account suggests that the mutations represented a complex adaptation.
b) Long's assertion that lowered optimal pH was essential for digestive function is a more urgent claim. It suggests the existence of tight functional constraints on changes leading to a complex adaptation. But whatever Long may suggest, the facts are otherwise. They so often are. "Although unproven," Zheng et al. write, "it is generally believed that foregut fermentation and leaf-eating emerged in the common ancestor of all colobines. Fossil evidence suggests that these changes occurred at least 10 Myr ago predating the duplications of RNASE1."
What follows is emphatic: "The shift in optimal pH of the pancreatic RNases was not necessary for the changes in diet and digestive physiology of colobines."
Not necessary, meaning not essential. "Rather, the latter changes provided a selective pressure for more efficient digestive RNases in acidiﬁed environments, while gene duplication offered raw genetic materials that enabled this functional improvement."
Only the monkeys are apt to remain impressed.
A member of the class of antifreeze glycoproteins used by Antarctic fish. Mike Behe is correct to note of these proteins that they have nothing to do with the demands of complex adaptation. Antifreeze proteins are not specific; they contain no robust secondary structure; they are poor in information and highly repetitive, surviving on a simple Thr-Ala-Ala repeat; and they do not interact with other proteins.
The features that AFGP employs to bind ice crystals would not work in forming lysyl oxidase. They would not work for any similarly specific protein family invented in the Cambrian. That required fold stability and well-defined tertiary structure for atomic-grade precision in orienting atoms to increase catalysis. AFGP is little more than a grunt-level blunt instrument.
Like so many other biologists, Matzke is persuaded that if in the modern protein theater, functional conversion is difficult, then in the ancient theater of the proteins, it must have been easy. This is as close to a transcendental deduction as an empirical science affords. But however valuable it may be as a metaphor, Deep Time is in the real world fickle as a factor, and while things may well have been different long ago, it hardly follows that they were easier. Various hypothetical scenarios tend to cancel one another. The play of forces, and the ensuing annihilation of advantage, is evident in Ohno's model of duplication and divergence. Where previously it had one gene, duplication provides an organism with two. One gene does the heavy cellular lifting and takes the obvious selective risks; the other is free to explore sequence space and serendipitously find new things to do. Genetic affairs do not get more flexible than this. This is surely a step in the right direction, no? It is by no means clear. Protein perturbation studies indicate that ~40% of all mutations "reduce or completely abolish function," a substantial portion (8%) leading to the "loss of all functions" (emphasis added). The rate of beneficial mutations, by way of contrast, stands at 103 or 0.1%. Absent selection, any duplicate will be crippled far faster than it will accumulate beneficial mutations.18 Selection is, of course, unavailing. It is the other gene that is busy testing its luck in the real world. The ancient theater of the proteins may well have contained proteins flexible enough to begin things and cause a commotion, but what good a starting point if there is nowhere to go?
These considerations are hardly new: They have been long in rattling around the literature. The accommodating story is now of a sort standard in Darwinian theory in which a happy ending is demanded well before the story ends. "Evolutionary biologists agree," Austin L. Hughes remarks, "that gene duplication has played an important role in the history of life on Earth, providing a supply of novel genes that make it possible for organisms to adapt to new environments." Something makes it possible for organisms to adapt. This is the happy ending. "But it is less certain," he goes on to add, "how this panoply of new functions actually arises, leaving room for ingenious speculation but not much rigor." Ingenious speculations? Not much rigor? This is less happy. "Cases where we can reconstruct with any confidence the evolutionary steps involved in the functional diversification," he goes on to say, "are relatively few." Not much confidence? This is not happy at all.19
Ingenious speculations are required when it comes to protein evolution, because nothing better is available. Ancient proteins may well have been promiscuous in their affinities; mutations may have come in compensatory pairs, the good cancelling out the bad and vice versa; and those ancient proteins might well have embarked on weirdly successful random walks along neutral evolutionary networks. These phenomena are real enough, but they are more efficient within the context of a cell than they might otherwise be in establishing the context of a cell. A world in which fitness costs are steep is a world carrying ancestral organisms into fitness bankruptcy. How did these organisms survive? How did they do what organisms must do -- condense chromosomes, translate proteins, harvest electrons? And to what constraining evolutionary pressures did they yield or succumb? We do not know. The questions are open.
What we do know is that questions of this kind are themselves promiscuous and reproduce freely. A recent study in Science indicates that in the case of antibiotic resistance, adaptive challenges can block a majority of Darwinian pathways. One relevant study showed that resistance to cefotaxime required 5 mutations. There are thus 5! =120 different trajectories for evolution to consider. But as the authors note: "In principle, evolution to this high-resistance beta-lactamase might follow any of the 120 mutational trajectories linking these alleles. However, we demonstrate that 102 trajectories are inaccessible to Darwinian selection and that many of the remaining trajectories have negligible probabilities of realization, because four of these five mutations fail to increase drug resistance in some combinations. Pervasive biophysical pleiotropy within the beta-lactamase seems to be responsible...we conclude that much protein evolution will be similarly constrained" (emphasis added).20
The prohibitive force is sign epistasis, an extreme and open-ended form of context dependence. "Sign epistasis means," write the authors who coined the term, "that the sign of the fitness effect of a mutation is under epistatic control; thus, such a mutation is beneficial on some genetic backgrounds and deleterious on others."21 If evolution is so easily confounded by an antibiotic challenge, how much more stringently will it be constrained given the intuitively harder, multi-level challenges involved in creating the histone protein complex? Or various Cambrian organisms, for that matter? In the absence of relevant epistatic factors, who knows? Were evolutionary biologists not professionally committed to happy endings, they would acknowledge with grace that the modern protein theater emerged by means of conditions that we cannot specify in organisms whose nature we cannot imagine.
This is, after all, what Axe, Gauger and Stephen Meyer are saying. It is not all that they are saying; but à chaque jour suffit sa peine.
(1) Susumo Ohno, 'The notion of a pananimalia Genome,' Proceedings of the National Academy of Sciences, Vol. 93, pp. 8475-8478, August 1996. Ohno is well-known for his thesis that evolutionary change proceeds by gene duplication and divergence, what is now called the Ohno model. See Evolution by Gene Duplication, Springer-Verlag, 1970.
(2) Cyrus Chothia, Julian Gough, Christine Vogel, Sarah A. Teichmann, 'Evolution of the Protein Repertoire,' Science, Vol 300, 13 June 2003.
(3) There is no completely general and mathematically rigorous account of Darwinian theory, a point ceded by mathematical biologists on those occasions when they are free to whisper into one another's ears. "Darwin's theory of evolution by natural selection has obstinately remained in words since 1859. Of course, there are many mathematical models that show natural selection at work, but they are all examples. None claims to capture Darwin's central argument in its entirety." A. Grafen, 'The Formal Darwinism Project: A Midterm Report,' J Evol Biol. 2007 July 20(4):1243-54. The phrase 'obstinately remained in words' must be assigned its proper meaning: Obstinately remaining on the level of anecdote, example, and gossip.
(4) It is worth observing that Fred Hoyle anticipated Mike Behe's idea of an irreducibly complex system in his discussion of the Histone protein complex. See his Mathematics of Evolution, Acorn Enterprises LLC, 1999. Hoyle's derivation of the principles of population genetics is well-worth reading, especially his discussion of Kimura's diffusion equations. His book is, needless to say, widely inaccessible.
(5) Responding to skeptics at the Wistar Symposium such as Murray Eden and M.P. Schutzenberger, Sewall Wright appealed to Twenty Questions as evidence that they had overlooked a fast algorithm for the generation of complexity. It was not his finest moment.
(6) Michael Lynch, 'Scaling expectations for the time to establishment of complex adaptations,' PNAS, 2010,
Vol. 107 no. 38, p 1 of 6.
(7) Jerry Coyne, 'The Great Mutator,' The New Republic (June 14, 2007).
(8) Doug Axe & Ann Gauger, 'The Evolutionary Accessibility of New Enzyme Functions: A Case Study from the Biotin Pathway.' BIO-Complexity 2011(1):1-17.
(9) Mike Behe, The Edge of Evolution, The Free Press, 2007. Long ago? See Hoyle, op cit.
(10) It is not impossible to transform a base metal into gold -- just very difficult. From alchemy to atomic theory is a progression governed in part by a sliding parameter, one measuring the difficulty of atomic transmutation. The ancients thought it easy, the moderns think it hard. In evolutionary biology, it is the other way around.
(11) Long et al., 'The origin of new genes: glimpses from the young and old.' (2003), Nature Reviews Genetics, 4, 865-875 (November 2003). Darwin's Doubt contains a fine discussion of Long et al., one starting on p. 222. Of course it does.
(12) See Rita Ponce & Daniel L. Hartl, 'The evolution of the novel Sdic gene cluster in Drosophila melanogaster,' Gene 376 (2006) 174-183.
(13) Shu-Dan Yeh et al., 'Functional evidence that a recently evolved Drosophila sperm-specific gene boosts sperm competition.' PNAS February 7, 2012 vol. 109 no. 6 2043-2048.
(14) Hongsheng Dai et al., 'The evolution of courtship behaviors through the origination of a new gene in Drosophila,' Proc Natl Acad Sci U S A. 2008 May 27; 105(21): 7478-7483.
(15) That non-coding RNA may play any number of important roles in the cell is a separate issue, one no longer in doubt but equally of no relevance to the point under discussion.
(16) Jiaming Zhang et al., 'Evolving protein functional diversity in new genes of Drosophila,' PNAS November 16, 2004, Vol. 101, no. 46, 16246-16250.
(17) Internal references to illustrations have been deleted.
(18) Soskine, M. and Tawfik, D. 'Mutational effects and the evolution of new protein functions' Nature Reviews Genetics, 11 572-282. (August, 2010). "Moreover, for a significant fraction of proteins, increased dosages result in reduced fitness owing to undesirable promiscuous interactions driven by high protein concentrations or disturbed balance of protein complexes. Thus, although increased protein doses can make a weak, promiscuous activity come into action and thereby provide an evolutionary starting point, these increased doses may also become deleterious owing to the very same effect" (emphasis added). This might be called the Monkey's Paw effect in molecular genetics.
(19) Austin L. Hughes, 'Gene Duplication and the Origin of Novel Proteins,' PNAS June 21, 2005 vol. 102 no. 25 8791-8792. Hughes does mention an example of a happier ending: "Thus the report in this issue of PNAS by Tocchini-Valentini and colleagues on tRNA endonucleases of Archaea is particularly welcome as a concrete example of how new protein functions can arise." Can arise, note, not has arisen.
(20) Daniel M. Weinreich, Nigel F. Delaney, Mark A. DePristo, Daniel L. Hartl, 'Darwinian Evolution Can Follow Only Very Few Mutational Paths to Fitter Proteins.' Science Vol 312 7 April 2006.
(21) See D. M. Weinreich, R. A. Watson, L. Chao, 'Sign Epistasis and Genetic Constraints on Evolutionary Theory,' Evolution, Jun 2005, Vol. 59, Issue 6, 1165-1174. Weinreich et al. would appear to be appealing to the concept of a context sensitive grammar in order to account for, or describe, sign epistasis. A context sensitive grammar is one whose production rules are of the form αAβ → αγβ, where the derivation of γ from A depends on the flanking parameters α and β.