A Tale of Three Creationists, Part 3
Reasons to Believe, Scurvy, and a (Guinea) Pig in a Poke
After the recent exchanges between Fazale Rana of Reasons to Believe (RTB) and myself, I had intended to move on to other important issues. However, Rana’s final response to my critique is so inaccurate - and so badly misleads his audience - that to not reply would be irresponsible. There is no room for a “pig-in-a-poke” in science or the church.
If you have been following this series, you will know that one of the strongest pieces of evidence in favor of common ancestry between humans and chimpanzees (and other organisms) is a large number of “broken” genes, otherwise known as pseudogenes. When they are damaged in exactly the same way and exactly the same spot in closely related organisms, by far the simplest explanation is that they are copies of a single gene that was damaged in a common ancestor.
One way of illustrating this is to think of a 25 page manuscript: pretend that there is a mistake of one letter out of the 75,000 letters. Let’s say that a “g” had been inadvertently removed from the word “missing”, so that it now reads “missin.” Now let’s say that you collect 10 copies of this manuscript from various people and five of them have exactly the same mistake—the word “missing” on page 22, line 15, word 7, is spelled “missin” instead of missing. Every one of the other 74,999 letters in the 25 pages is perfect. Would you not conclude that the five identically mistaken manuscripts are derived from a single source? Would you not have to strain very hard to come up with an alternative explanation?
Rana is similarly desperate. The RTB model—based as it is on their specific interpretation of scripture— requires that they find an alternative to the common descent model overwhelmingly accepted by biologists. Rana feels a need to respond to pseudogene evidence since it so powerfully supports the creation of humans through an evolutionary process. The fact that different mammalian species, including humans, have many pseudogenes with multiple identical abnormalities (mutations) shared between them is a problem for any sort of non-evolutionary, special independent creation model. The fact that shared mutations are present in patterns that match other independent lines of evidence such as sequence identity (see here and here) and shared synteny (see here) only makes matters worse. The data from these different areas are congruent and I am not aware of any paper in the peer-reviewed scientific literature that attempts to argue for a non-evolutionary interpretation of these data.*
The topic of Rana’s rebuttal is a single portion (termed the tenth exon) of a single pseudogene, the GLO pseudogene, which, when not broken, makes an enzyme required for vitamin C biosynthesis. In “dry-nosed” primates (a group which includes humans,) this gene is “broken.” It is because of this that we get scurvy if we do not ingest enough vitamin C in our diet. Many of the mutations in the GLO pseudogene are shared between humans and other primates, indicating that these mutations have been inherited from a common ancestor. The only other possible explanation is that a large number of identical mutations occurred independently in many species – something so improbable as to be not worth considering. Yet this is exactly the strained interpretation that Rana chooses to argue for:
Comparison of the rat’s exon X (10th exon) DNA sequence with those in humans, chimpanzees, and orangutans reveals a number of the same mutations in the same locations for the primate sequences. Of particular significance is position 97 in which it appears as if a deletion took place in the primate sequences. Evolutionary biologists argue that this deletion and the other shared mutations are clear evidence for common ancestry, with these changes having occurred before apes diverged from Old World monkeys.
The RTB genomics model offers a different explanation for the similarities in the primate DNA sequences. The shared features are interpreted as the outworking of nonrandom, reproducible changes that happened independently in humans, chimpanzees, and orangutans.
Rana then goes on to claim that there is empirical evidence for the RTB view. The GLO gene is also a pseudogene in guinea pigs, but this species lost GLO function independently of the primate common ancestor. Not surprisingly, the mutations in the guinea pig pseudogene are different from those in the primate pseudogene, but Rana claims that there are a few shared mutations between primates and guinea pigs in exon X:
Support for this interpretation comes from comparisons of the primate exon X sequences with the corresponding region of the guinea pig GLO pseudogene. The structure of the guinea pig GLO pseudogene is dramatically different than that of the human pseudogene. Presumably, guinea pigs and primates lost this gene independently. If mutations were random, then few if any of the changes in the primate and guinea pig exon X sequences would be the same. Yet, as biologist Peter Borger points out, fifty percent of the mutations in the primate and guinea pig exon X sequence are identical. In addition, the guinea pig exon X region shows a mutation at position 97, the location in the primate genomes where a deletion took place. These shared features could not have resulted because guinea pigs and primates shared a common ancestor. Instead, they must reflect nonrandom, reproducible changes.
In other words, the RTB genomics model can reasonably account for the shared features of the GLO pseudogene in primates without resorting to common ancestry as the explanation.
There are several problems with this argument, but none would be apparent to a non-specialist. On the face of it, it seems like a strong case. If indeed there are shared mutations in two independent gene-to-pseudogene conversion events, it would lend (some) credence to Rana’s assertion that all shared mutations are the result of non-random events. The GLO pseudogene sequences in question are as follows. At six sites (in yellow) the human and guinea pig sequences match each other but differ from the rat (purple). The anti-evolutionary source Rana cites (PDF) claims that these sites are the result of independent mutation to the same nucleotide. (Position 97, which has a single-nucleotide deletion in humans and other primates, is shown in blue.)
Despite Rana’s insistence that there is only one possible explanation for this pattern, there are two possibilities. The possibility that Rana does not discuss is this: the human and guinea pig sequences match because they are not mutations, but rather the original ancestral sequence of the functional GLO gene that was present in the last human – guinea pig common ancestor. This would make the rat sequence at these six positions the mutations, not the human / guinea pig sequences. There is a simple way to test this hypothesis, of course: examine the GLO sequence of other mammalian species in this region. If many other species match the human and guinea pig sequences, then we know this sequence is the ancestral sequence. If many other species match the rat sequence, the rat sequence is ancestral and the evidence supports multiple independent mutations in the primate and guinea pig lineages. Unfortunately for RTB, even a cursory examination of other species in this region demonstrates that the human / guinea pig sequence is ancestral. At each of the six positions, the consensus sequence matches the human / guinea pig sequence. The so-called “shared mutations” are not mutations at all:
The RTB model thus has no support for the proposal that shared mutations observed between species are the result of non-random, parallel mutation events. As such, Rana’s assertion that the RTB model adequately accounts for shared mutations in pseudogenes is incorrect. Unitary pseudogenes with shared mutations remain an insurmountable problem for RTB.
Beyond the scientific flaws in this argument, what I found troubling about Rana’s response is that this analysis I have presented here is nothing new – it has been known for years (see for example here and here). Rana should have known about this. But even if he didn’t the fact remains that it would have been easy for Rana to test the claims himself. For example, I assembled the alignments in this post using public databases and search tools freely available on the web. (It would have been possible simply to use figures presented elsewhere, but I wanted to (a) confirm the data myself, and (b) ensure that the relevant sequences have not been updated with corrections in the last few years). If I can do this, Rana should be able to do so as well, especially before presenting flawed arguments to non-specialists (such as RTB supporters or public in general) who will take him at his word.
As the young earth creationist, Todd Wood, pointed out in the conclusion to his recent series on RTB, the reasons behind this unfortunate pattern in RTB scholarship are somewhat irrelevant. What matters is that believers know that until RTB makes amends, their model is untrustworthy. Indeed, if RTB had developed a scientifically credible model to explain shared pseudogenes from an anti-common descent perspective, Todd Wood would be very interested (since he, being a young earth creationist, does not believe in common descent). Todd’s view of the RTB model, however, agrees with mine - and he too has called on RTB to correct their errors:
Venema and I have documented a sad but consistent and ongoing pattern of erroneous summaries of published works on the part of RTB (Rana and Ross, but mostly Rana).There's really no way to deny these mistakes have been made or to explain them away, so what are you going to do about them? I recommend apologizing for the mistakes, correcting them if you can, and instituting some kind of serious fact-checking filter on everything you publish…
As far as I'm concerned, RTB's credibility is completely shot. I would recommend that no one accept any of RTB's arguments without fact-checking their claims first. I do not know whether these problems are due to lazy scholarship, ignorance, intentional deception, or ideological blinders. What I do know is that you cannot trust Reasons to Believe.
Hopefully in the coming days RTB will follow Todd’s advice. Please, Dr. Rana, for the sake of RTB’s followers (and the church as a whole), let your audience know that you have made a mistake. Let them know that the RTB model is flawed. Admit that it misrepresents well-established science. Then start the work of making it right. This, after all is how real science works. We admit when old ideas don’t work anymore and we move on.
I would also suggest that you surround yourself with people who have expertise in this area of biology. Please have your ideas thoroughly vetted by those who know the field. As your brother in Christ, I would be happy to work with you in arranging that. While I cannot speak for Todd personally, I have every confidence that he too would be willing to help with an honest overhaul of the RTB model. Without that (significant) overhaul, however, the RTB model is a (guinea) pig in a poke. Caveat emptor.
* - “I previously stated in this post that I was convinced that no geneticist at a secular, research university would see this data differently. I have now received what I consider reliable information that this is not the case, but rather that there are at least a few biologists at secular institutions who privately hold to a young earth creationist view. As such, they do see the data differently, though for religious reasons. I should have been more clear: I am convinced that there is no biologist at a secular institution who objects to a common ancestry interpretation of these data for scientific reasons. Certainly, if there is such an individual, they have not made their case in the scientific literature, the proper place for contesting current scientific consensus.”
Dennis Venema is Fellow of Biology for The BioLogos Foundation and associate professor of biology at Trinity Western University in Langley, British Columbia. His research is focused on the genetics of pattern formation and signalling.