A thread over at the Richard Dawkins forums got me thinking:
Creationism (which, despite the best efforts of its followers, includes “intelligent design”) is never, ever supported by evidence. Any alleged support for this idiocy is almost universally in the form of either argument from personal incredulity (“I can’t conceive of how life could have evolved from a single common ancestor, therefore god created life”) or baseless critique of evolutionary theory (and often of evolutionary theory that is woefully out of date).
All of this made me wonder, what evidence, if true, would falsify creationism for the creationist? After all, creationists have fought hard to get this stuff taught in science classes, and falsifiability is a key component of the scientific method. Thus, it seems logical to ask the creationist: what would you expect to observe if creationism is false?
I thought about this a little more (and posted some of this to the Dawkins forums), and here are my thoughts on the topic (coming from the point of view of a cell/molecular biologist, remember):
1. The genetic code
The genetic code allows us (and the ribosome) to translate the sequence of nucleotides in RNA (A, C, U, and G; which are themselves encoded by the sequence of nucleotides in DNA) into the amino acid sequence of a given protein. It is a triplet code, meaning that each amino acid is encoded by three consecutive nucleotides; the amino acid methionine (more on this in a moment) is encoded by the RNA sequence A-U-G (or, if you like, the DNA sequence A-T-G…thymine is replaced by uracil in RNA). Each set of three nucleotides in RNA is called a ‘codon’. The code is also redundant, meaning that a single amino acid can be encoded by several codons; the amino acid serine, for instance, is encoded by UCU, UCC, UCG, UCA, AGU, and AGC. The opposite, however, is never true; each codon codes for only one amino acid.
The intriguing thing here is that the code is universal (or very nearly so); the same three nucleotide codon mechanism is used in all organisms, and there exist only very minor, infrequent exceptions to the code itself. Bacteria use the same code as do humans (which allows us to use bacteria to produce human proteins for research or therapy), as do sharks, as do platypuses, as do lobsters.
The interesting thing to consider is why this should be so. Why should the code be universally applicable? There are two approaches to this question.
First, if all life originated from a common ancestor, we would expect that the code, which was laid down in the very earliest organisms which used the RNA-protein system, was simply used by all descended organisms, and that any divergence from the universal code should occur most in distantly related organisms; that is, the degree of similarity in the code should be greatest for those organisms that are most closely related to one another and least for those organisms that are less closely related to one another.
Alternatively, if life were created as we now see it (i.e. there is no common descent), then we would not expect to see a universal code; indeed, there is no a priori reason to expect a universal code. Why should lobster proteins be synthesized using the same code as human proteins? There is no advantage to having UUU encode phenylalanine and not serine (or vice versa), so why should the code be universal?
In fact, the code is very nearly universal, and, as predicted, the divergence from universality is correlated to the degree of relatedness of the organism; this should serve as the first falsification of creationism. Justifying a universal genetic code under creationism requires some sort of statement that “god is mysterious” or that there is some unique strength in the particular code we see. Of course, neither of these statements is supported by a lick of evidence, and so they get us nowhere.
The fact remains, a universal genetic code is a falsification of creationism.
2. Shared Genetic and Protein Homology
DNA contains the four bases A, T, C, and G, the sequence of which determines the sequence of the RNA that will be encoded by the DNA, and, through the genetic code, the protein that will ultimately be produced. As we have seen, however, the genetic code is highly redundant; if you were creating a DNA sequence to encode for a given gene (to ultimately encode for a given protein), you have options. Lets say you needed to put the amino acid serine into your growing protein. You could cause serine to be inserted into the growing protein by altering the DNA in a number of ways: TCT, TCC, TCG, TCA, AGT, and AGC; all of these encode serine.
Now, one might assume that if serine was required in a given protein in a specific organism that it would also be required in that same protein in other organisms (serine, for instance, is often modified by enzymes called kinases, which can drastically alter the function of a protein).
So, if we hypothesize that life was created as we see it (i.e. not descended from a common ancestor), we would expect to see, across species, a random allotment of these 6 serine codons. That is, if we accept that the presence of serine is required for the function of a given protein (which is well-documented), then the DNA sequences encoding that serine in different, allegedly unrelated organisms should show a random assortment of the six codons that encode serine; there is no reason to expect the TCT codon to occur more frequently than the TCC codon (or any other codon), since they both have exactly the same effect on the final product: the protein.
What we observe is quite different. Take the protein called ‘p53’, which is absolutely critical for the proper sensing and repair of damaged DNA, as well as for the death of individual cells when the damage is too severe to be fixed. The 15th and 20th amino acid of human p53 is serine; both are modified by phosphorylation, and both are required for the normal function of the protein.
Human serine 15 is encoded by the DNA sequence AGT; human serine 20 is encoded by TCA. Now let’s look at mouse p53. Mouse serine 15 (which is actually at position 18 in the mouse gene) is encoded by AGC; the T in human p53 has been changed to a C in the mouse p53, but this has no effect on the protein sequence; both encode serine. Mouse serine 20 (which is actually at position 23 in the mouse gene) is encoded by TCA, the same as the human sequence.
Now, let’s look at chimpanzee p53. Serine 15 is encoded by AGT, the same as human p53. Serine 20 is encoded by TCA, the same as human p53.
Do you see something happening here?
The human, mouse, and chimpanzee p53 proteins are functionally very similar (as evidenced by the over 43000 research papers published on the subject), and they all have conserved amino acids in very specific places (since these amino acids perform very specific functions). And yet, while all three forms encode serine 15 and serine 20 (the mouse serines are pushed forward through the insertion of three additional amino acids that are lacking in the human and chimp p53), the human and chimpanzee p53 are encoded by THE SAME DNA sequence, whereas the mouse p53 serine 15 (18) is encoded by a different DNA sequence altogether*.
If we repeat this exercise for other organisms, we see precisely the same thing; the degree of relatedness between the organism predicts the extent to which the SAME amino acid is encoded by the SAME DNA sequence. The protein sequence has to remain the same in order to maintain the same function, but the DNA sequence is under no such restriction; there are options at the DNA level.
If all life were created as is, we would expect a random allotment of DNA sequences making up the same codon. Thus, the observation that the degree of relatedness of two species predicts the extent to which a codon for a specific amino acid will be identical across the two species is a falsification of creationism. I have yet to hear even an attempt to rationalize this finding based on creationism.
3. Protein Homology in Organisms Sharing Similar Environments
One might argue that protein sequences in organisms sharing a common environment, particularly of proteins that are absolutely required for survival in that environment, should be very similar. If this were true, we would, for instance, expect that cytochrome c, a protein that is absolutely required for efficient energy metabolism, in bats would be very similar to that found in pigeons, since they both must meet the energy requirements of using flight as a primary means of locomotion. Similarly, we would expect that bat cytochrome c would be dissimilar to that found in humans, since the energy requirements created by the environments in which the two organisms live is likewise dissimilar.
That is to say, if life was created as we see it, we would expect that the sequence of proteins required for survival in a given environment for a given organism (irrespective of those factors I have discussed above) would be more similar in other organisms living in the same environment than in those who live in entirely disparate environments.
In fact, this is not at all what we observe. Bat cytochrome c is more homologous to whale and human cytochrome c than it is to pigeon cytochrome c, or to cytochrome c observed in other birds. This is a falsification of creationism.
Note that when I say “falsification of creationism”, I don’t commit the creationist’s blunder of claiming that this is evidence of common descent; evolution by common descent, like all scientific theories, requires positive evidence in its favour. It just so happens that, taken together, the data I’ve shown here are highly consistent with the theory that all life is descended from a common ancestor. But remember, it is a false dichotomy to state that evidence against one theory is evidence for another theory, unless the two theories are the only possible ones.
The next time a creationist provides you with some ridiculous criticism of evolution, take the following approach. First, if you can, respond to the criticism and set the creationist straight. Now that probably wont work (or you may not have the detailed scientific knowledge to respond immediately), so you might also want to point them to the appropriate authorities on the subject (this is really good start, which I always recommend). But don’t let the conversation end without asking them for an example of a piece of data that, if true, would falsify creationism for them. If they can’t answer, they’re not talking science, and they’re only interested in dogma. At this point, you can safely walk away, thus protecting your brain from further idiocy-induced atrophy.
* – For those who are really interested in the evolutionary biology here, it follows that serine 20 is better conserved than serine 15; serine 20 (23 in mouse) lies within the binding site for a protein called HDM2 (or, more commonly, MDM2), which inhibits the function of p53. When serine 20 is modified by the addition of a phosphate group, this inhibits the binding of MDM2, which promotes the function of p53 (by inhibiting the inhibitor). MDM2 binding is a VERY important process for the regulation of DNA damage sensing, DNA repair, and apoptosis (cell death), and thus serine 20 is likely to be VERY well conserved by natural selection. Serine 15, by contrast, may be somewhat less important in this regard, and thus somewhat more poorly conserved…I haven’t done any detailed studies to look at this, but it could make an interesting paper…